diff options
Diffstat (limited to 'drivers/net/ethernet/netronome/nfp')
42 files changed, 4275 insertions, 1700 deletions
diff --git a/drivers/net/ethernet/netronome/nfp/Makefile b/drivers/net/ethernet/netronome/nfp/Makefile index 4afb10375397..47c708f08ade 100644 --- a/drivers/net/ethernet/netronome/nfp/Makefile +++ b/drivers/net/ethernet/netronome/nfp/Makefile @@ -56,7 +56,9 @@ endif ifeq ($(CONFIG_NFP_APP_ABM_NIC),y) nfp-objs += \ + abm/cls.o \ abm/ctrl.o \ + abm/qdisc.o \ abm/main.o endif diff --git a/drivers/net/ethernet/netronome/nfp/abm/cls.c b/drivers/net/ethernet/netronome/nfp/abm/cls.c new file mode 100644 index 000000000000..ff3913085665 --- /dev/null +++ b/drivers/net/ethernet/netronome/nfp/abm/cls.c @@ -0,0 +1,283 @@ +// SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +/* Copyright (C) 2018 Netronome Systems, Inc. */ + +#include <linux/bitfield.h> +#include <net/pkt_cls.h> + +#include "../nfpcore/nfp_cpp.h" +#include "../nfp_app.h" +#include "../nfp_net_repr.h" +#include "main.h" + +struct nfp_abm_u32_match { + u32 handle; + u32 band; + u8 mask; + u8 val; + struct list_head list; +}; + +static bool +nfp_abm_u32_check_knode(struct nfp_abm *abm, struct tc_cls_u32_knode *knode, + __be16 proto, struct netlink_ext_ack *extack) +{ + struct tc_u32_key *k; + unsigned int tos_off; + + if (knode->exts && tcf_exts_has_actions(knode->exts)) { + NL_SET_ERR_MSG_MOD(extack, "action offload not supported"); + return false; + } + if (knode->link_handle) { + NL_SET_ERR_MSG_MOD(extack, "linking not supported"); + return false; + } + if (knode->sel->flags != TC_U32_TERMINAL) { + NL_SET_ERR_MSG_MOD(extack, + "flags must be equal to TC_U32_TERMINAL"); + return false; + } + if (knode->sel->off || knode->sel->offshift || knode->sel->offmask || + knode->sel->offoff || knode->fshift) { + NL_SET_ERR_MSG_MOD(extack, "variable offsetting not supported"); + return false; + } + if (knode->sel->hoff || knode->sel->hmask) { + NL_SET_ERR_MSG_MOD(extack, "hashing not supported"); + return false; + } + if (knode->val || knode->mask) { + NL_SET_ERR_MSG_MOD(extack, "matching on mark not supported"); + return false; + } + if (knode->res && knode->res->class) { + NL_SET_ERR_MSG_MOD(extack, "setting non-0 class not supported"); + return false; + } + if (knode->res && knode->res->classid >= abm->num_bands) { + NL_SET_ERR_MSG_MOD(extack, + "classid higher than number of bands"); + return false; + } + if (knode->sel->nkeys != 1) { + NL_SET_ERR_MSG_MOD(extack, "exactly one key required"); + return false; + } + + switch (proto) { + case htons(ETH_P_IP): + tos_off = 16; + break; + case htons(ETH_P_IPV6): + tos_off = 20; + break; + default: + NL_SET_ERR_MSG_MOD(extack, "only IP and IPv6 supported as filter protocol"); + return false; + } + + k = &knode->sel->keys[0]; + if (k->offmask) { + NL_SET_ERR_MSG_MOD(extack, "offset mask - variable offsetting not supported"); + return false; + } + if (k->off) { + NL_SET_ERR_MSG_MOD(extack, "only DSCP fields can be matched"); + return false; + } + if (k->val & ~k->mask) { + NL_SET_ERR_MSG_MOD(extack, "mask does not cover the key"); + return false; + } + if (be32_to_cpu(k->mask) >> tos_off & ~abm->dscp_mask) { + NL_SET_ERR_MSG_MOD(extack, "only high DSCP class selector bits can be used"); + nfp_err(abm->app->cpp, + "u32 offload: requested mask %x FW can support only %x\n", + be32_to_cpu(k->mask) >> tos_off, abm->dscp_mask); + return false; + } + + return true; +} + +/* This filter list -> map conversion is O(n * m), we expect single digit or + * low double digit number of prios and likewise for the filters. Also u32 + * doesn't report stats, so it's really only setup time cost. + */ +static unsigned int +nfp_abm_find_band_for_prio(struct nfp_abm_link *alink, unsigned int prio) +{ + struct nfp_abm_u32_match *iter; + + list_for_each_entry(iter, &alink->dscp_map, list) + if ((prio & iter->mask) == iter->val) + return iter->band; + + return alink->def_band; +} + +static int nfp_abm_update_band_map(struct nfp_abm_link *alink) +{ + unsigned int i, bits_per_prio, prios_per_word, base_shift; + struct nfp_abm *abm = alink->abm; + u32 field_mask; + + alink->has_prio = !list_empty(&alink->dscp_map); + + bits_per_prio = roundup_pow_of_two(order_base_2(abm->num_bands)); + field_mask = (1 << bits_per_prio) - 1; + prios_per_word = sizeof(u32) * BITS_PER_BYTE / bits_per_prio; + + /* FW mask applies from top bits */ + base_shift = 8 - order_base_2(abm->num_prios); + + for (i = 0; i < abm->num_prios; i++) { + unsigned int offset; + u32 *word; + u8 band; + + word = &alink->prio_map[i / prios_per_word]; + offset = (i % prios_per_word) * bits_per_prio; + + band = nfp_abm_find_band_for_prio(alink, i << base_shift); + + *word &= ~(field_mask << offset); + *word |= band << offset; + } + + /* Qdisc offload status may change if has_prio changed */ + nfp_abm_qdisc_offload_update(alink); + + return nfp_abm_ctrl_prio_map_update(alink, alink->prio_map); +} + +static void +nfp_abm_u32_knode_delete(struct nfp_abm_link *alink, + struct tc_cls_u32_knode *knode) +{ + struct nfp_abm_u32_match *iter; + + list_for_each_entry(iter, &alink->dscp_map, list) + if (iter->handle == knode->handle) { + list_del(&iter->list); + kfree(iter); + nfp_abm_update_band_map(alink); + return; + } +} + +static int +nfp_abm_u32_knode_replace(struct nfp_abm_link *alink, + struct tc_cls_u32_knode *knode, + __be16 proto, struct netlink_ext_ack *extack) +{ + struct nfp_abm_u32_match *match = NULL, *iter; + unsigned int tos_off; + u8 mask, val; + int err; + + if (!nfp_abm_u32_check_knode(alink->abm, knode, proto, extack)) + goto err_delete; + + tos_off = proto == htons(ETH_P_IP) ? 16 : 20; + + /* Extract the DSCP Class Selector bits */ + val = be32_to_cpu(knode->sel->keys[0].val) >> tos_off & 0xff; + mask = be32_to_cpu(knode->sel->keys[0].mask) >> tos_off & 0xff; + + /* Check if there is no conflicting mapping and find match by handle */ + list_for_each_entry(iter, &alink->dscp_map, list) { + u32 cmask; + + if (iter->handle == knode->handle) { + match = iter; + continue; + } + + cmask = iter->mask & mask; + if ((iter->val & cmask) == (val & cmask) && + iter->band != knode->res->classid) { + NL_SET_ERR_MSG_MOD(extack, "conflict with already offloaded filter"); + goto err_delete; + } + } + + if (!match) { + match = kzalloc(sizeof(*match), GFP_KERNEL); + if (!match) + return -ENOMEM; + list_add(&match->list, &alink->dscp_map); + } + match->handle = knode->handle; + match->band = knode->res->classid; + match->mask = mask; + match->val = val; + + err = nfp_abm_update_band_map(alink); + if (err) + goto err_delete; + + return 0; + +err_delete: + nfp_abm_u32_knode_delete(alink, knode); + return -EOPNOTSUPP; +} + +static int nfp_abm_setup_tc_block_cb(enum tc_setup_type type, + void *type_data, void *cb_priv) +{ + struct tc_cls_u32_offload *cls_u32 = type_data; + struct nfp_repr *repr = cb_priv; + struct nfp_abm_link *alink; + + alink = repr->app_priv; + + if (type != TC_SETUP_CLSU32) { + NL_SET_ERR_MSG_MOD(cls_u32->common.extack, + "only offload of u32 classifier supported"); + return -EOPNOTSUPP; + } + if (!tc_cls_can_offload_and_chain0(repr->netdev, &cls_u32->common)) + return -EOPNOTSUPP; + + if (cls_u32->common.protocol != htons(ETH_P_IP) && + cls_u32->common.protocol != htons(ETH_P_IPV6)) { + NL_SET_ERR_MSG_MOD(cls_u32->common.extack, + "only IP and IPv6 supported as filter protocol"); + return -EOPNOTSUPP; + } + + switch (cls_u32->command) { + case TC_CLSU32_NEW_KNODE: + case TC_CLSU32_REPLACE_KNODE: + return nfp_abm_u32_knode_replace(alink, &cls_u32->knode, + cls_u32->common.protocol, + cls_u32->common.extack); + case TC_CLSU32_DELETE_KNODE: + nfp_abm_u32_knode_delete(alink, &cls_u32->knode); + return 0; + default: + return -EOPNOTSUPP; + } +} + +int nfp_abm_setup_cls_block(struct net_device *netdev, struct nfp_repr *repr, + struct tc_block_offload *f) +{ + if (f->binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS) + return -EOPNOTSUPP; + + switch (f->command) { + case TC_BLOCK_BIND: + return tcf_block_cb_register(f->block, + nfp_abm_setup_tc_block_cb, + repr, repr, f->extack); + case TC_BLOCK_UNBIND: + tcf_block_cb_unregister(f->block, nfp_abm_setup_tc_block_cb, + repr); + return 0; + default: + return -EOPNOTSUPP; + } +} diff --git a/drivers/net/ethernet/netronome/nfp/abm/ctrl.c b/drivers/net/ethernet/netronome/nfp/abm/ctrl.c index 3c661f422688..9584f03f3efa 100644 --- a/drivers/net/ethernet/netronome/nfp/abm/ctrl.c +++ b/drivers/net/ethernet/netronome/nfp/abm/ctrl.c @@ -1,7 +1,9 @@ // SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) /* Copyright (C) 2018 Netronome Systems, Inc. */ +#include <linux/bitops.h> #include <linux/kernel.h> +#include <linux/log2.h> #include "../nfpcore/nfp_cpp.h" #include "../nfpcore/nfp_nffw.h" @@ -11,38 +13,58 @@ #include "../nfp_net.h" #include "main.h" -#define NFP_QLVL_SYM_NAME "_abi_nfd_out_q_lvls_%u" +#define NFP_NUM_PRIOS_SYM_NAME "_abi_pci_dscp_num_prio_%u" +#define NFP_NUM_BANDS_SYM_NAME "_abi_pci_dscp_num_band_%u" +#define NFP_ACT_MASK_SYM_NAME "_abi_nfd_out_q_actions_%u" + +#define NFP_RED_SUPPORT_SYM_NAME "_abi_nfd_out_red_offload_%u" + +#define NFP_QLVL_SYM_NAME "_abi_nfd_out_q_lvls_%u%s" #define NFP_QLVL_STRIDE 16 #define NFP_QLVL_BLOG_BYTES 0 #define NFP_QLVL_BLOG_PKTS 4 #define NFP_QLVL_THRS 8 +#define NFP_QLVL_ACT 12 -#define NFP_QMSTAT_SYM_NAME "_abi_nfdqm%u_stats" +#define NFP_QMSTAT_SYM_NAME "_abi_nfdqm%u_stats%s" #define NFP_QMSTAT_STRIDE 32 #define NFP_QMSTAT_NON_STO 0 #define NFP_QMSTAT_STO 8 #define NFP_QMSTAT_DROP 16 #define NFP_QMSTAT_ECN 24 +#define NFP_Q_STAT_SYM_NAME "_abi_nfd_rxq_stats%u%s" +#define NFP_Q_STAT_STRIDE 16 +#define NFP_Q_STAT_PKTS 0 +#define NFP_Q_STAT_BYTES 8 + +#define NFP_NET_ABM_MBOX_CMD NFP_NET_CFG_MBOX_SIMPLE_CMD +#define NFP_NET_ABM_MBOX_RET NFP_NET_CFG_MBOX_SIMPLE_RET +#define NFP_NET_ABM_MBOX_DATALEN NFP_NET_CFG_MBOX_SIMPLE_VAL +#define NFP_NET_ABM_MBOX_RESERVED (NFP_NET_CFG_MBOX_SIMPLE_VAL + 4) +#define NFP_NET_ABM_MBOX_DATA (NFP_NET_CFG_MBOX_SIMPLE_VAL + 8) + static int nfp_abm_ctrl_stat(struct nfp_abm_link *alink, const struct nfp_rtsym *sym, - unsigned int stride, unsigned int offset, unsigned int i, - bool is_u64, u64 *res) + unsigned int stride, unsigned int offset, unsigned int band, + unsigned int queue, bool is_u64, u64 *res) { struct nfp_cpp *cpp = alink->abm->app->cpp; u64 val, sym_offset; + unsigned int qid; u32 val32; int err; - sym_offset = (alink->queue_base + i) * stride + offset; + qid = band * NFP_NET_MAX_RX_RINGS + alink->queue_base + queue; + + sym_offset = qid * stride + offset; if (is_u64) err = __nfp_rtsym_readq(cpp, sym, 3, 0, sym_offset, &val); else err = __nfp_rtsym_readl(cpp, sym, 3, 0, sym_offset, &val32); if (err) { - nfp_err(cpp, - "RED offload reading stat failed on vNIC %d queue %d\n", - alink->id, i); + nfp_err(cpp, "RED offload reading stat failed on vNIC %d band %d queue %d (+ %d)\n", + alink->id, band, queue, alink->queue_base); return err; } @@ -50,175 +72,179 @@ nfp_abm_ctrl_stat(struct nfp_abm_link *alink, const struct nfp_rtsym *sym, return 0; } -static int -nfp_abm_ctrl_stat_all(struct nfp_abm_link *alink, const struct nfp_rtsym *sym, - unsigned int stride, unsigned int offset, bool is_u64, - u64 *res) +int __nfp_abm_ctrl_set_q_lvl(struct nfp_abm *abm, unsigned int id, u32 val) { - u64 val, sum = 0; - unsigned int i; + struct nfp_cpp *cpp = abm->app->cpp; + u64 sym_offset; int err; - for (i = 0; i < alink->vnic->max_rx_rings; i++) { - err = nfp_abm_ctrl_stat(alink, sym, stride, offset, i, - is_u64, &val); - if (err) - return err; - sum += val; + __clear_bit(id, abm->threshold_undef); + if (abm->thresholds[id] == val) + return 0; + + sym_offset = id * NFP_QLVL_STRIDE + NFP_QLVL_THRS; + err = __nfp_rtsym_writel(cpp, abm->q_lvls, 4, 0, sym_offset, val); + if (err) { + nfp_err(cpp, + "RED offload setting level failed on subqueue %d\n", + id); + return err; } - *res = sum; + abm->thresholds[id] = val; return 0; } -int nfp_abm_ctrl_set_q_lvl(struct nfp_abm_link *alink, unsigned int i, u32 val) +int nfp_abm_ctrl_set_q_lvl(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, u32 val) { - struct nfp_cpp *cpp = alink->abm->app->cpp; + unsigned int threshold; + + threshold = band * NFP_NET_MAX_RX_RINGS + alink->queue_base + queue; + + return __nfp_abm_ctrl_set_q_lvl(alink->abm, threshold, val); +} + +int __nfp_abm_ctrl_set_q_act(struct nfp_abm *abm, unsigned int id, + enum nfp_abm_q_action act) +{ + struct nfp_cpp *cpp = abm->app->cpp; u64 sym_offset; int err; - sym_offset = (alink->queue_base + i) * NFP_QLVL_STRIDE + NFP_QLVL_THRS; - err = __nfp_rtsym_writel(cpp, alink->abm->q_lvls, 4, 0, - sym_offset, val); + if (abm->actions[id] == act) + return 0; + + sym_offset = id * NFP_QLVL_STRIDE + NFP_QLVL_ACT; + err = __nfp_rtsym_writel(cpp, abm->q_lvls, 4, 0, sym_offset, act); if (err) { - nfp_err(cpp, "RED offload setting level failed on vNIC %d queue %d\n", - alink->id, i); + nfp_err(cpp, + "RED offload setting action failed on subqueue %d\n", + id); return err; } + abm->actions[id] = act; return 0; } -int nfp_abm_ctrl_set_all_q_lvls(struct nfp_abm_link *alink, u32 val) +int nfp_abm_ctrl_set_q_act(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, enum nfp_abm_q_action act) +{ + unsigned int qid; + + qid = band * NFP_NET_MAX_RX_RINGS + alink->queue_base + queue; + + return __nfp_abm_ctrl_set_q_act(alink->abm, qid, act); +} + +u64 nfp_abm_ctrl_stat_non_sto(struct nfp_abm_link *alink, unsigned int queue) { - int i, err; + unsigned int band; + u64 val, sum = 0; - for (i = 0; i < alink->vnic->max_rx_rings; i++) { - err = nfp_abm_ctrl_set_q_lvl(alink, i, val); - if (err) - return err; + for (band = 0; band < alink->abm->num_bands; band++) { + if (nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, + NFP_QMSTAT_STRIDE, NFP_QMSTAT_NON_STO, + band, queue, true, &val)) + return 0; + sum += val; } - return 0; + return sum; } -u64 nfp_abm_ctrl_stat_non_sto(struct nfp_abm_link *alink, unsigned int i) +u64 nfp_abm_ctrl_stat_sto(struct nfp_abm_link *alink, unsigned int queue) { - u64 val; + unsigned int band; + u64 val, sum = 0; - if (nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, NFP_QMSTAT_STRIDE, - NFP_QMSTAT_NON_STO, i, true, &val)) - return 0; - return val; + for (band = 0; band < alink->abm->num_bands; band++) { + if (nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, + NFP_QMSTAT_STRIDE, NFP_QMSTAT_STO, + band, queue, true, &val)) + return 0; + sum += val; + } + + return sum; } -u64 nfp_abm_ctrl_stat_sto(struct nfp_abm_link *alink, unsigned int i) +static int +nfp_abm_ctrl_stat_basic(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, unsigned int off, u64 *val) { - u64 val; + if (!nfp_abm_has_prio(alink->abm)) { + if (!band) { + unsigned int id = alink->queue_base + queue; + + *val = nn_readq(alink->vnic, + NFP_NET_CFG_RXR_STATS(id) + off); + } else { + *val = 0; + } - if (nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, NFP_QMSTAT_STRIDE, - NFP_QMSTAT_STO, i, true, &val)) return 0; - return val; + } else { + return nfp_abm_ctrl_stat(alink, alink->abm->q_stats, + NFP_Q_STAT_STRIDE, off, band, queue, + true, val); + } } -int nfp_abm_ctrl_read_q_stats(struct nfp_abm_link *alink, unsigned int i, - struct nfp_alink_stats *stats) +int nfp_abm_ctrl_read_q_stats(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, struct nfp_alink_stats *stats) { int err; - stats->tx_pkts = nn_readq(alink->vnic, NFP_NET_CFG_RXR_STATS(i)); - stats->tx_bytes = nn_readq(alink->vnic, NFP_NET_CFG_RXR_STATS(i) + 8); + err = nfp_abm_ctrl_stat_basic(alink, band, queue, NFP_Q_STAT_PKTS, + &stats->tx_pkts); + if (err) + return err; - err = nfp_abm_ctrl_stat(alink, alink->abm->q_lvls, - NFP_QLVL_STRIDE, NFP_QLVL_BLOG_BYTES, - i, false, &stats->backlog_bytes); + err = nfp_abm_ctrl_stat_basic(alink, band, queue, NFP_Q_STAT_BYTES, + &stats->tx_bytes); + if (err) + return err; + + err = nfp_abm_ctrl_stat(alink, alink->abm->q_lvls, NFP_QLVL_STRIDE, + NFP_QLVL_BLOG_BYTES, band, queue, false, + &stats->backlog_bytes); if (err) return err; err = nfp_abm_ctrl_stat(alink, alink->abm->q_lvls, NFP_QLVL_STRIDE, NFP_QLVL_BLOG_PKTS, - i, false, &stats->backlog_pkts); + band, queue, false, &stats->backlog_pkts); if (err) return err; err = nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, NFP_QMSTAT_STRIDE, NFP_QMSTAT_DROP, - i, true, &stats->drops); + band, queue, true, &stats->drops); if (err) return err; return nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, NFP_QMSTAT_STRIDE, NFP_QMSTAT_ECN, - i, true, &stats->overlimits); + band, queue, true, &stats->overlimits); } -int nfp_abm_ctrl_read_stats(struct nfp_abm_link *alink, - struct nfp_alink_stats *stats) -{ - u64 pkts = 0, bytes = 0; - int i, err; - - for (i = 0; i < alink->vnic->max_rx_rings; i++) { - pkts += nn_readq(alink->vnic, NFP_NET_CFG_RXR_STATS(i)); - bytes += nn_readq(alink->vnic, NFP_NET_CFG_RXR_STATS(i) + 8); - } - stats->tx_pkts = pkts; - stats->tx_bytes = bytes; - - err = nfp_abm_ctrl_stat_all(alink, alink->abm->q_lvls, - NFP_QLVL_STRIDE, NFP_QLVL_BLOG_BYTES, - false, &stats->backlog_bytes); - if (err) - return err; - - err = nfp_abm_ctrl_stat_all(alink, alink->abm->q_lvls, - NFP_QLVL_STRIDE, NFP_QLVL_BLOG_PKTS, - false, &stats->backlog_pkts); - if (err) - return err; - - err = nfp_abm_ctrl_stat_all(alink, alink->abm->qm_stats, - NFP_QMSTAT_STRIDE, NFP_QMSTAT_DROP, - true, &stats->drops); - if (err) - return err; - - return nfp_abm_ctrl_stat_all(alink, alink->abm->qm_stats, - NFP_QMSTAT_STRIDE, NFP_QMSTAT_ECN, - true, &stats->overlimits); -} - -int nfp_abm_ctrl_read_q_xstats(struct nfp_abm_link *alink, unsigned int i, +int nfp_abm_ctrl_read_q_xstats(struct nfp_abm_link *alink, + unsigned int band, unsigned int queue, struct nfp_alink_xstats *xstats) { int err; err = nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, NFP_QMSTAT_STRIDE, NFP_QMSTAT_DROP, - i, true, &xstats->pdrop); + band, queue, true, &xstats->pdrop); if (err) return err; return nfp_abm_ctrl_stat(alink, alink->abm->qm_stats, NFP_QMSTAT_STRIDE, NFP_QMSTAT_ECN, - i, true, &xstats->ecn_marked); -} - -int nfp_abm_ctrl_read_xstats(struct nfp_abm_link *alink, - struct nfp_alink_xstats *xstats) -{ - int err; - - err = nfp_abm_ctrl_stat_all(alink, alink->abm->qm_stats, - NFP_QMSTAT_STRIDE, NFP_QMSTAT_DROP, - true, &xstats->pdrop); - if (err) - return err; - - return nfp_abm_ctrl_stat_all(alink, alink->abm->qm_stats, - NFP_QMSTAT_STRIDE, NFP_QMSTAT_ECN, - true, &xstats->ecn_marked); + band, queue, true, &xstats->ecn_marked); } int nfp_abm_ctrl_qm_enable(struct nfp_abm *abm) @@ -233,10 +259,64 @@ int nfp_abm_ctrl_qm_disable(struct nfp_abm *abm) NULL, 0, NULL, 0); } -void nfp_abm_ctrl_read_params(struct nfp_abm_link *alink) +int nfp_abm_ctrl_prio_map_update(struct nfp_abm_link *alink, u32 *packed) +{ + struct nfp_net *nn = alink->vnic; + unsigned int i; + int err; + + /* Write data_len and wipe reserved */ + nn_writeq(nn, nn->tlv_caps.mbox_off + NFP_NET_ABM_MBOX_DATALEN, + alink->abm->prio_map_len); + + for (i = 0; i < alink->abm->prio_map_len; i += sizeof(u32)) + nn_writel(nn, nn->tlv_caps.mbox_off + NFP_NET_ABM_MBOX_DATA + i, + packed[i / sizeof(u32)]); + + err = nfp_net_reconfig_mbox(nn, + NFP_NET_CFG_MBOX_CMD_PCI_DSCP_PRIOMAP_SET); + if (err) + nfp_err(alink->abm->app->cpp, + "setting DSCP -> VQ map failed with error %d\n", err); + return err; +} + +static int nfp_abm_ctrl_prio_check_params(struct nfp_abm_link *alink) +{ + struct nfp_abm *abm = alink->abm; + struct nfp_net *nn = alink->vnic; + unsigned int min_mbox_sz; + + if (!nfp_abm_has_prio(alink->abm)) + return 0; + + min_mbox_sz = NFP_NET_ABM_MBOX_DATA + alink->abm->prio_map_len; + if (nn->tlv_caps.mbox_len < min_mbox_sz) { + nfp_err(abm->app->pf->cpp, "vNIC mailbox too small for prio offload: %u, need: %u\n", + nn->tlv_caps.mbox_len, min_mbox_sz); + return -EINVAL; + } + + return 0; +} + +int nfp_abm_ctrl_read_params(struct nfp_abm_link *alink) { alink->queue_base = nn_readl(alink->vnic, NFP_NET_CFG_START_RXQ); alink->queue_base /= alink->vnic->stride_rx; + + return nfp_abm_ctrl_prio_check_params(alink); +} + +static unsigned int nfp_abm_ctrl_prio_map_size(struct nfp_abm *abm) +{ + unsigned int size; + + size = roundup_pow_of_two(order_base_2(abm->num_bands)); + size = DIV_ROUND_UP(size * abm->num_prios, BITS_PER_BYTE); + size = round_up(size, sizeof(u32)); + + return size; } static const struct nfp_rtsym * @@ -260,33 +340,86 @@ nfp_abm_ctrl_find_rtsym(struct nfp_pf *pf, const char *name, unsigned int size) } static const struct nfp_rtsym * -nfp_abm_ctrl_find_q_rtsym(struct nfp_pf *pf, const char *name, - unsigned int size) +nfp_abm_ctrl_find_q_rtsym(struct nfp_abm *abm, const char *name_fmt, + size_t size) { - return nfp_abm_ctrl_find_rtsym(pf, name, size * NFP_NET_MAX_RX_RINGS); + char pf_symbol[64]; + + size = array3_size(size, abm->num_bands, NFP_NET_MAX_RX_RINGS); + snprintf(pf_symbol, sizeof(pf_symbol), name_fmt, + abm->pf_id, nfp_abm_has_prio(abm) ? "_per_band" : ""); + + return nfp_abm_ctrl_find_rtsym(abm->app->pf, pf_symbol, size); } int nfp_abm_ctrl_find_addrs(struct nfp_abm *abm) { struct nfp_pf *pf = abm->app->pf; const struct nfp_rtsym *sym; - unsigned int pf_id; - char pf_symbol[64]; + int res; + + abm->pf_id = nfp_cppcore_pcie_unit(pf->cpp); + + /* Check if Qdisc offloads are supported */ + res = nfp_pf_rtsym_read_optional(pf, NFP_RED_SUPPORT_SYM_NAME, 1); + if (res < 0) + return res; + abm->red_support = res; + + /* Read count of prios and prio bands */ + res = nfp_pf_rtsym_read_optional(pf, NFP_NUM_BANDS_SYM_NAME, 1); + if (res < 0) + return res; + abm->num_bands = res; + + res = nfp_pf_rtsym_read_optional(pf, NFP_NUM_PRIOS_SYM_NAME, 1); + if (res < 0) + return res; + abm->num_prios = res; + + /* Read available actions */ + res = nfp_pf_rtsym_read_optional(pf, NFP_ACT_MASK_SYM_NAME, + BIT(NFP_ABM_ACT_MARK_DROP)); + if (res < 0) + return res; + abm->action_mask = res; + + abm->prio_map_len = nfp_abm_ctrl_prio_map_size(abm); + abm->dscp_mask = GENMASK(7, 8 - order_base_2(abm->num_prios)); + + /* Check values are sane, U16_MAX is arbitrarily chosen as max */ + if (!is_power_of_2(abm->num_bands) || !is_power_of_2(abm->num_prios) || + abm->num_bands > U16_MAX || abm->num_prios > U16_MAX || + (abm->num_bands == 1) != (abm->num_prios == 1)) { + nfp_err(pf->cpp, + "invalid priomap description num bands: %u and num prios: %u\n", + abm->num_bands, abm->num_prios); + return -EINVAL; + } - pf_id = nfp_cppcore_pcie_unit(pf->cpp); - abm->pf_id = pf_id; + /* Find level and stat symbols */ + if (!abm->red_support) + return 0; - snprintf(pf_symbol, sizeof(pf_symbol), NFP_QLVL_SYM_NAME, pf_id); - sym = nfp_abm_ctrl_find_q_rtsym(pf, pf_symbol, NFP_QLVL_STRIDE); + sym = nfp_abm_ctrl_find_q_rtsym(abm, NFP_QLVL_SYM_NAME, + NFP_QLVL_STRIDE); if (IS_ERR(sym)) return PTR_ERR(sym); abm->q_lvls = sym; - snprintf(pf_symbol, sizeof(pf_symbol), NFP_QMSTAT_SYM_NAME, pf_id); - sym = nfp_abm_ctrl_find_q_rtsym(pf, pf_symbol, NFP_QMSTAT_STRIDE); + sym = nfp_abm_ctrl_find_q_rtsym(abm, NFP_QMSTAT_SYM_NAME, + NFP_QMSTAT_STRIDE); if (IS_ERR(sym)) return PTR_ERR(sym); abm->qm_stats = sym; + if (nfp_abm_has_prio(abm)) { + sym = nfp_abm_ctrl_find_q_rtsym(abm, NFP_Q_STAT_SYM_NAME, + NFP_Q_STAT_STRIDE); + if (IS_ERR(sym)) + return PTR_ERR(sym); + abm->q_stats = sym; + } + return 0; } diff --git a/drivers/net/ethernet/netronome/nfp/abm/main.c b/drivers/net/ethernet/netronome/nfp/abm/main.c index c0830c0c2c3f..4d4ff5844c47 100644 --- a/drivers/net/ethernet/netronome/nfp/abm/main.c +++ b/drivers/net/ethernet/netronome/nfp/abm/main.c @@ -2,14 +2,13 @@ /* Copyright (C) 2018 Netronome Systems, Inc. */ #include <linux/bitfield.h> +#include <linux/bitmap.h> #include <linux/etherdevice.h> #include <linux/lockdep.h> #include <linux/netdevice.h> #include <linux/rcupdate.h> +#include <linux/rtnetlink.h> #include <linux/slab.h> -#include <net/pkt_cls.h> -#include <net/pkt_sched.h> -#include <net/red.h> #include "../nfpcore/nfp.h" #include "../nfpcore/nfp_cpp.h" @@ -28,269 +27,6 @@ static u32 nfp_abm_portid(enum nfp_repr_type rtype, unsigned int id) } static int -__nfp_abm_reset_root(struct net_device *netdev, struct nfp_abm_link *alink, - u32 handle, unsigned int qs, u32 init_val) -{ - struct nfp_port *port = nfp_port_from_netdev(netdev); - int ret; - - ret = nfp_abm_ctrl_set_all_q_lvls(alink, init_val); - memset(alink->qdiscs, 0, sizeof(*alink->qdiscs) * alink->num_qdiscs); - - alink->parent = handle; - alink->num_qdiscs = qs; - port->tc_offload_cnt = qs; - - return ret; -} - -static void -nfp_abm_reset_root(struct net_device *netdev, struct nfp_abm_link *alink, - u32 handle, unsigned int qs) -{ - __nfp_abm_reset_root(netdev, alink, handle, qs, ~0); -} - -static int -nfp_abm_red_find(struct nfp_abm_link *alink, struct tc_red_qopt_offload *opt) -{ - unsigned int i = TC_H_MIN(opt->parent) - 1; - - if (opt->parent == TC_H_ROOT) - i = 0; - else if (TC_H_MAJ(alink->parent) == TC_H_MAJ(opt->parent)) - i = TC_H_MIN(opt->parent) - 1; - else - return -EOPNOTSUPP; - - if (i >= alink->num_qdiscs || opt->handle != alink->qdiscs[i].handle) - return -EOPNOTSUPP; - - return i; -} - -static void -nfp_abm_red_destroy(struct net_device *netdev, struct nfp_abm_link *alink, - u32 handle) -{ - unsigned int i; - - for (i = 0; i < alink->num_qdiscs; i++) - if (handle == alink->qdiscs[i].handle) - break; - if (i == alink->num_qdiscs) - return; - - if (alink->parent == TC_H_ROOT) { - nfp_abm_reset_root(netdev, alink, TC_H_ROOT, 0); - } else { - nfp_abm_ctrl_set_q_lvl(alink, i, ~0); - memset(&alink->qdiscs[i], 0, sizeof(*alink->qdiscs)); - } -} - -static int -nfp_abm_red_replace(struct net_device *netdev, struct nfp_abm_link *alink, - struct tc_red_qopt_offload *opt) -{ - bool existing; - int i, err; - - i = nfp_abm_red_find(alink, opt); - existing = i >= 0; - - if (opt->set.min != opt->set.max || !opt->set.is_ecn) { - nfp_warn(alink->abm->app->cpp, - "RED offload failed - unsupported parameters\n"); - err = -EINVAL; - goto err_destroy; - } - - if (existing) { - if (alink->parent == TC_H_ROOT) - err = nfp_abm_ctrl_set_all_q_lvls(alink, opt->set.min); - else - err = nfp_abm_ctrl_set_q_lvl(alink, i, opt->set.min); - if (err) - goto err_destroy; - return 0; - } - - if (opt->parent == TC_H_ROOT) { - i = 0; - err = __nfp_abm_reset_root(netdev, alink, TC_H_ROOT, 1, - opt->set.min); - } else if (TC_H_MAJ(alink->parent) == TC_H_MAJ(opt->parent)) { - i = TC_H_MIN(opt->parent) - 1; - err = nfp_abm_ctrl_set_q_lvl(alink, i, opt->set.min); - } else { - return -EINVAL; - } - /* Set the handle to try full clean up, in case IO failed */ - alink->qdiscs[i].handle = opt->handle; - if (err) - goto err_destroy; - - if (opt->parent == TC_H_ROOT) - err = nfp_abm_ctrl_read_stats(alink, &alink->qdiscs[i].stats); - else - err = nfp_abm_ctrl_read_q_stats(alink, i, - &alink->qdiscs[i].stats); - if (err) - goto err_destroy; - - if (opt->parent == TC_H_ROOT) - err = nfp_abm_ctrl_read_xstats(alink, - &alink->qdiscs[i].xstats); - else - err = nfp_abm_ctrl_read_q_xstats(alink, i, - &alink->qdiscs[i].xstats); - if (err) - goto err_destroy; - - alink->qdiscs[i].stats.backlog_pkts = 0; - alink->qdiscs[i].stats.backlog_bytes = 0; - - return 0; -err_destroy: - /* If the qdisc keeps on living, but we can't offload undo changes */ - if (existing) { - opt->set.qstats->qlen -= alink->qdiscs[i].stats.backlog_pkts; - opt->set.qstats->backlog -= - alink->qdiscs[i].stats.backlog_bytes; - } - nfp_abm_red_destroy(netdev, alink, opt->handle); - - return err; -} - -static void -nfp_abm_update_stats(struct nfp_alink_stats *new, struct nfp_alink_stats *old, - struct tc_qopt_offload_stats *stats) -{ - _bstats_update(stats->bstats, new->tx_bytes - old->tx_bytes, - new->tx_pkts - old->tx_pkts); - stats->qstats->qlen += new->backlog_pkts - old->backlog_pkts; - stats->qstats->backlog += new->backlog_bytes - old->backlog_bytes; - stats->qstats->overlimits += new->overlimits - old->overlimits; - stats->qstats->drops += new->drops - old->drops; -} - -static int -nfp_abm_red_stats(struct nfp_abm_link *alink, struct tc_red_qopt_offload *opt) -{ - struct nfp_alink_stats *prev_stats; - struct nfp_alink_stats stats; - int i, err; - - i = nfp_abm_red_find(alink, opt); - if (i < 0) - return i; - prev_stats = &alink->qdiscs[i].stats; - - if (alink->parent == TC_H_ROOT) - err = nfp_abm_ctrl_read_stats(alink, &stats); - else - err = nfp_abm_ctrl_read_q_stats(alink, i, &stats); - if (err) - return err; - - nfp_abm_update_stats(&stats, prev_stats, &opt->stats); - - *prev_stats = stats; - - return 0; -} - -static int -nfp_abm_red_xstats(struct nfp_abm_link *alink, struct tc_red_qopt_offload *opt) -{ - struct nfp_alink_xstats *prev_xstats; - struct nfp_alink_xstats xstats; - int i, err; - - i = nfp_abm_red_find(alink, opt); - if (i < 0) - return i; - prev_xstats = &alink->qdiscs[i].xstats; - - if (alink->parent == TC_H_ROOT) - err = nfp_abm_ctrl_read_xstats(alink, &xstats); - else - err = nfp_abm_ctrl_read_q_xstats(alink, i, &xstats); - if (err) - return err; - - opt->xstats->forced_mark += xstats.ecn_marked - prev_xstats->ecn_marked; - opt->xstats->pdrop += xstats.pdrop - prev_xstats->pdrop; - - *prev_xstats = xstats; - - return 0; -} - -static int -nfp_abm_setup_tc_red(struct net_device *netdev, struct nfp_abm_link *alink, - struct tc_red_qopt_offload *opt) -{ - switch (opt->command) { - case TC_RED_REPLACE: - return nfp_abm_red_replace(netdev, alink, opt); - case TC_RED_DESTROY: - nfp_abm_red_destroy(netdev, alink, opt->handle); - return 0; - case TC_RED_STATS: - return nfp_abm_red_stats(alink, opt); - case TC_RED_XSTATS: - return nfp_abm_red_xstats(alink, opt); - default: - return -EOPNOTSUPP; - } -} - -static int -nfp_abm_mq_stats(struct nfp_abm_link *alink, struct tc_mq_qopt_offload *opt) -{ - struct nfp_alink_stats stats; - unsigned int i; - int err; - - for (i = 0; i < alink->num_qdiscs; i++) { - if (alink->qdiscs[i].handle == TC_H_UNSPEC) - continue; - - err = nfp_abm_ctrl_read_q_stats(alink, i, &stats); - if (err) - return err; - - nfp_abm_update_stats(&stats, &alink->qdiscs[i].stats, - &opt->stats); - } - - return 0; -} - -static int -nfp_abm_setup_tc_mq(struct net_device *netdev, struct nfp_abm_link *alink, - struct tc_mq_qopt_offload *opt) -{ - switch (opt->command) { - case TC_MQ_CREATE: - nfp_abm_reset_root(netdev, alink, opt->handle, - alink->total_queues); - return 0; - case TC_MQ_DESTROY: - if (opt->handle == alink->parent) - nfp_abm_reset_root(netdev, alink, TC_H_ROOT, 0); - return 0; - case TC_MQ_STATS: - return nfp_abm_mq_stats(alink, opt); - default: - return -EOPNOTSUPP; - } -} - -static int nfp_abm_setup_tc(struct nfp_app *app, struct net_device *netdev, enum tc_setup_type type, void *type_data) { @@ -302,10 +38,16 @@ nfp_abm_setup_tc(struct nfp_app *app, struct net_device *netdev, return -EOPNOTSUPP; switch (type) { + case TC_SETUP_ROOT_QDISC: + return nfp_abm_setup_root(netdev, repr->app_priv, type_data); case TC_SETUP_QDISC_MQ: return nfp_abm_setup_tc_mq(netdev, repr->app_priv, type_data); case TC_SETUP_QDISC_RED: return nfp_abm_setup_tc_red(netdev, repr->app_priv, type_data); + case TC_SETUP_QDISC_GRED: + return nfp_abm_setup_tc_gred(netdev, repr->app_priv, type_data); + case TC_SETUP_BLOCK: + return nfp_abm_setup_cls_block(netdev, repr, type_data); default: return -EOPNOTSUPP; } @@ -384,7 +126,9 @@ nfp_abm_spawn_repr(struct nfp_app *app, struct nfp_abm_link *alink, reprs = nfp_reprs_get_locked(app, rtype); WARN(nfp_repr_get_locked(app, reprs, alink->id), "duplicate repr"); + rtnl_lock(); rcu_assign_pointer(reprs->reprs[alink->id], netdev); + rtnl_unlock(); nfp_info(app->cpp, "%s Port %d Representor(%s) created\n", ptype == NFP_PORT_PF_PORT ? "PCIe" : "Phys", @@ -410,7 +154,9 @@ nfp_abm_kill_repr(struct nfp_app *app, struct nfp_abm_link *alink, netdev = nfp_repr_get_locked(app, reprs, alink->id); if (!netdev) return; + rtnl_lock(); rcu_assign_pointer(reprs->reprs[alink->id], NULL); + rtnl_unlock(); synchronize_rcu(); /* Cast to make sure nfp_repr_clean_and_free() takes a nfp_repr */ nfp_repr_clean_and_free((struct nfp_repr *)netdev_priv(netdev)); @@ -461,6 +207,9 @@ static int nfp_abm_eswitch_set_switchdev(struct nfp_abm *abm) struct nfp_net *nn; int err; + if (!abm->red_support) + return -EOPNOTSUPP; + err = nfp_abm_ctrl_qm_enable(abm); if (err) return err; @@ -573,31 +322,34 @@ nfp_abm_vnic_alloc(struct nfp_app *app, struct nfp_net *nn, unsigned int id) alink->abm = abm; alink->vnic = nn; alink->id = id; - alink->parent = TC_H_ROOT; alink->total_queues = alink->vnic->max_rx_rings; - alink->qdiscs = kvcalloc(alink->total_queues, sizeof(*alink->qdiscs), - GFP_KERNEL); - if (!alink->qdiscs) { - err = -ENOMEM; + + INIT_LIST_HEAD(&alink->dscp_map); + + err = nfp_abm_ctrl_read_params(alink); + if (err) + goto err_free_alink; + + alink->prio_map = kzalloc(abm->prio_map_len, GFP_KERNEL); + if (!alink->prio_map) goto err_free_alink; - } /* This is a multi-host app, make sure MAC/PHY is up, but don't * make the MAC/PHY state follow the state of any of the ports. */ err = nfp_eth_set_configured(app->cpp, eth_port->index, true); if (err < 0) - goto err_free_qdiscs; + goto err_free_priomap; netif_keep_dst(nn->dp.netdev); nfp_abm_vnic_set_mac(app->pf, abm, nn, id); - nfp_abm_ctrl_read_params(alink); + INIT_RADIX_TREE(&alink->qdiscs, GFP_KERNEL); return 0; -err_free_qdiscs: - kvfree(alink->qdiscs); +err_free_priomap: + kfree(alink->prio_map); err_free_alink: kfree(alink); return err; @@ -608,10 +360,20 @@ static void nfp_abm_vnic_free(struct nfp_app *app, struct nfp_net *nn) struct nfp_abm_link *alink = nn->app_priv; nfp_abm_kill_reprs(alink->abm, alink); - kvfree(alink->qdiscs); + WARN(!radix_tree_empty(&alink->qdiscs), "left over qdiscs\n"); + kfree(alink->prio_map); kfree(alink); } +static int nfp_abm_vnic_init(struct nfp_app *app, struct nfp_net *nn) +{ + struct nfp_abm_link *alink = nn->app_priv; + + if (nfp_abm_has_prio(alink->abm)) + return nfp_abm_ctrl_prio_map_update(alink, alink->prio_map); + return 0; +} + static u64 * nfp_abm_port_get_stats(struct nfp_app *app, struct nfp_port *port, u64 *data) { @@ -659,6 +421,21 @@ nfp_abm_port_get_stats_strings(struct nfp_app *app, struct nfp_port *port, return data; } +static int nfp_abm_fw_init_reset(struct nfp_abm *abm) +{ + unsigned int i; + + if (!abm->red_support) + return 0; + + for (i = 0; i < abm->num_bands * NFP_NET_MAX_RX_RINGS; i++) { + __nfp_abm_ctrl_set_q_lvl(abm, i, NFP_ABM_LVL_INFINITY); + __nfp_abm_ctrl_set_q_act(abm, i, NFP_ABM_ACT_DROP); + } + + return nfp_abm_ctrl_qm_disable(abm); +} + static int nfp_abm_init(struct nfp_app *app) { struct nfp_pf *pf = app->pf; @@ -690,15 +467,31 @@ static int nfp_abm_init(struct nfp_app *app) if (err) goto err_free_abm; + err = -ENOMEM; + abm->num_thresholds = array_size(abm->num_bands, NFP_NET_MAX_RX_RINGS); + abm->threshold_undef = bitmap_zalloc(abm->num_thresholds, GFP_KERNEL); + if (!abm->threshold_undef) + goto err_free_abm; + + abm->thresholds = kvcalloc(abm->num_thresholds, + sizeof(*abm->thresholds), GFP_KERNEL); + if (!abm->thresholds) + goto err_free_thresh_umap; + + abm->actions = kvcalloc(abm->num_thresholds, sizeof(*abm->actions), + GFP_KERNEL); + if (!abm->actions) + goto err_free_thresh; + /* We start in legacy mode, make sure advanced queuing is disabled */ - err = nfp_abm_ctrl_qm_disable(abm); + err = nfp_abm_fw_init_reset(abm); if (err) - goto err_free_abm; + goto err_free_act; err = -ENOMEM; reprs = nfp_reprs_alloc(pf->max_data_vnics); if (!reprs) - goto err_free_abm; + goto err_free_act; RCU_INIT_POINTER(app->reprs[NFP_REPR_TYPE_PHYS_PORT], reprs); reprs = nfp_reprs_alloc(pf->max_data_vnics); @@ -710,6 +503,12 @@ static int nfp_abm_init(struct nfp_app *app) err_free_phys: nfp_reprs_clean_and_free_by_type(app, NFP_REPR_TYPE_PHYS_PORT); +err_free_act: + kvfree(abm->actions); +err_free_thresh: + kvfree(abm->thresholds); +err_free_thresh_umap: + bitmap_free(abm->threshold_undef); err_free_abm: kfree(abm); app->priv = NULL; @@ -723,6 +522,9 @@ static void nfp_abm_clean(struct nfp_app *app) nfp_abm_eswitch_clean_up(abm); nfp_reprs_clean_and_free_by_type(app, NFP_REPR_TYPE_PF); nfp_reprs_clean_and_free_by_type(app, NFP_REPR_TYPE_PHYS_PORT); + bitmap_free(abm->threshold_undef); + kvfree(abm->actions); + kvfree(abm->thresholds); kfree(abm); app->priv = NULL; } @@ -736,6 +538,7 @@ const struct nfp_app_type app_abm = { .vnic_alloc = nfp_abm_vnic_alloc, .vnic_free = nfp_abm_vnic_free, + .vnic_init = nfp_abm_vnic_init, .port_get_stats = nfp_abm_port_get_stats, .port_get_stats_count = nfp_abm_port_get_stats_count, diff --git a/drivers/net/ethernet/netronome/nfp/abm/main.h b/drivers/net/ethernet/netronome/nfp/abm/main.h index f907b7d98917..49749c60885e 100644 --- a/drivers/net/ethernet/netronome/nfp/abm/main.h +++ b/drivers/net/ethernet/netronome/nfp/abm/main.h @@ -4,7 +4,19 @@ #ifndef __NFP_ABM_H__ #define __NFP_ABM_H__ 1 +#include <linux/bits.h> +#include <linux/list.h> +#include <linux/radix-tree.h> #include <net/devlink.h> +#include <net/pkt_cls.h> +#include <net/pkt_sched.h> + +/* Dump of 64 PRIOs and 256 REDs seems to take 850us on Xeon v4 @ 2.20GHz; + * 2.5ms / 400Hz seems more than sufficient for stats resolution. + */ +#define NFP_ABM_STATS_REFRESH_IVAL (2500 * 1000) /* ns */ + +#define NFP_ABM_LVL_INFINITY S32_MAX struct nfp_app; struct nfp_net; @@ -12,21 +24,64 @@ struct nfp_net; #define NFP_ABM_PORTID_TYPE GENMASK(23, 16) #define NFP_ABM_PORTID_ID GENMASK(7, 0) +/* The possible actions if thresholds are exceeded */ +enum nfp_abm_q_action { + /* mark if ECN capable, otherwise drop */ + NFP_ABM_ACT_MARK_DROP = 0, + /* mark if ECN capable, otherwise goto QM */ + NFP_ABM_ACT_MARK_QUEUE = 1, + NFP_ABM_ACT_DROP = 2, + NFP_ABM_ACT_QUEUE = 3, + NFP_ABM_ACT_NOQUEUE = 4, +}; + /** * struct nfp_abm - ABM NIC app structure * @app: back pointer to nfp_app * @pf_id: ID of our PF link + * + * @red_support: is RED offload supported + * @num_prios: number of supported DSCP priorities + * @num_bands: number of supported DSCP priority bands + * @action_mask: bitmask of supported actions + * + * @thresholds: current threshold configuration + * @threshold_undef: bitmap of thresholds which have not been set + * @actions: current FW action configuration + * @num_thresholds: number of @thresholds and bits in @threshold_undef + * + * @prio_map_len: computed length of FW priority map (in bytes) + * @dscp_mask: mask FW will apply on DSCP field + * * @eswitch_mode: devlink eswitch mode, advanced functions only visible * in switchdev mode + * * @q_lvls: queue level control area * @qm_stats: queue statistics symbol + * @q_stats: basic queue statistics (only in per-band case) */ struct nfp_abm { struct nfp_app *app; unsigned int pf_id; + + unsigned int red_support; + unsigned int num_prios; + unsigned int num_bands; + unsigned int action_mask; + + u32 *thresholds; + unsigned long *threshold_undef; + u8 *actions; + size_t num_thresholds; + + unsigned int prio_map_len; + u8 dscp_mask; + enum devlink_eswitch_mode eswitch_mode; + const struct nfp_rtsym *q_lvls; const struct nfp_rtsym *qm_stats; + const struct nfp_rtsym *q_stats; }; /** @@ -57,16 +112,76 @@ struct nfp_alink_xstats { u64 pdrop; }; +enum nfp_qdisc_type { + NFP_QDISC_NONE = 0, + NFP_QDISC_MQ, + NFP_QDISC_RED, + NFP_QDISC_GRED, +}; + +#define NFP_QDISC_UNTRACKED ((struct nfp_qdisc *)1UL) + /** - * struct nfp_red_qdisc - representation of single RED Qdisc - * @handle: handle of currently offloaded RED Qdisc - * @stats: statistics from last refresh - * @xstats: base of extended statistics + * struct nfp_qdisc - tracked TC Qdisc + * @netdev: netdev on which Qdisc was created + * @type: Qdisc type + * @handle: handle of this Qdisc + * @parent_handle: handle of the parent (unreliable if Qdisc was grafted) + * @use_cnt: number of attachment points in the hierarchy + * @num_children: current size of the @children array + * @children: pointers to children + * + * @params_ok: parameters of this Qdisc are OK for offload + * @offload_mark: offload refresh state - selected for offload + * @offloaded: Qdisc is currently offloaded to the HW + * + * @mq: MQ Qdisc specific parameters and state + * @mq.stats: current stats of the MQ Qdisc + * @mq.prev_stats: previously reported @mq.stats + * + * @red: RED Qdisc specific parameters and state + * @red.num_bands: Number of valid entries in the @red.band table + * @red.band: Per-band array of RED instances + * @red.band.ecn: ECN marking is enabled (rather than drop) + * @red.band.threshold: ECN marking threshold + * @red.band.stats: current stats of the RED Qdisc + * @red.band.prev_stats: previously reported @red.stats + * @red.band.xstats: extended stats for RED - current + * @red.band.prev_xstats: extended stats for RED - previously reported */ -struct nfp_red_qdisc { +struct nfp_qdisc { + struct net_device *netdev; + enum nfp_qdisc_type type; u32 handle; - struct nfp_alink_stats stats; - struct nfp_alink_xstats xstats; + u32 parent_handle; + unsigned int use_cnt; + unsigned int num_children; + struct nfp_qdisc **children; + + bool params_ok; + bool offload_mark; + bool offloaded; + + union { + /* NFP_QDISC_MQ */ + struct { + struct nfp_alink_stats stats; + struct nfp_alink_stats prev_stats; + } mq; + /* TC_SETUP_QDISC_RED, TC_SETUP_QDISC_GRED */ + struct { + unsigned int num_bands; + + struct { + bool ecn; + u32 threshold; + struct nfp_alink_stats stats; + struct nfp_alink_stats prev_stats; + struct nfp_alink_xstats xstats; + struct nfp_alink_xstats prev_xstats; + } band[MAX_DPs]; + } red; + }; }; /** @@ -76,9 +191,17 @@ struct nfp_red_qdisc { * @id: id of the data vNIC * @queue_base: id of base to host queue within PCIe (not QC idx) * @total_queues: number of PF queues - * @parent: handle of expected parent, i.e. handle of MQ, or TC_H_ROOT - * @num_qdiscs: number of currently used qdiscs - * @qdiscs: array of qdiscs + * + * @last_stats_update: ktime of last stats update + * + * @prio_map: current map of priorities + * @has_prio: @prio_map is valid + * + * @def_band: default band to use + * @dscp_map: list of DSCP to band mappings + * + * @root_qdisc: pointer to the current root of the Qdisc hierarchy + * @qdiscs: all qdiscs recorded by major part of the handle */ struct nfp_abm_link { struct nfp_abm *abm; @@ -86,26 +209,65 @@ struct nfp_abm_link { unsigned int id; unsigned int queue_base; unsigned int total_queues; - u32 parent; - unsigned int num_qdiscs; - struct nfp_red_qdisc *qdiscs; + + u64 last_stats_update; + + u32 *prio_map; + bool has_prio; + + u8 def_band; + struct list_head dscp_map; + + struct nfp_qdisc *root_qdisc; + struct radix_tree_root qdiscs; }; -void nfp_abm_ctrl_read_params(struct nfp_abm_link *alink); +static inline bool nfp_abm_has_prio(struct nfp_abm *abm) +{ + return abm->num_bands > 1; +} + +static inline bool nfp_abm_has_drop(struct nfp_abm *abm) +{ + return abm->action_mask & BIT(NFP_ABM_ACT_DROP); +} + +static inline bool nfp_abm_has_mark(struct nfp_abm *abm) +{ + return abm->action_mask & BIT(NFP_ABM_ACT_MARK_DROP); +} + +void nfp_abm_qdisc_offload_update(struct nfp_abm_link *alink); +int nfp_abm_setup_root(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_root_qopt_offload *opt); +int nfp_abm_setup_tc_red(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_red_qopt_offload *opt); +int nfp_abm_setup_tc_mq(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_mq_qopt_offload *opt); +int nfp_abm_setup_tc_gred(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_gred_qopt_offload *opt); +int nfp_abm_setup_cls_block(struct net_device *netdev, struct nfp_repr *repr, + struct tc_block_offload *opt); + +int nfp_abm_ctrl_read_params(struct nfp_abm_link *alink); int nfp_abm_ctrl_find_addrs(struct nfp_abm *abm); -int nfp_abm_ctrl_set_all_q_lvls(struct nfp_abm_link *alink, u32 val); -int nfp_abm_ctrl_set_q_lvl(struct nfp_abm_link *alink, unsigned int i, - u32 val); -int nfp_abm_ctrl_read_stats(struct nfp_abm_link *alink, - struct nfp_alink_stats *stats); -int nfp_abm_ctrl_read_q_stats(struct nfp_abm_link *alink, unsigned int i, +int __nfp_abm_ctrl_set_q_lvl(struct nfp_abm *abm, unsigned int id, u32 val); +int nfp_abm_ctrl_set_q_lvl(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, u32 val); +int __nfp_abm_ctrl_set_q_act(struct nfp_abm *abm, unsigned int id, + enum nfp_abm_q_action act); +int nfp_abm_ctrl_set_q_act(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, enum nfp_abm_q_action act); +int nfp_abm_ctrl_read_q_stats(struct nfp_abm_link *alink, + unsigned int band, unsigned int queue, struct nfp_alink_stats *stats); -int nfp_abm_ctrl_read_xstats(struct nfp_abm_link *alink, - struct nfp_alink_xstats *xstats); -int nfp_abm_ctrl_read_q_xstats(struct nfp_abm_link *alink, unsigned int i, +int nfp_abm_ctrl_read_q_xstats(struct nfp_abm_link *alink, + unsigned int band, unsigned int queue, struct nfp_alink_xstats *xstats); u64 nfp_abm_ctrl_stat_non_sto(struct nfp_abm_link *alink, unsigned int i); u64 nfp_abm_ctrl_stat_sto(struct nfp_abm_link *alink, unsigned int i); int nfp_abm_ctrl_qm_enable(struct nfp_abm *abm); int nfp_abm_ctrl_qm_disable(struct nfp_abm *abm); +void nfp_abm_prio_map_update(struct nfp_abm *abm); +int nfp_abm_ctrl_prio_map_update(struct nfp_abm_link *alink, u32 *packed); #endif diff --git a/drivers/net/ethernet/netronome/nfp/abm/qdisc.c b/drivers/net/ethernet/netronome/nfp/abm/qdisc.c new file mode 100644 index 000000000000..2473fb5f75e5 --- /dev/null +++ b/drivers/net/ethernet/netronome/nfp/abm/qdisc.c @@ -0,0 +1,850 @@ +// SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +/* Copyright (C) 2018 Netronome Systems, Inc. */ + +#include <linux/rtnetlink.h> +#include <net/pkt_cls.h> +#include <net/pkt_sched.h> +#include <net/red.h> + +#include "../nfpcore/nfp_cpp.h" +#include "../nfp_app.h" +#include "../nfp_main.h" +#include "../nfp_net.h" +#include "../nfp_port.h" +#include "main.h" + +static bool nfp_abm_qdisc_is_red(struct nfp_qdisc *qdisc) +{ + return qdisc->type == NFP_QDISC_RED || qdisc->type == NFP_QDISC_GRED; +} + +static bool nfp_abm_qdisc_child_valid(struct nfp_qdisc *qdisc, unsigned int id) +{ + return qdisc->children[id] && + qdisc->children[id] != NFP_QDISC_UNTRACKED; +} + +static void *nfp_abm_qdisc_tree_deref_slot(void __rcu **slot) +{ + return rtnl_dereference(*slot); +} + +static void +nfp_abm_stats_propagate(struct nfp_alink_stats *parent, + struct nfp_alink_stats *child) +{ + parent->tx_pkts += child->tx_pkts; + parent->tx_bytes += child->tx_bytes; + parent->backlog_pkts += child->backlog_pkts; + parent->backlog_bytes += child->backlog_bytes; + parent->overlimits += child->overlimits; + parent->drops += child->drops; +} + +static void +nfp_abm_stats_update_red(struct nfp_abm_link *alink, struct nfp_qdisc *qdisc, + unsigned int queue) +{ + struct nfp_cpp *cpp = alink->abm->app->cpp; + unsigned int i; + int err; + + if (!qdisc->offloaded) + return; + + for (i = 0; i < qdisc->red.num_bands; i++) { + err = nfp_abm_ctrl_read_q_stats(alink, i, queue, + &qdisc->red.band[i].stats); + if (err) + nfp_err(cpp, "RED stats (%d, %d) read failed with error %d\n", + i, queue, err); + + err = nfp_abm_ctrl_read_q_xstats(alink, i, queue, + &qdisc->red.band[i].xstats); + if (err) + nfp_err(cpp, "RED xstats (%d, %d) read failed with error %d\n", + i, queue, err); + } +} + +static void +nfp_abm_stats_update_mq(struct nfp_abm_link *alink, struct nfp_qdisc *qdisc) +{ + unsigned int i; + + if (qdisc->type != NFP_QDISC_MQ) + return; + + for (i = 0; i < alink->total_queues; i++) + if (nfp_abm_qdisc_child_valid(qdisc, i)) + nfp_abm_stats_update_red(alink, qdisc->children[i], i); +} + +static void __nfp_abm_stats_update(struct nfp_abm_link *alink, u64 time_now) +{ + alink->last_stats_update = time_now; + if (alink->root_qdisc) + nfp_abm_stats_update_mq(alink, alink->root_qdisc); +} + +static void nfp_abm_stats_update(struct nfp_abm_link *alink) +{ + u64 now; + + /* Limit the frequency of updates - stats of non-leaf qdiscs are a sum + * of all their leafs, so we would read the same stat multiple times + * for every dump. + */ + now = ktime_get(); + if (now - alink->last_stats_update < NFP_ABM_STATS_REFRESH_IVAL) + return; + + __nfp_abm_stats_update(alink, now); +} + +static void +nfp_abm_qdisc_unlink_children(struct nfp_qdisc *qdisc, + unsigned int start, unsigned int end) +{ + unsigned int i; + + for (i = start; i < end; i++) + if (nfp_abm_qdisc_child_valid(qdisc, i)) { + qdisc->children[i]->use_cnt--; + qdisc->children[i] = NULL; + } +} + +static void +nfp_abm_qdisc_offload_stop(struct nfp_abm_link *alink, struct nfp_qdisc *qdisc) +{ + unsigned int i; + + /* Don't complain when qdisc is getting unlinked */ + if (qdisc->use_cnt) + nfp_warn(alink->abm->app->cpp, "Offload of '%08x' stopped\n", + qdisc->handle); + + if (!nfp_abm_qdisc_is_red(qdisc)) + return; + + for (i = 0; i < qdisc->red.num_bands; i++) { + qdisc->red.band[i].stats.backlog_pkts = 0; + qdisc->red.band[i].stats.backlog_bytes = 0; + } +} + +static int +__nfp_abm_stats_init(struct nfp_abm_link *alink, unsigned int band, + unsigned int queue, struct nfp_alink_stats *prev_stats, + struct nfp_alink_xstats *prev_xstats) +{ + u64 backlog_pkts, backlog_bytes; + int err; + + /* Don't touch the backlog, backlog can only be reset after it has + * been reported back to the tc qdisc stats. + */ + backlog_pkts = prev_stats->backlog_pkts; + backlog_bytes = prev_stats->backlog_bytes; + + err = nfp_abm_ctrl_read_q_stats(alink, band, queue, prev_stats); + if (err) { + nfp_err(alink->abm->app->cpp, + "RED stats init (%d, %d) failed with error %d\n", + band, queue, err); + return err; + } + + err = nfp_abm_ctrl_read_q_xstats(alink, band, queue, prev_xstats); + if (err) { + nfp_err(alink->abm->app->cpp, + "RED xstats init (%d, %d) failed with error %d\n", + band, queue, err); + return err; + } + + prev_stats->backlog_pkts = backlog_pkts; + prev_stats->backlog_bytes = backlog_bytes; + return 0; +} + +static int +nfp_abm_stats_init(struct nfp_abm_link *alink, struct nfp_qdisc *qdisc, + unsigned int queue) +{ + unsigned int i; + int err; + + for (i = 0; i < qdisc->red.num_bands; i++) { + err = __nfp_abm_stats_init(alink, i, queue, + &qdisc->red.band[i].prev_stats, + &qdisc->red.band[i].prev_xstats); + if (err) + return err; + } + + return 0; +} + +static void +nfp_abm_offload_compile_red(struct nfp_abm_link *alink, struct nfp_qdisc *qdisc, + unsigned int queue) +{ + bool good_red, good_gred; + unsigned int i; + + good_red = qdisc->type == NFP_QDISC_RED && + qdisc->params_ok && + qdisc->use_cnt == 1 && + !alink->has_prio && + !qdisc->children[0]; + good_gred = qdisc->type == NFP_QDISC_GRED && + qdisc->params_ok && + qdisc->use_cnt == 1; + qdisc->offload_mark = good_red || good_gred; + + /* If we are starting offload init prev_stats */ + if (qdisc->offload_mark && !qdisc->offloaded) + if (nfp_abm_stats_init(alink, qdisc, queue)) + qdisc->offload_mark = false; + + if (!qdisc->offload_mark) + return; + + for (i = 0; i < alink->abm->num_bands; i++) { + enum nfp_abm_q_action act; + + nfp_abm_ctrl_set_q_lvl(alink, i, queue, + qdisc->red.band[i].threshold); + act = qdisc->red.band[i].ecn ? + NFP_ABM_ACT_MARK_DROP : NFP_ABM_ACT_DROP; + nfp_abm_ctrl_set_q_act(alink, i, queue, act); + } +} + +static void +nfp_abm_offload_compile_mq(struct nfp_abm_link *alink, struct nfp_qdisc *qdisc) +{ + unsigned int i; + + qdisc->offload_mark = qdisc->type == NFP_QDISC_MQ; + if (!qdisc->offload_mark) + return; + + for (i = 0; i < alink->total_queues; i++) { + struct nfp_qdisc *child = qdisc->children[i]; + + if (!nfp_abm_qdisc_child_valid(qdisc, i)) + continue; + + nfp_abm_offload_compile_red(alink, child, i); + } +} + +void nfp_abm_qdisc_offload_update(struct nfp_abm_link *alink) +{ + struct nfp_abm *abm = alink->abm; + struct radix_tree_iter iter; + struct nfp_qdisc *qdisc; + void __rcu **slot; + size_t i; + + /* Mark all thresholds as unconfigured */ + for (i = 0; i < abm->num_bands; i++) + __bitmap_set(abm->threshold_undef, + i * NFP_NET_MAX_RX_RINGS + alink->queue_base, + alink->total_queues); + + /* Clear offload marks */ + radix_tree_for_each_slot(slot, &alink->qdiscs, &iter, 0) { + qdisc = nfp_abm_qdisc_tree_deref_slot(slot); + qdisc->offload_mark = false; + } + + if (alink->root_qdisc) + nfp_abm_offload_compile_mq(alink, alink->root_qdisc); + + /* Refresh offload status */ + radix_tree_for_each_slot(slot, &alink->qdiscs, &iter, 0) { + qdisc = nfp_abm_qdisc_tree_deref_slot(slot); + if (!qdisc->offload_mark && qdisc->offloaded) + nfp_abm_qdisc_offload_stop(alink, qdisc); + qdisc->offloaded = qdisc->offload_mark; + } + + /* Reset the unconfigured thresholds */ + for (i = 0; i < abm->num_thresholds; i++) + if (test_bit(i, abm->threshold_undef)) + __nfp_abm_ctrl_set_q_lvl(abm, i, NFP_ABM_LVL_INFINITY); + + __nfp_abm_stats_update(alink, ktime_get()); +} + +static void +nfp_abm_qdisc_clear_mq(struct net_device *netdev, struct nfp_abm_link *alink, + struct nfp_qdisc *qdisc) +{ + struct radix_tree_iter iter; + unsigned int mq_refs = 0; + void __rcu **slot; + + if (!qdisc->use_cnt) + return; + /* MQ doesn't notify well on destruction, we need special handling of + * MQ's children. + */ + if (qdisc->type == NFP_QDISC_MQ && + qdisc == alink->root_qdisc && + netdev->reg_state == NETREG_UNREGISTERING) + return; + + /* Count refs held by MQ instances and clear pointers */ + radix_tree_for_each_slot(slot, &alink->qdiscs, &iter, 0) { + struct nfp_qdisc *mq = nfp_abm_qdisc_tree_deref_slot(slot); + unsigned int i; + + if (mq->type != NFP_QDISC_MQ || mq->netdev != netdev) + continue; + for (i = 0; i < mq->num_children; i++) + if (mq->children[i] == qdisc) { + mq->children[i] = NULL; + mq_refs++; + } + } + + WARN(qdisc->use_cnt != mq_refs, "non-zero qdisc use count: %d (- %d)\n", + qdisc->use_cnt, mq_refs); +} + +static void +nfp_abm_qdisc_free(struct net_device *netdev, struct nfp_abm_link *alink, + struct nfp_qdisc *qdisc) +{ + struct nfp_port *port = nfp_port_from_netdev(netdev); + + if (!qdisc) + return; + nfp_abm_qdisc_clear_mq(netdev, alink, qdisc); + WARN_ON(radix_tree_delete(&alink->qdiscs, + TC_H_MAJ(qdisc->handle)) != qdisc); + + kfree(qdisc->children); + kfree(qdisc); + + port->tc_offload_cnt--; +} + +static struct nfp_qdisc * +nfp_abm_qdisc_alloc(struct net_device *netdev, struct nfp_abm_link *alink, + enum nfp_qdisc_type type, u32 parent_handle, u32 handle, + unsigned int children) +{ + struct nfp_port *port = nfp_port_from_netdev(netdev); + struct nfp_qdisc *qdisc; + int err; + + qdisc = kzalloc(sizeof(*qdisc), GFP_KERNEL); + if (!qdisc) + return NULL; + + if (children) { + qdisc->children = kcalloc(children, sizeof(void *), GFP_KERNEL); + if (!qdisc->children) + goto err_free_qdisc; + } + + qdisc->netdev = netdev; + qdisc->type = type; + qdisc->parent_handle = parent_handle; + qdisc->handle = handle; + qdisc->num_children = children; + + err = radix_tree_insert(&alink->qdiscs, TC_H_MAJ(qdisc->handle), qdisc); + if (err) { + nfp_err(alink->abm->app->cpp, + "Qdisc insertion into radix tree failed: %d\n", err); + goto err_free_child_tbl; + } + + port->tc_offload_cnt++; + return qdisc; + +err_free_child_tbl: + kfree(qdisc->children); +err_free_qdisc: + kfree(qdisc); + return NULL; +} + +static struct nfp_qdisc * +nfp_abm_qdisc_find(struct nfp_abm_link *alink, u32 handle) +{ + return radix_tree_lookup(&alink->qdiscs, TC_H_MAJ(handle)); +} + +static int +nfp_abm_qdisc_replace(struct net_device *netdev, struct nfp_abm_link *alink, + enum nfp_qdisc_type type, u32 parent_handle, u32 handle, + unsigned int children, struct nfp_qdisc **qdisc) +{ + *qdisc = nfp_abm_qdisc_find(alink, handle); + if (*qdisc) { + if (WARN_ON((*qdisc)->type != type)) + return -EINVAL; + return 1; + } + + *qdisc = nfp_abm_qdisc_alloc(netdev, alink, type, parent_handle, handle, + children); + return *qdisc ? 0 : -ENOMEM; +} + +static void +nfp_abm_qdisc_destroy(struct net_device *netdev, struct nfp_abm_link *alink, + u32 handle) +{ + struct nfp_qdisc *qdisc; + + qdisc = nfp_abm_qdisc_find(alink, handle); + if (!qdisc) + return; + + /* We don't get TC_SETUP_ROOT_QDISC w/ MQ when netdev is unregistered */ + if (alink->root_qdisc == qdisc) + qdisc->use_cnt--; + + nfp_abm_qdisc_unlink_children(qdisc, 0, qdisc->num_children); + nfp_abm_qdisc_free(netdev, alink, qdisc); + + if (alink->root_qdisc == qdisc) { + alink->root_qdisc = NULL; + /* Only root change matters, other changes are acted upon on + * the graft notification. + */ + nfp_abm_qdisc_offload_update(alink); + } +} + +static int +nfp_abm_qdisc_graft(struct nfp_abm_link *alink, u32 handle, u32 child_handle, + unsigned int id) +{ + struct nfp_qdisc *parent, *child; + + parent = nfp_abm_qdisc_find(alink, handle); + if (!parent) + return 0; + + if (WARN(id >= parent->num_children, + "graft child out of bound %d >= %d\n", + id, parent->num_children)) + return -EINVAL; + + nfp_abm_qdisc_unlink_children(parent, id, id + 1); + + child = nfp_abm_qdisc_find(alink, child_handle); + if (child) + child->use_cnt++; + else + child = NFP_QDISC_UNTRACKED; + parent->children[id] = child; + + nfp_abm_qdisc_offload_update(alink); + + return 0; +} + +static void +nfp_abm_stats_calculate(struct nfp_alink_stats *new, + struct nfp_alink_stats *old, + struct gnet_stats_basic_packed *bstats, + struct gnet_stats_queue *qstats) +{ + _bstats_update(bstats, new->tx_bytes - old->tx_bytes, + new->tx_pkts - old->tx_pkts); + qstats->qlen += new->backlog_pkts - old->backlog_pkts; + qstats->backlog += new->backlog_bytes - old->backlog_bytes; + qstats->overlimits += new->overlimits - old->overlimits; + qstats->drops += new->drops - old->drops; +} + +static void +nfp_abm_stats_red_calculate(struct nfp_alink_xstats *new, + struct nfp_alink_xstats *old, + struct red_stats *stats) +{ + stats->forced_mark += new->ecn_marked - old->ecn_marked; + stats->pdrop += new->pdrop - old->pdrop; +} + +static int +nfp_abm_gred_stats(struct nfp_abm_link *alink, u32 handle, + struct tc_gred_qopt_offload_stats *stats) +{ + struct nfp_qdisc *qdisc; + unsigned int i; + + nfp_abm_stats_update(alink); + + qdisc = nfp_abm_qdisc_find(alink, handle); + if (!qdisc) + return -EOPNOTSUPP; + /* If the qdisc offload has stopped we may need to adjust the backlog + * counters back so carry on even if qdisc is not currently offloaded. + */ + + for (i = 0; i < qdisc->red.num_bands; i++) { + if (!stats->xstats[i]) + continue; + + nfp_abm_stats_calculate(&qdisc->red.band[i].stats, + &qdisc->red.band[i].prev_stats, + &stats->bstats[i], &stats->qstats[i]); + qdisc->red.band[i].prev_stats = qdisc->red.band[i].stats; + + nfp_abm_stats_red_calculate(&qdisc->red.band[i].xstats, + &qdisc->red.band[i].prev_xstats, + stats->xstats[i]); + qdisc->red.band[i].prev_xstats = qdisc->red.band[i].xstats; + } + + return qdisc->offloaded ? 0 : -EOPNOTSUPP; +} + +static bool +nfp_abm_gred_check_params(struct nfp_abm_link *alink, + struct tc_gred_qopt_offload *opt) +{ + struct nfp_cpp *cpp = alink->abm->app->cpp; + struct nfp_abm *abm = alink->abm; + unsigned int i; + + if (opt->set.grio_on || opt->set.wred_on) { + nfp_warn(cpp, "GRED offload failed - GRIO and WRED not supported (p:%08x h:%08x)\n", + opt->parent, opt->handle); + return false; + } + if (opt->set.dp_def != alink->def_band) { + nfp_warn(cpp, "GRED offload failed - default band must be %d (p:%08x h:%08x)\n", + alink->def_band, opt->parent, opt->handle); + return false; + } + if (opt->set.dp_cnt != abm->num_bands) { + nfp_warn(cpp, "GRED offload failed - band count must be %d (p:%08x h:%08x)\n", + abm->num_bands, opt->parent, opt->handle); + return false; + } + + for (i = 0; i < abm->num_bands; i++) { + struct tc_gred_vq_qopt_offload_params *band = &opt->set.tab[i]; + + if (!band->present) + return false; + if (!band->is_ecn && !nfp_abm_has_drop(abm)) { + nfp_warn(cpp, "GRED offload failed - drop is not supported (ECN option required) (p:%08x h:%08x vq:%d)\n", + opt->parent, opt->handle, i); + return false; + } + if (band->is_ecn && !nfp_abm_has_mark(abm)) { + nfp_warn(cpp, "GRED offload failed - ECN marking not supported (p:%08x h:%08x vq:%d)\n", + opt->parent, opt->handle, i); + return false; + } + if (band->is_harddrop) { + nfp_warn(cpp, "GRED offload failed - harddrop is not supported (p:%08x h:%08x vq:%d)\n", + opt->parent, opt->handle, i); + return false; + } + if (band->min != band->max) { + nfp_warn(cpp, "GRED offload failed - threshold mismatch (p:%08x h:%08x vq:%d)\n", + opt->parent, opt->handle, i); + return false; + } + if (band->min > S32_MAX) { + nfp_warn(cpp, "GRED offload failed - threshold too large %d > %d (p:%08x h:%08x vq:%d)\n", + band->min, S32_MAX, opt->parent, opt->handle, + i); + return false; + } + } + + return true; +} + +static int +nfp_abm_gred_replace(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_gred_qopt_offload *opt) +{ + struct nfp_qdisc *qdisc; + unsigned int i; + int ret; + + ret = nfp_abm_qdisc_replace(netdev, alink, NFP_QDISC_GRED, opt->parent, + opt->handle, 0, &qdisc); + if (ret < 0) + return ret; + + qdisc->params_ok = nfp_abm_gred_check_params(alink, opt); + if (qdisc->params_ok) { + qdisc->red.num_bands = opt->set.dp_cnt; + for (i = 0; i < qdisc->red.num_bands; i++) { + qdisc->red.band[i].ecn = opt->set.tab[i].is_ecn; + qdisc->red.band[i].threshold = opt->set.tab[i].min; + } + } + + if (qdisc->use_cnt) + nfp_abm_qdisc_offload_update(alink); + + return 0; +} + +int nfp_abm_setup_tc_gred(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_gred_qopt_offload *opt) +{ + switch (opt->command) { + case TC_GRED_REPLACE: + return nfp_abm_gred_replace(netdev, alink, opt); + case TC_GRED_DESTROY: + nfp_abm_qdisc_destroy(netdev, alink, opt->handle); + return 0; + case TC_GRED_STATS: + return nfp_abm_gred_stats(alink, opt->handle, &opt->stats); + default: + return -EOPNOTSUPP; + } +} + +static int +nfp_abm_red_xstats(struct nfp_abm_link *alink, struct tc_red_qopt_offload *opt) +{ + struct nfp_qdisc *qdisc; + + nfp_abm_stats_update(alink); + + qdisc = nfp_abm_qdisc_find(alink, opt->handle); + if (!qdisc || !qdisc->offloaded) + return -EOPNOTSUPP; + + nfp_abm_stats_red_calculate(&qdisc->red.band[0].xstats, + &qdisc->red.band[0].prev_xstats, + opt->xstats); + qdisc->red.band[0].prev_xstats = qdisc->red.band[0].xstats; + return 0; +} + +static int +nfp_abm_red_stats(struct nfp_abm_link *alink, u32 handle, + struct tc_qopt_offload_stats *stats) +{ + struct nfp_qdisc *qdisc; + + nfp_abm_stats_update(alink); + + qdisc = nfp_abm_qdisc_find(alink, handle); + if (!qdisc) + return -EOPNOTSUPP; + /* If the qdisc offload has stopped we may need to adjust the backlog + * counters back so carry on even if qdisc is not currently offloaded. + */ + + nfp_abm_stats_calculate(&qdisc->red.band[0].stats, + &qdisc->red.band[0].prev_stats, + stats->bstats, stats->qstats); + qdisc->red.band[0].prev_stats = qdisc->red.band[0].stats; + + return qdisc->offloaded ? 0 : -EOPNOTSUPP; +} + +static bool +nfp_abm_red_check_params(struct nfp_abm_link *alink, + struct tc_red_qopt_offload *opt) +{ + struct nfp_cpp *cpp = alink->abm->app->cpp; + struct nfp_abm *abm = alink->abm; + + if (!opt->set.is_ecn && !nfp_abm_has_drop(abm)) { + nfp_warn(cpp, "RED offload failed - drop is not supported (ECN option required) (p:%08x h:%08x)\n", + opt->parent, opt->handle); + return false; + } + if (opt->set.is_ecn && !nfp_abm_has_mark(abm)) { + nfp_warn(cpp, "RED offload failed - ECN marking not supported (p:%08x h:%08x)\n", + opt->parent, opt->handle); + return false; + } + if (opt->set.is_harddrop) { + nfp_warn(cpp, "RED offload failed - harddrop is not supported (p:%08x h:%08x)\n", + opt->parent, opt->handle); + return false; + } + if (opt->set.min != opt->set.max) { + nfp_warn(cpp, "RED offload failed - unsupported min/max parameters (p:%08x h:%08x)\n", + opt->parent, opt->handle); + return false; + } + if (opt->set.min > NFP_ABM_LVL_INFINITY) { + nfp_warn(cpp, "RED offload failed - threshold too large %d > %d (p:%08x h:%08x)\n", + opt->set.min, NFP_ABM_LVL_INFINITY, opt->parent, + opt->handle); + return false; + } + + return true; +} + +static int +nfp_abm_red_replace(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_red_qopt_offload *opt) +{ + struct nfp_qdisc *qdisc; + int ret; + + ret = nfp_abm_qdisc_replace(netdev, alink, NFP_QDISC_RED, opt->parent, + opt->handle, 1, &qdisc); + if (ret < 0) + return ret; + + /* If limit != 0 child gets reset */ + if (opt->set.limit) { + if (nfp_abm_qdisc_child_valid(qdisc, 0)) + qdisc->children[0]->use_cnt--; + qdisc->children[0] = NULL; + } else { + /* Qdisc was just allocated without a limit will use noop_qdisc, + * i.e. a block hole. + */ + if (!ret) + qdisc->children[0] = NFP_QDISC_UNTRACKED; + } + + qdisc->params_ok = nfp_abm_red_check_params(alink, opt); + if (qdisc->params_ok) { + qdisc->red.num_bands = 1; + qdisc->red.band[0].ecn = opt->set.is_ecn; + qdisc->red.band[0].threshold = opt->set.min; + } + + if (qdisc->use_cnt == 1) + nfp_abm_qdisc_offload_update(alink); + + return 0; +} + +int nfp_abm_setup_tc_red(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_red_qopt_offload *opt) +{ + switch (opt->command) { + case TC_RED_REPLACE: + return nfp_abm_red_replace(netdev, alink, opt); + case TC_RED_DESTROY: + nfp_abm_qdisc_destroy(netdev, alink, opt->handle); + return 0; + case TC_RED_STATS: + return nfp_abm_red_stats(alink, opt->handle, &opt->stats); + case TC_RED_XSTATS: + return nfp_abm_red_xstats(alink, opt); + case TC_RED_GRAFT: + return nfp_abm_qdisc_graft(alink, opt->handle, + opt->child_handle, 0); + default: + return -EOPNOTSUPP; + } +} + +static int +nfp_abm_mq_create(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_mq_qopt_offload *opt) +{ + struct nfp_qdisc *qdisc; + int ret; + + ret = nfp_abm_qdisc_replace(netdev, alink, NFP_QDISC_MQ, + TC_H_ROOT, opt->handle, alink->total_queues, + &qdisc); + if (ret < 0) + return ret; + + qdisc->params_ok = true; + qdisc->offloaded = true; + nfp_abm_qdisc_offload_update(alink); + return 0; +} + +static int +nfp_abm_mq_stats(struct nfp_abm_link *alink, u32 handle, + struct tc_qopt_offload_stats *stats) +{ + struct nfp_qdisc *qdisc, *red; + unsigned int i, j; + + qdisc = nfp_abm_qdisc_find(alink, handle); + if (!qdisc) + return -EOPNOTSUPP; + + nfp_abm_stats_update(alink); + + /* MQ stats are summed over the children in the core, so we need + * to add up the unreported child values. + */ + memset(&qdisc->mq.stats, 0, sizeof(qdisc->mq.stats)); + memset(&qdisc->mq.prev_stats, 0, sizeof(qdisc->mq.prev_stats)); + + for (i = 0; i < qdisc->num_children; i++) { + if (!nfp_abm_qdisc_child_valid(qdisc, i)) + continue; + + if (!nfp_abm_qdisc_is_red(qdisc->children[i])) + continue; + red = qdisc->children[i]; + + for (j = 0; j < red->red.num_bands; j++) { + nfp_abm_stats_propagate(&qdisc->mq.stats, + &red->red.band[j].stats); + nfp_abm_stats_propagate(&qdisc->mq.prev_stats, + &red->red.band[j].prev_stats); + } + } + + nfp_abm_stats_calculate(&qdisc->mq.stats, &qdisc->mq.prev_stats, + stats->bstats, stats->qstats); + + return qdisc->offloaded ? 0 : -EOPNOTSUPP; +} + +int nfp_abm_setup_tc_mq(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_mq_qopt_offload *opt) +{ + switch (opt->command) { + case TC_MQ_CREATE: + return nfp_abm_mq_create(netdev, alink, opt); + case TC_MQ_DESTROY: + nfp_abm_qdisc_destroy(netdev, alink, opt->handle); + return 0; + case TC_MQ_STATS: + return nfp_abm_mq_stats(alink, opt->handle, &opt->stats); + case TC_MQ_GRAFT: + return nfp_abm_qdisc_graft(alink, opt->handle, + opt->graft_params.child_handle, + opt->graft_params.queue); + default: + return -EOPNOTSUPP; + } +} + +int nfp_abm_setup_root(struct net_device *netdev, struct nfp_abm_link *alink, + struct tc_root_qopt_offload *opt) +{ + if (opt->ingress) + return -EOPNOTSUPP; + if (alink->root_qdisc) + alink->root_qdisc->use_cnt--; + alink->root_qdisc = nfp_abm_qdisc_find(alink, opt->handle); + if (alink->root_qdisc) + alink->root_qdisc->use_cnt++; + + nfp_abm_qdisc_offload_update(alink); + + return 0; +} diff --git a/drivers/net/ethernet/netronome/nfp/bpf/jit.c b/drivers/net/ethernet/netronome/nfp/bpf/jit.c index 97d33bb4d84d..f272247d1708 100644 --- a/drivers/net/ethernet/netronome/nfp/bpf/jit.c +++ b/drivers/net/ethernet/netronome/nfp/bpf/jit.c @@ -1266,7 +1266,7 @@ wrp_alu64_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, u64 imm = insn->imm; /* sign extend */ if (skip) { - meta->skip = true; + meta->flags |= FLAG_INSN_SKIP_NOOP; return 0; } @@ -1291,15 +1291,10 @@ wrp_alu64_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, static int wrp_alu32_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, - enum alu_op alu_op, bool skip) + enum alu_op alu_op) { const struct bpf_insn *insn = &meta->insn; - if (skip) { - meta->skip = true; - return 0; - } - wrp_alu_imm(nfp_prog, insn->dst_reg * 2, alu_op, insn->imm); wrp_immed(nfp_prog, reg_both(insn->dst_reg * 2 + 1), 0); @@ -1334,8 +1329,9 @@ wrp_test_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, wrp_test_reg_one(nfp_prog, insn->dst_reg * 2, alu_op, insn->src_reg * 2, br_mask, insn->off); - wrp_test_reg_one(nfp_prog, insn->dst_reg * 2 + 1, alu_op, - insn->src_reg * 2 + 1, br_mask, insn->off); + if (is_mbpf_jmp64(meta)) + wrp_test_reg_one(nfp_prog, insn->dst_reg * 2 + 1, alu_op, + insn->src_reg * 2 + 1, br_mask, insn->off); return 0; } @@ -1390,13 +1386,15 @@ static int cmp_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) else emit_alu(nfp_prog, reg_none(), tmp_reg, alu_op, reg_a(reg)); - tmp_reg = ur_load_imm_any(nfp_prog, imm >> 32, imm_b(nfp_prog)); - if (!code->swap) - emit_alu(nfp_prog, reg_none(), - reg_a(reg + 1), carry_op, tmp_reg); - else - emit_alu(nfp_prog, reg_none(), - tmp_reg, carry_op, reg_a(reg + 1)); + if (is_mbpf_jmp64(meta)) { + tmp_reg = ur_load_imm_any(nfp_prog, imm >> 32, imm_b(nfp_prog)); + if (!code->swap) + emit_alu(nfp_prog, reg_none(), + reg_a(reg + 1), carry_op, tmp_reg); + else + emit_alu(nfp_prog, reg_none(), + tmp_reg, carry_op, reg_a(reg + 1)); + } emit_br(nfp_prog, code->br_mask, insn->off, 0); @@ -1423,8 +1421,9 @@ static int cmp_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) } emit_alu(nfp_prog, reg_none(), reg_a(areg), ALU_OP_SUB, reg_b(breg)); - emit_alu(nfp_prog, reg_none(), - reg_a(areg + 1), ALU_OP_SUB_C, reg_b(breg + 1)); + if (is_mbpf_jmp64(meta)) + emit_alu(nfp_prog, reg_none(), + reg_a(areg + 1), ALU_OP_SUB_C, reg_b(breg + 1)); emit_br(nfp_prog, code->br_mask, insn->off, 0); return 0; @@ -1963,6 +1962,9 @@ static int neg_reg64(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) */ static int __shl_imm64(struct nfp_prog *nfp_prog, u8 dst, u8 shift_amt) { + if (!shift_amt) + return 0; + if (shift_amt < 32) { emit_shf(nfp_prog, reg_both(dst + 1), reg_a(dst + 1), SHF_OP_NONE, reg_b(dst), SHF_SC_R_DSHF, @@ -2075,6 +2077,9 @@ static int shl_reg64(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) */ static int __shr_imm64(struct nfp_prog *nfp_prog, u8 dst, u8 shift_amt) { + if (!shift_amt) + return 0; + if (shift_amt < 32) { emit_shf(nfp_prog, reg_both(dst), reg_a(dst + 1), SHF_OP_NONE, reg_b(dst), SHF_SC_R_DSHF, shift_amt); @@ -2176,6 +2181,9 @@ static int shr_reg64(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) */ static int __ashr_imm64(struct nfp_prog *nfp_prog, u8 dst, u8 shift_amt) { + if (!shift_amt) + return 0; + if (shift_amt < 32) { emit_shf(nfp_prog, reg_both(dst), reg_a(dst + 1), SHF_OP_NONE, reg_b(dst), SHF_SC_R_DSHF, shift_amt); @@ -2309,7 +2317,7 @@ static int xor_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) static int xor_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { - return wrp_alu32_imm(nfp_prog, meta, ALU_OP_XOR, !~meta->insn.imm); + return wrp_alu32_imm(nfp_prog, meta, ALU_OP_XOR); } static int and_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) @@ -2319,7 +2327,7 @@ static int and_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) static int and_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { - return wrp_alu32_imm(nfp_prog, meta, ALU_OP_AND, !~meta->insn.imm); + return wrp_alu32_imm(nfp_prog, meta, ALU_OP_AND); } static int or_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) @@ -2329,7 +2337,7 @@ static int or_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) static int or_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { - return wrp_alu32_imm(nfp_prog, meta, ALU_OP_OR, !meta->insn.imm); + return wrp_alu32_imm(nfp_prog, meta, ALU_OP_OR); } static int add_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) @@ -2339,7 +2347,7 @@ static int add_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) static int add_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { - return wrp_alu32_imm(nfp_prog, meta, ALU_OP_ADD, !meta->insn.imm); + return wrp_alu32_imm(nfp_prog, meta, ALU_OP_ADD); } static int sub_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) @@ -2349,7 +2357,7 @@ static int sub_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) static int sub_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { - return wrp_alu32_imm(nfp_prog, meta, ALU_OP_SUB, !meta->insn.imm); + return wrp_alu32_imm(nfp_prog, meta, ALU_OP_SUB); } static int mul_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) @@ -2382,18 +2390,121 @@ static int neg_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) return 0; } +static int __ashr_imm(struct nfp_prog *nfp_prog, u8 dst, u8 shift_amt) +{ + if (shift_amt) { + /* Set signedness bit (MSB of result). */ + emit_alu(nfp_prog, reg_none(), reg_a(dst), ALU_OP_OR, + reg_imm(0)); + emit_shf(nfp_prog, reg_both(dst), reg_none(), SHF_OP_ASHR, + reg_b(dst), SHF_SC_R_SHF, shift_amt); + } + wrp_immed(nfp_prog, reg_both(dst + 1), 0); + + return 0; +} + +static int ashr_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +{ + const struct bpf_insn *insn = &meta->insn; + u64 umin, umax; + u8 dst, src; + + dst = insn->dst_reg * 2; + umin = meta->umin_src; + umax = meta->umax_src; + if (umin == umax) + return __ashr_imm(nfp_prog, dst, umin); + + src = insn->src_reg * 2; + /* NOTE: the first insn will set both indirect shift amount (source A) + * and signedness bit (MSB of result). + */ + emit_alu(nfp_prog, reg_none(), reg_a(src), ALU_OP_OR, reg_b(dst)); + emit_shf_indir(nfp_prog, reg_both(dst), reg_none(), SHF_OP_ASHR, + reg_b(dst), SHF_SC_R_SHF); + wrp_immed(nfp_prog, reg_both(dst + 1), 0); + + return 0; +} + +static int ashr_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +{ + const struct bpf_insn *insn = &meta->insn; + u8 dst = insn->dst_reg * 2; + + return __ashr_imm(nfp_prog, dst, insn->imm); +} + +static int __shr_imm(struct nfp_prog *nfp_prog, u8 dst, u8 shift_amt) +{ + if (shift_amt) + emit_shf(nfp_prog, reg_both(dst), reg_none(), SHF_OP_NONE, + reg_b(dst), SHF_SC_R_SHF, shift_amt); + wrp_immed(nfp_prog, reg_both(dst + 1), 0); + return 0; +} + +static int shr_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +{ + const struct bpf_insn *insn = &meta->insn; + u8 dst = insn->dst_reg * 2; + + return __shr_imm(nfp_prog, dst, insn->imm); +} + +static int shr_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +{ + const struct bpf_insn *insn = &meta->insn; + u64 umin, umax; + u8 dst, src; + + dst = insn->dst_reg * 2; + umin = meta->umin_src; + umax = meta->umax_src; + if (umin == umax) + return __shr_imm(nfp_prog, dst, umin); + + src = insn->src_reg * 2; + emit_alu(nfp_prog, reg_none(), reg_a(src), ALU_OP_OR, reg_imm(0)); + emit_shf_indir(nfp_prog, reg_both(dst), reg_none(), SHF_OP_NONE, + reg_b(dst), SHF_SC_R_SHF); + wrp_immed(nfp_prog, reg_both(dst + 1), 0); + return 0; +} + +static int __shl_imm(struct nfp_prog *nfp_prog, u8 dst, u8 shift_amt) +{ + if (shift_amt) + emit_shf(nfp_prog, reg_both(dst), reg_none(), SHF_OP_NONE, + reg_b(dst), SHF_SC_L_SHF, shift_amt); + wrp_immed(nfp_prog, reg_both(dst + 1), 0); + return 0; +} + static int shl_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { const struct bpf_insn *insn = &meta->insn; + u8 dst = insn->dst_reg * 2; - if (!insn->imm) - return 1; /* TODO: zero shift means indirect */ + return __shl_imm(nfp_prog, dst, insn->imm); +} - emit_shf(nfp_prog, reg_both(insn->dst_reg * 2), - reg_none(), SHF_OP_NONE, reg_b(insn->dst_reg * 2), - SHF_SC_L_SHF, insn->imm); - wrp_immed(nfp_prog, reg_both(insn->dst_reg * 2 + 1), 0); +static int shl_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +{ + const struct bpf_insn *insn = &meta->insn; + u64 umin, umax; + u8 dst, src; + dst = insn->dst_reg * 2; + umin = meta->umin_src; + umax = meta->umax_src; + if (umin == umax) + return __shl_imm(nfp_prog, dst, umin); + + src = insn->src_reg * 2; + shl_reg64_lt32_low(nfp_prog, dst, src); + wrp_immed(nfp_prog, reg_both(dst + 1), 0); return 0; } @@ -3005,30 +3116,37 @@ static int jeq_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) return 0; } -static int jset_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +static int jeq32_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { const struct bpf_insn *insn = &meta->insn; - u64 imm = insn->imm; /* sign extend */ swreg tmp_reg; - if (!imm) { - meta->skip = true; - return 0; - } + tmp_reg = ur_load_imm_any(nfp_prog, insn->imm, imm_b(nfp_prog)); + emit_alu(nfp_prog, reg_none(), + reg_a(insn->dst_reg * 2), ALU_OP_XOR, tmp_reg); + emit_br(nfp_prog, BR_BEQ, insn->off, 0); - if (imm & ~0U) { - tmp_reg = ur_load_imm_any(nfp_prog, imm & ~0U, imm_b(nfp_prog)); - emit_alu(nfp_prog, reg_none(), - reg_a(insn->dst_reg * 2), ALU_OP_AND, tmp_reg); - emit_br(nfp_prog, BR_BNE, insn->off, 0); - } + return 0; +} - if (imm >> 32) { - tmp_reg = ur_load_imm_any(nfp_prog, imm >> 32, imm_b(nfp_prog)); +static int jset_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) +{ + const struct bpf_insn *insn = &meta->insn; + u64 imm = insn->imm; /* sign extend */ + u8 dst_gpr = insn->dst_reg * 2; + swreg tmp_reg; + + tmp_reg = ur_load_imm_any(nfp_prog, imm & ~0U, imm_b(nfp_prog)); + emit_alu(nfp_prog, imm_b(nfp_prog), + reg_a(dst_gpr), ALU_OP_AND, tmp_reg); + /* Upper word of the mask can only be 0 or ~0 from sign extension, + * so either ignore it or OR the whole thing in. + */ + if (is_mbpf_jmp64(meta) && imm >> 32) { emit_alu(nfp_prog, reg_none(), - reg_a(insn->dst_reg * 2 + 1), ALU_OP_AND, tmp_reg); - emit_br(nfp_prog, BR_BNE, insn->off, 0); + reg_a(dst_gpr + 1), ALU_OP_OR, imm_b(nfp_prog)); } + emit_br(nfp_prog, BR_BNE, insn->off, 0); return 0; } @@ -3037,11 +3155,16 @@ static int jne_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) { const struct bpf_insn *insn = &meta->insn; u64 imm = insn->imm; /* sign extend */ + bool is_jmp32 = is_mbpf_jmp32(meta); swreg tmp_reg; if (!imm) { - emit_alu(nfp_prog, reg_none(), reg_a(insn->dst_reg * 2), - ALU_OP_OR, reg_b(insn->dst_reg * 2 + 1)); + if (is_jmp32) + emit_alu(nfp_prog, reg_none(), reg_none(), ALU_OP_NONE, + reg_b(insn->dst_reg * 2)); + else + emit_alu(nfp_prog, reg_none(), reg_a(insn->dst_reg * 2), + ALU_OP_OR, reg_b(insn->dst_reg * 2 + 1)); emit_br(nfp_prog, BR_BNE, insn->off, 0); return 0; } @@ -3051,6 +3174,9 @@ static int jne_imm(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) reg_a(insn->dst_reg * 2), ALU_OP_XOR, tmp_reg); emit_br(nfp_prog, BR_BNE, insn->off, 0); + if (is_jmp32) + return 0; + tmp_reg = ur_load_imm_any(nfp_prog, imm >> 32, imm_b(nfp_prog)); emit_alu(nfp_prog, reg_none(), reg_a(insn->dst_reg * 2 + 1), ALU_OP_XOR, tmp_reg); @@ -3065,10 +3191,13 @@ static int jeq_reg(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) emit_alu(nfp_prog, imm_a(nfp_prog), reg_a(insn->dst_reg * 2), ALU_OP_XOR, reg_b(insn->src_reg * 2)); - emit_alu(nfp_prog, imm_b(nfp_prog), reg_a(insn->dst_reg * 2 + 1), - ALU_OP_XOR, reg_b(insn->src_reg * 2 + 1)); - emit_alu(nfp_prog, reg_none(), - imm_a(nfp_prog), ALU_OP_OR, imm_b(nfp_prog)); + if (is_mbpf_jmp64(meta)) { + emit_alu(nfp_prog, imm_b(nfp_prog), + reg_a(insn->dst_reg * 2 + 1), ALU_OP_XOR, + reg_b(insn->src_reg * 2 + 1)); + emit_alu(nfp_prog, reg_none(), imm_a(nfp_prog), ALU_OP_OR, + imm_b(nfp_prog)); + } emit_br(nfp_prog, BR_BEQ, insn->off, 0); return 0; @@ -3146,7 +3275,7 @@ bpf_to_bpf_call(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta) wrp_immed_relo(nfp_prog, imm_b(nfp_prog), 0, RELO_IMMED_REL); } else { ret_tgt = nfp_prog_current_offset(nfp_prog) + 2; - emit_br(nfp_prog, BR_UNC, meta->n + 1 + meta->insn.imm, 1); + emit_br(nfp_prog, BR_UNC, meta->insn.imm, 1); offset_br = nfp_prog_current_offset(nfp_prog); } wrp_immed_relo(nfp_prog, ret_reg(nfp_prog), ret_tgt, RELO_IMMED_REL); @@ -3285,7 +3414,12 @@ static const instr_cb_t instr_cb[256] = { [BPF_ALU | BPF_DIV | BPF_X] = div_reg, [BPF_ALU | BPF_DIV | BPF_K] = div_imm, [BPF_ALU | BPF_NEG] = neg_reg, + [BPF_ALU | BPF_LSH | BPF_X] = shl_reg, [BPF_ALU | BPF_LSH | BPF_K] = shl_imm, + [BPF_ALU | BPF_RSH | BPF_X] = shr_reg, + [BPF_ALU | BPF_RSH | BPF_K] = shr_imm, + [BPF_ALU | BPF_ARSH | BPF_X] = ashr_reg, + [BPF_ALU | BPF_ARSH | BPF_K] = ashr_imm, [BPF_ALU | BPF_END | BPF_X] = end_reg32, [BPF_LD | BPF_IMM | BPF_DW] = imm_ld8, [BPF_LD | BPF_ABS | BPF_B] = data_ld1, @@ -3331,6 +3465,28 @@ static const instr_cb_t instr_cb[256] = { [BPF_JMP | BPF_JSLE | BPF_X] = cmp_reg, [BPF_JMP | BPF_JSET | BPF_X] = jset_reg, [BPF_JMP | BPF_JNE | BPF_X] = jne_reg, + [BPF_JMP32 | BPF_JEQ | BPF_K] = jeq32_imm, + [BPF_JMP32 | BPF_JGT | BPF_K] = cmp_imm, + [BPF_JMP32 | BPF_JGE | BPF_K] = cmp_imm, + [BPF_JMP32 | BPF_JLT | BPF_K] = cmp_imm, + [BPF_JMP32 | BPF_JLE | BPF_K] = cmp_imm, + [BPF_JMP32 | BPF_JSGT | BPF_K] =cmp_imm, + [BPF_JMP32 | BPF_JSGE | BPF_K] =cmp_imm, + [BPF_JMP32 | BPF_JSLT | BPF_K] =cmp_imm, + [BPF_JMP32 | BPF_JSLE | BPF_K] =cmp_imm, + [BPF_JMP32 | BPF_JSET | BPF_K] =jset_imm, + [BPF_JMP32 | BPF_JNE | BPF_K] = jne_imm, + [BPF_JMP32 | BPF_JEQ | BPF_X] = jeq_reg, + [BPF_JMP32 | BPF_JGT | BPF_X] = cmp_reg, + [BPF_JMP32 | BPF_JGE | BPF_X] = cmp_reg, + [BPF_JMP32 | BPF_JLT | BPF_X] = cmp_reg, + [BPF_JMP32 | BPF_JLE | BPF_X] = cmp_reg, + [BPF_JMP32 | BPF_JSGT | BPF_X] =cmp_reg, + [BPF_JMP32 | BPF_JSGE | BPF_X] =cmp_reg, + [BPF_JMP32 | BPF_JSLT | BPF_X] =cmp_reg, + [BPF_JMP32 | BPF_JSLE | BPF_X] =cmp_reg, + [BPF_JMP32 | BPF_JSET | BPF_X] =jset_reg, + [BPF_JMP32 | BPF_JNE | BPF_X] = jne_reg, [BPF_JMP | BPF_CALL] = call, [BPF_JMP | BPF_EXIT] = jmp_exit, }; @@ -3357,9 +3513,9 @@ static int nfp_fixup_branches(struct nfp_prog *nfp_prog) int err; list_for_each_entry(meta, &nfp_prog->insns, l) { - if (meta->skip) + if (meta->flags & FLAG_INSN_SKIP_MASK) continue; - if (BPF_CLASS(meta->insn.code) != BPF_JMP) + if (!is_mbpf_jmp(meta)) continue; if (meta->insn.code == (BPF_JMP | BPF_EXIT) && !nfp_is_main_function(meta)) @@ -3401,7 +3557,7 @@ static int nfp_fixup_branches(struct nfp_prog *nfp_prog) jmp_dst = meta->jmp_dst; - if (jmp_dst->skip) { + if (jmp_dst->flags & FLAG_INSN_SKIP_PREC_DEPENDENT) { pr_err("Branch landing on removed instruction!!\n"); return -ELOOP; } @@ -3651,7 +3807,7 @@ static int nfp_translate(struct nfp_prog *nfp_prog) return nfp_prog->error; } - if (meta->skip) { + if (meta->flags & FLAG_INSN_SKIP_MASK) { nfp_prog->n_translated++; continue; } @@ -3699,10 +3855,10 @@ static void nfp_bpf_opt_reg_init(struct nfp_prog *nfp_prog) /* Programs start with R6 = R1 but we ignore the skb pointer */ if (insn.code == (BPF_ALU64 | BPF_MOV | BPF_X) && insn.src_reg == 1 && insn.dst_reg == 6) - meta->skip = true; + meta->flags |= FLAG_INSN_SKIP_PREC_DEPENDENT; /* Return as soon as something doesn't match */ - if (!meta->skip) + if (!(meta->flags & FLAG_INSN_SKIP_MASK)) return; } } @@ -3717,19 +3873,17 @@ static void nfp_bpf_opt_neg_add_sub(struct nfp_prog *nfp_prog) list_for_each_entry(meta, &nfp_prog->insns, l) { struct bpf_insn insn = meta->insn; - if (meta->skip) + if (meta->flags & FLAG_INSN_SKIP_MASK) continue; - if (BPF_CLASS(insn.code) != BPF_ALU && - BPF_CLASS(insn.code) != BPF_ALU64 && - BPF_CLASS(insn.code) != BPF_JMP) + if (!is_mbpf_alu(meta) && !is_mbpf_jmp(meta)) continue; if (BPF_SRC(insn.code) != BPF_K) continue; if (insn.imm >= 0) continue; - if (BPF_CLASS(insn.code) == BPF_JMP) { + if (is_mbpf_jmp(meta)) { switch (BPF_OP(insn.code)) { case BPF_JGE: case BPF_JSGE: @@ -3791,7 +3945,7 @@ static void nfp_bpf_opt_ld_mask(struct nfp_prog *nfp_prog) if (meta2->flags & FLAG_INSN_IS_JUMP_DST) continue; - meta2->skip = true; + meta2->flags |= FLAG_INSN_SKIP_PREC_DEPENDENT; } } @@ -3831,8 +3985,8 @@ static void nfp_bpf_opt_ld_shift(struct nfp_prog *nfp_prog) meta3->flags & FLAG_INSN_IS_JUMP_DST) continue; - meta2->skip = true; - meta3->skip = true; + meta2->flags |= FLAG_INSN_SKIP_PREC_DEPENDENT; + meta3->flags |= FLAG_INSN_SKIP_PREC_DEPENDENT; } } @@ -4027,7 +4181,8 @@ static void nfp_bpf_opt_ldst_gather(struct nfp_prog *nfp_prog) } head_ld_meta->paired_st = &head_st_meta->insn; - head_st_meta->skip = true; + head_st_meta->flags |= + FLAG_INSN_SKIP_PREC_DEPENDENT; } else { head_ld_meta->ldst_gather_len = 0; } @@ -4060,8 +4215,8 @@ static void nfp_bpf_opt_ldst_gather(struct nfp_prog *nfp_prog) head_ld_meta = meta1; head_st_meta = meta2; } else { - meta1->skip = true; - meta2->skip = true; + meta1->flags |= FLAG_INSN_SKIP_PREC_DEPENDENT; + meta2->flags |= FLAG_INSN_SKIP_PREC_DEPENDENT; } head_ld_meta->ldst_gather_len += BPF_LDST_BYTES(ld); @@ -4086,7 +4241,7 @@ static void nfp_bpf_opt_pkt_cache(struct nfp_prog *nfp_prog) if (meta->flags & FLAG_INSN_IS_JUMP_DST) cache_avail = false; - if (meta->skip) + if (meta->flags & FLAG_INSN_SKIP_MASK) continue; insn = &meta->insn; @@ -4172,7 +4327,7 @@ start_new: } list_for_each_entry(meta, &nfp_prog->insns, l) { - if (meta->skip) + if (meta->flags & FLAG_INSN_SKIP_MASK) continue; if (is_mbpf_load_pkt(meta) && !meta->ldst_gather_len) { @@ -4208,7 +4363,8 @@ static int nfp_bpf_replace_map_ptrs(struct nfp_prog *nfp_prog) u32 id; nfp_for_each_insn_walk2(nfp_prog, meta1, meta2) { - if (meta1->skip || meta2->skip) + if (meta1->flags & FLAG_INSN_SKIP_MASK || + meta2->flags & FLAG_INSN_SKIP_MASK) continue; if (meta1->insn.code != (BPF_LD | BPF_IMM | BPF_DW) || @@ -4287,7 +4443,7 @@ int nfp_bpf_jit(struct nfp_prog *nfp_prog) return ret; } -void nfp_bpf_jit_prepare(struct nfp_prog *nfp_prog, unsigned int cnt) +void nfp_bpf_jit_prepare(struct nfp_prog *nfp_prog) { struct nfp_insn_meta *meta; @@ -4298,7 +4454,7 @@ void nfp_bpf_jit_prepare(struct nfp_prog *nfp_prog, unsigned int cnt) unsigned int dst_idx; bool pseudo_call; - if (BPF_CLASS(code) != BPF_JMP) + if (!is_mbpf_jmp(meta)) continue; if (BPF_OP(code) == BPF_EXIT) continue; @@ -4315,7 +4471,7 @@ void nfp_bpf_jit_prepare(struct nfp_prog *nfp_prog, unsigned int cnt) else dst_idx = meta->n + 1 + meta->insn.off; - dst_meta = nfp_bpf_goto_meta(nfp_prog, meta, dst_idx, cnt); + dst_meta = nfp_bpf_goto_meta(nfp_prog, meta, dst_idx); if (pseudo_call) dst_meta->flags |= FLAG_INSN_IS_SUBPROG_START; diff --git a/drivers/net/ethernet/netronome/nfp/bpf/main.c b/drivers/net/ethernet/netronome/nfp/bpf/main.c index 6243af0ab025..275de9f4c61c 100644 --- a/drivers/net/ethernet/netronome/nfp/bpf/main.c +++ b/drivers/net/ethernet/netronome/nfp/bpf/main.c @@ -465,7 +465,7 @@ static int nfp_bpf_init(struct nfp_app *app) app->ctrl_mtu = nfp_bpf_ctrl_cmsg_mtu(bpf); } - bpf->bpf_dev = bpf_offload_dev_create(); + bpf->bpf_dev = bpf_offload_dev_create(&nfp_bpf_dev_ops, bpf); err = PTR_ERR_OR_ZERO(bpf->bpf_dev); if (err) goto err_free_neutral_maps; diff --git a/drivers/net/ethernet/netronome/nfp/bpf/main.h b/drivers/net/ethernet/netronome/nfp/bpf/main.h index 7f591d71ab28..b25a48218bcf 100644 --- a/drivers/net/ethernet/netronome/nfp/bpf/main.h +++ b/drivers/net/ethernet/netronome/nfp/bpf/main.h @@ -243,6 +243,16 @@ struct nfp_bpf_reg_state { #define FLAG_INSN_IS_JUMP_DST BIT(0) #define FLAG_INSN_IS_SUBPROG_START BIT(1) #define FLAG_INSN_PTR_CALLER_STACK_FRAME BIT(2) +/* Instruction is pointless, noop even on its own */ +#define FLAG_INSN_SKIP_NOOP BIT(3) +/* Instruction is optimized out based on preceding instructions */ +#define FLAG_INSN_SKIP_PREC_DEPENDENT BIT(4) +/* Instruction is optimized by the verifier */ +#define FLAG_INSN_SKIP_VERIFIER_OPT BIT(5) + +#define FLAG_INSN_SKIP_MASK (FLAG_INSN_SKIP_NOOP | \ + FLAG_INSN_SKIP_PREC_DEPENDENT | \ + FLAG_INSN_SKIP_VERIFIER_OPT) /** * struct nfp_insn_meta - BPF instruction wrapper @@ -271,7 +281,6 @@ struct nfp_bpf_reg_state { * @n: eBPF instruction number * @flags: eBPF instruction extra optimization flags * @subprog_idx: index of subprogram to which the instruction belongs - * @skip: skip this instruction (optimized out) * @double_cb: callback for second part of the instruction * @l: link on nfp_prog->insns list */ @@ -319,7 +328,6 @@ struct nfp_insn_meta { unsigned short n; unsigned short flags; unsigned short subprog_idx; - bool skip; instr_cb_t double_cb; struct list_head l; @@ -357,6 +365,21 @@ static inline bool is_mbpf_load(const struct nfp_insn_meta *meta) return (meta->insn.code & ~BPF_SIZE_MASK) == (BPF_LDX | BPF_MEM); } +static inline bool is_mbpf_jmp32(const struct nfp_insn_meta *meta) +{ + return mbpf_class(meta) == BPF_JMP32; +} + +static inline bool is_mbpf_jmp64(const struct nfp_insn_meta *meta) +{ + return mbpf_class(meta) == BPF_JMP; +} + +static inline bool is_mbpf_jmp(const struct nfp_insn_meta *meta) +{ + return is_mbpf_jmp32(meta) || is_mbpf_jmp64(meta); +} + static inline bool is_mbpf_store(const struct nfp_insn_meta *meta) { return (meta->insn.code & ~BPF_SIZE_MASK) == (BPF_STX | BPF_MEM); @@ -407,6 +430,20 @@ static inline bool is_mbpf_div(const struct nfp_insn_meta *meta) return is_mbpf_alu(meta) && mbpf_op(meta) == BPF_DIV; } +static inline bool is_mbpf_cond_jump(const struct nfp_insn_meta *meta) +{ + u8 op; + + if (is_mbpf_jmp32(meta)) + return true; + + if (!is_mbpf_jmp64(meta)) + return false; + + op = mbpf_op(meta); + return op != BPF_JA && op != BPF_EXIT && op != BPF_CALL; +} + static inline bool is_mbpf_helper_call(const struct nfp_insn_meta *meta) { struct bpf_insn insn = meta->insn; @@ -457,6 +494,7 @@ struct nfp_bpf_subprog_info { * @subprog_cnt: number of sub-programs, including main function * @map_records: the map record pointers from bpf->maps_neutral * @subprog: pointer to an array of objects holding info about sub-programs + * @n_insns: number of instructions on @insns list * @insns: list of BPF instruction wrappers (struct nfp_insn_meta) */ struct nfp_prog { @@ -489,6 +527,7 @@ struct nfp_prog { struct nfp_bpf_neutral_map **map_records; struct nfp_bpf_subprog_info *subprog; + unsigned int n_insns; struct list_head insns; }; @@ -505,11 +544,19 @@ struct nfp_bpf_vnic { }; bool nfp_is_subprog_start(struct nfp_insn_meta *meta); -void nfp_bpf_jit_prepare(struct nfp_prog *nfp_prog, unsigned int cnt); +void nfp_bpf_jit_prepare(struct nfp_prog *nfp_prog); int nfp_bpf_jit(struct nfp_prog *prog); bool nfp_bpf_supported_opcode(u8 code); -extern const struct bpf_prog_offload_ops nfp_bpf_analyzer_ops; +int nfp_verify_insn(struct bpf_verifier_env *env, int insn_idx, + int prev_insn_idx); +int nfp_bpf_finalize(struct bpf_verifier_env *env); + +int nfp_bpf_opt_replace_insn(struct bpf_verifier_env *env, u32 off, + struct bpf_insn *insn); +int nfp_bpf_opt_remove_insns(struct bpf_verifier_env *env, u32 off, u32 cnt); + +extern const struct bpf_prog_offload_ops nfp_bpf_dev_ops; struct netdev_bpf; struct nfp_app; @@ -522,7 +569,7 @@ int nfp_net_bpf_offload(struct nfp_net *nn, struct bpf_prog *prog, struct nfp_insn_meta * nfp_bpf_goto_meta(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, - unsigned int insn_idx, unsigned int n_insns); + unsigned int insn_idx); void *nfp_bpf_relo_for_vnic(struct nfp_prog *nfp_prog, struct nfp_bpf_vnic *bv); diff --git a/drivers/net/ethernet/netronome/nfp/bpf/offload.c b/drivers/net/ethernet/netronome/nfp/bpf/offload.c index ba8ceedcf6a2..15dce97650a5 100644 --- a/drivers/net/ethernet/netronome/nfp/bpf/offload.c +++ b/drivers/net/ethernet/netronome/nfp/bpf/offload.c @@ -33,9 +33,6 @@ nfp_map_ptr_record(struct nfp_app_bpf *bpf, struct nfp_prog *nfp_prog, struct nfp_bpf_neutral_map *record; int err; - /* Map record paths are entered via ndo, update side is protected. */ - ASSERT_RTNL(); - /* Reuse path - other offloaded program is already tracking this map. */ record = rhashtable_lookup_fast(&bpf->maps_neutral, &map->id, nfp_bpf_maps_neutral_params); @@ -84,8 +81,6 @@ nfp_map_ptrs_forget(struct nfp_app_bpf *bpf, struct nfp_prog *nfp_prog) bool freed = false; int i; - ASSERT_RTNL(); - for (i = 0; i < nfp_prog->map_records_cnt; i++) { if (--nfp_prog->map_records[i]->count) { nfp_prog->map_records[i] = NULL; @@ -168,8 +163,9 @@ nfp_prog_prepare(struct nfp_prog *nfp_prog, const struct bpf_insn *prog, list_add_tail(&meta->l, &nfp_prog->insns); } + nfp_prog->n_insns = cnt; - nfp_bpf_jit_prepare(nfp_prog, cnt); + nfp_bpf_jit_prepare(nfp_prog); return 0; } @@ -187,11 +183,8 @@ static void nfp_prog_free(struct nfp_prog *nfp_prog) kfree(nfp_prog); } -static int -nfp_bpf_verifier_prep(struct nfp_app *app, struct nfp_net *nn, - struct netdev_bpf *bpf) +static int nfp_bpf_verifier_prep(struct bpf_prog *prog) { - struct bpf_prog *prog = bpf->verifier.prog; struct nfp_prog *nfp_prog; int ret; @@ -202,14 +195,13 @@ nfp_bpf_verifier_prep(struct nfp_app *app, struct nfp_net *nn, INIT_LIST_HEAD(&nfp_prog->insns); nfp_prog->type = prog->type; - nfp_prog->bpf = app->priv; + nfp_prog->bpf = bpf_offload_dev_priv(prog->aux->offload->offdev); ret = nfp_prog_prepare(nfp_prog, prog->insnsi, prog->len); if (ret) goto err_free; nfp_prog->verifier_meta = nfp_prog_first_meta(nfp_prog); - bpf->verifier.ops = &nfp_bpf_analyzer_ops; return 0; @@ -219,12 +211,17 @@ err_free: return ret; } -static int nfp_bpf_translate(struct nfp_net *nn, struct bpf_prog *prog) +static int nfp_bpf_translate(struct bpf_prog *prog) { + struct nfp_net *nn = netdev_priv(prog->aux->offload->netdev); struct nfp_prog *nfp_prog = prog->aux->offload->dev_priv; unsigned int max_instr; int err; + /* We depend on dead code elimination succeeding */ + if (prog->aux->offload->opt_failed) + return -EINVAL; + max_instr = nn_readw(nn, NFP_NET_CFG_BPF_MAX_LEN); nfp_prog->__prog_alloc_len = max_instr * sizeof(u64); @@ -242,15 +239,13 @@ static int nfp_bpf_translate(struct nfp_net *nn, struct bpf_prog *prog) return nfp_map_ptrs_record(nfp_prog->bpf, nfp_prog, prog); } -static int nfp_bpf_destroy(struct nfp_net *nn, struct bpf_prog *prog) +static void nfp_bpf_destroy(struct bpf_prog *prog) { struct nfp_prog *nfp_prog = prog->aux->offload->dev_priv; kvfree(nfp_prog->prog); nfp_map_ptrs_forget(nfp_prog->bpf, nfp_prog); nfp_prog_free(nfp_prog); - - return 0; } /* Atomic engine requires values to be in big endian, we need to byte swap @@ -422,12 +417,6 @@ nfp_bpf_map_free(struct nfp_app_bpf *bpf, struct bpf_offloaded_map *offmap) int nfp_ndo_bpf(struct nfp_app *app, struct nfp_net *nn, struct netdev_bpf *bpf) { switch (bpf->command) { - case BPF_OFFLOAD_VERIFIER_PREP: - return nfp_bpf_verifier_prep(app, nn, bpf); - case BPF_OFFLOAD_TRANSLATE: - return nfp_bpf_translate(nn, bpf->offload.prog); - case BPF_OFFLOAD_DESTROY: - return nfp_bpf_destroy(nn, bpf->offload.prog); case BPF_OFFLOAD_MAP_ALLOC: return nfp_bpf_map_alloc(app->priv, bpf->offmap); case BPF_OFFLOAD_MAP_FREE: @@ -489,14 +478,15 @@ nfp_net_bpf_load(struct nfp_net *nn, struct bpf_prog *prog, struct netlink_ext_ack *extack) { struct nfp_prog *nfp_prog = prog->aux->offload->dev_priv; - unsigned int max_mtu, max_stack, max_prog_len; + unsigned int fw_mtu, pkt_off, max_stack, max_prog_len; dma_addr_t dma_addr; void *img; int err; - max_mtu = nn_readb(nn, NFP_NET_CFG_BPF_INL_MTU) * 64 - 32; - if (max_mtu < nn->dp.netdev->mtu) { - NL_SET_ERR_MSG_MOD(extack, "BPF offload not supported with MTU larger than HW packet split boundary"); + fw_mtu = nn_readb(nn, NFP_NET_CFG_BPF_INL_MTU) * 64 - 32; + pkt_off = min(prog->aux->max_pkt_offset, nn->dp.netdev->mtu); + if (fw_mtu < pkt_off) { + NL_SET_ERR_MSG_MOD(extack, "BPF offload not supported with potential packet access beyond HW packet split boundary"); return -EOPNOTSUPP; } @@ -600,3 +590,13 @@ int nfp_net_bpf_offload(struct nfp_net *nn, struct bpf_prog *prog, return 0; } + +const struct bpf_prog_offload_ops nfp_bpf_dev_ops = { + .insn_hook = nfp_verify_insn, + .finalize = nfp_bpf_finalize, + .replace_insn = nfp_bpf_opt_replace_insn, + .remove_insns = nfp_bpf_opt_remove_insns, + .prepare = nfp_bpf_verifier_prep, + .translate = nfp_bpf_translate, + .destroy = nfp_bpf_destroy, +}; diff --git a/drivers/net/ethernet/netronome/nfp/bpf/verifier.c b/drivers/net/ethernet/netronome/nfp/bpf/verifier.c index 99f977bfd8cc..36f56eb4cbe2 100644 --- a/drivers/net/ethernet/netronome/nfp/bpf/verifier.c +++ b/drivers/net/ethernet/netronome/nfp/bpf/verifier.c @@ -18,15 +18,15 @@ struct nfp_insn_meta * nfp_bpf_goto_meta(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, - unsigned int insn_idx, unsigned int n_insns) + unsigned int insn_idx) { unsigned int forward, backward, i; backward = meta->n - insn_idx; forward = insn_idx - meta->n; - if (min(forward, backward) > n_insns - insn_idx - 1) { - backward = n_insns - insn_idx - 1; + if (min(forward, backward) > nfp_prog->n_insns - insn_idx - 1) { + backward = nfp_prog->n_insns - insn_idx - 1; meta = nfp_prog_last_meta(nfp_prog); } if (min(forward, backward) > insn_idx && backward > insn_idx) { @@ -623,13 +623,13 @@ nfp_bpf_check_alu(struct nfp_prog *nfp_prog, struct nfp_insn_meta *meta, return 0; } -static int -nfp_verify_insn(struct bpf_verifier_env *env, int insn_idx, int prev_insn_idx) +int nfp_verify_insn(struct bpf_verifier_env *env, int insn_idx, + int prev_insn_idx) { struct nfp_prog *nfp_prog = env->prog->aux->offload->dev_priv; struct nfp_insn_meta *meta = nfp_prog->verifier_meta; - meta = nfp_bpf_goto_meta(nfp_prog, meta, insn_idx, env->prog->len); + meta = nfp_bpf_goto_meta(nfp_prog, meta, insn_idx); nfp_prog->verifier_meta = meta; if (!nfp_bpf_supported_opcode(meta->insn.code)) { @@ -690,8 +690,7 @@ nfp_assign_subprog_idx_and_regs(struct bpf_verifier_env *env, return 0; } -static unsigned int -nfp_bpf_get_stack_usage(struct nfp_prog *nfp_prog, unsigned int cnt) +static unsigned int nfp_bpf_get_stack_usage(struct nfp_prog *nfp_prog) { struct nfp_insn_meta *meta = nfp_prog_first_meta(nfp_prog); unsigned int max_depth = 0, depth = 0, frame = 0; @@ -726,7 +725,7 @@ continue_subprog: /* Find the callee and start processing it. */ meta = nfp_bpf_goto_meta(nfp_prog, meta, - meta->n + 1 + meta->insn.imm, cnt); + meta->n + 1 + meta->insn.imm); idx = meta->subprog_idx; frame++; goto process_subprog; @@ -745,7 +744,7 @@ continue_subprog: goto continue_subprog; } -static int nfp_bpf_finalize(struct bpf_verifier_env *env) +int nfp_bpf_finalize(struct bpf_verifier_env *env) { struct bpf_subprog_info *info; struct nfp_prog *nfp_prog; @@ -778,8 +777,7 @@ static int nfp_bpf_finalize(struct bpf_verifier_env *env) nn = netdev_priv(env->prog->aux->offload->netdev); max_stack = nn_readb(nn, NFP_NET_CFG_BPF_STACK_SZ) * 64; - nfp_prog->stack_size = nfp_bpf_get_stack_usage(nfp_prog, - env->prog->len); + nfp_prog->stack_size = nfp_bpf_get_stack_usage(nfp_prog); if (nfp_prog->stack_size > max_stack) { pr_vlog(env, "stack too large: program %dB > FW stack %dB\n", nfp_prog->stack_size, max_stack); @@ -789,7 +787,60 @@ static int nfp_bpf_finalize(struct bpf_verifier_env *env) return 0; } -const struct bpf_prog_offload_ops nfp_bpf_analyzer_ops = { - .insn_hook = nfp_verify_insn, - .finalize = nfp_bpf_finalize, -}; +int nfp_bpf_opt_replace_insn(struct bpf_verifier_env *env, u32 off, + struct bpf_insn *insn) +{ + struct nfp_prog *nfp_prog = env->prog->aux->offload->dev_priv; + struct bpf_insn_aux_data *aux_data = env->insn_aux_data; + struct nfp_insn_meta *meta = nfp_prog->verifier_meta; + + meta = nfp_bpf_goto_meta(nfp_prog, meta, aux_data[off].orig_idx); + nfp_prog->verifier_meta = meta; + + /* conditional jump to jump conversion */ + if (is_mbpf_cond_jump(meta) && + insn->code == (BPF_JMP | BPF_JA | BPF_K)) { + unsigned int tgt_off; + + tgt_off = off + insn->off + 1; + + if (!insn->off) { + meta->jmp_dst = list_next_entry(meta, l); + meta->jump_neg_op = false; + } else if (meta->jmp_dst->n != aux_data[tgt_off].orig_idx) { + pr_vlog(env, "branch hard wire at %d changes target %d -> %d\n", + off, meta->jmp_dst->n, + aux_data[tgt_off].orig_idx); + return -EINVAL; + } + return 0; + } + + pr_vlog(env, "unsupported instruction replacement %hhx -> %hhx\n", + meta->insn.code, insn->code); + return -EINVAL; +} + +int nfp_bpf_opt_remove_insns(struct bpf_verifier_env *env, u32 off, u32 cnt) +{ + struct nfp_prog *nfp_prog = env->prog->aux->offload->dev_priv; + struct bpf_insn_aux_data *aux_data = env->insn_aux_data; + struct nfp_insn_meta *meta = nfp_prog->verifier_meta; + unsigned int i; + + meta = nfp_bpf_goto_meta(nfp_prog, meta, aux_data[off].orig_idx); + + for (i = 0; i < cnt; i++) { + if (WARN_ON_ONCE(&meta->l == &nfp_prog->insns)) + return -EINVAL; + + /* doesn't count if it already has the flag */ + if (meta->flags & FLAG_INSN_SKIP_VERIFIER_OPT) + i--; + + meta->flags |= FLAG_INSN_SKIP_VERIFIER_OPT; + meta = list_next_entry(meta, l); + } + + return 0; +} diff --git a/drivers/net/ethernet/netronome/nfp/flower/action.c b/drivers/net/ethernet/netronome/nfp/flower/action.c index 244dc261006e..e336f6ee94f5 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/action.c +++ b/drivers/net/ethernet/netronome/nfp/flower/action.c @@ -2,9 +2,7 @@ /* Copyright (C) 2017-2018 Netronome Systems, Inc. */ #include <linux/bitfield.h> -#include <net/geneve.h> #include <net/pkt_cls.h> -#include <net/switchdev.h> #include <net/tc_act/tc_csum.h> #include <net/tc_act/tc_gact.h> #include <net/tc_act/tc_mirred.h> @@ -38,7 +36,7 @@ static void nfp_fl_pop_vlan(struct nfp_fl_pop_vlan *pop_vlan) static void nfp_fl_push_vlan(struct nfp_fl_push_vlan *push_vlan, - const struct tc_action *action) + const struct flow_action_entry *act) { size_t act_size = sizeof(struct nfp_fl_push_vlan); u16 tmp_push_vlan_tci; @@ -46,17 +44,16 @@ nfp_fl_push_vlan(struct nfp_fl_push_vlan *push_vlan, push_vlan->head.jump_id = NFP_FL_ACTION_OPCODE_PUSH_VLAN; push_vlan->head.len_lw = act_size >> NFP_FL_LW_SIZ; push_vlan->reserved = 0; - push_vlan->vlan_tpid = tcf_vlan_push_proto(action); + push_vlan->vlan_tpid = act->vlan.proto; tmp_push_vlan_tci = - FIELD_PREP(NFP_FL_PUSH_VLAN_PRIO, tcf_vlan_push_prio(action)) | - FIELD_PREP(NFP_FL_PUSH_VLAN_VID, tcf_vlan_push_vid(action)) | - NFP_FL_PUSH_VLAN_CFI; + FIELD_PREP(NFP_FL_PUSH_VLAN_PRIO, act->vlan.prio) | + FIELD_PREP(NFP_FL_PUSH_VLAN_VID, act->vlan.vid); push_vlan->vlan_tci = cpu_to_be16(tmp_push_vlan_tci); } static int -nfp_fl_pre_lag(struct nfp_app *app, const struct tc_action *action, +nfp_fl_pre_lag(struct nfp_app *app, const struct flow_action_entry *act, struct nfp_fl_payload *nfp_flow, int act_len) { size_t act_size = sizeof(struct nfp_fl_pre_lag); @@ -64,7 +61,7 @@ nfp_fl_pre_lag(struct nfp_app *app, const struct tc_action *action, struct net_device *out_dev; int err; - out_dev = tcf_mirred_dev(action); + out_dev = act->dev; if (!out_dev || !netif_is_lag_master(out_dev)) return 0; @@ -91,24 +88,10 @@ nfp_fl_pre_lag(struct nfp_app *app, const struct tc_action *action, return act_size; } -static bool nfp_fl_netdev_is_tunnel_type(struct net_device *out_dev, - enum nfp_flower_tun_type tun_type) -{ - if (!out_dev->rtnl_link_ops) - return false; - - if (!strcmp(out_dev->rtnl_link_ops->kind, "vxlan")) - return tun_type == NFP_FL_TUNNEL_VXLAN; - - if (!strcmp(out_dev->rtnl_link_ops->kind, "geneve")) - return tun_type == NFP_FL_TUNNEL_GENEVE; - - return false; -} - static int nfp_fl_output(struct nfp_app *app, struct nfp_fl_output *output, - const struct tc_action *action, struct nfp_fl_payload *nfp_flow, + const struct flow_action_entry *act, + struct nfp_fl_payload *nfp_flow, bool last, struct net_device *in_dev, enum nfp_flower_tun_type tun_type, int *tun_out_cnt) { @@ -120,7 +103,7 @@ nfp_fl_output(struct nfp_app *app, struct nfp_fl_output *output, output->head.jump_id = NFP_FL_ACTION_OPCODE_OUTPUT; output->head.len_lw = act_size >> NFP_FL_LW_SIZ; - out_dev = tcf_mirred_dev(action); + out_dev = act->dev; if (!out_dev) return -EOPNOTSUPP; @@ -151,11 +134,12 @@ nfp_fl_output(struct nfp_app *app, struct nfp_fl_output *output, /* Set action output parameters. */ output->flags = cpu_to_be16(tmp_flags); - /* Only offload if egress ports are on the same device as the - * ingress port. - */ - if (!switchdev_port_same_parent_id(in_dev, out_dev)) - return -EOPNOTSUPP; + if (nfp_netdev_is_nfp_repr(in_dev)) { + /* Confirm ingress and egress are on same device. */ + if (!netdev_port_same_parent_id(in_dev, out_dev)) + return -EOPNOTSUPP; + } + if (!nfp_netdev_is_nfp_repr(out_dev)) return -EOPNOTSUPP; @@ -170,9 +154,9 @@ nfp_fl_output(struct nfp_app *app, struct nfp_fl_output *output, static enum nfp_flower_tun_type nfp_fl_get_tun_from_act_l4_port(struct nfp_app *app, - const struct tc_action *action) + const struct flow_action_entry *act) { - struct ip_tunnel_info *tun = tcf_tunnel_info(action); + const struct ip_tunnel_info *tun = act->tunnel; struct nfp_flower_priv *priv = app->priv; switch (tun->key.tp_dst) { @@ -210,9 +194,9 @@ static struct nfp_fl_pre_tunnel *nfp_fl_pre_tunnel(char *act_data, int act_len) static int nfp_fl_push_geneve_options(struct nfp_fl_payload *nfp_fl, int *list_len, - const struct tc_action *action) + const struct flow_action_entry *act) { - struct ip_tunnel_info *ip_tun = tcf_tunnel_info(action); + struct ip_tunnel_info *ip_tun = (struct ip_tunnel_info *)act->tunnel; int opt_len, opt_cnt, act_start, tot_push_len; u8 *src = ip_tunnel_info_opts(ip_tun); @@ -274,13 +258,13 @@ nfp_fl_push_geneve_options(struct nfp_fl_payload *nfp_fl, int *list_len, static int nfp_fl_set_ipv4_udp_tun(struct nfp_app *app, struct nfp_fl_set_ipv4_udp_tun *set_tun, - const struct tc_action *action, + const struct flow_action_entry *act, struct nfp_fl_pre_tunnel *pre_tun, enum nfp_flower_tun_type tun_type, struct net_device *netdev) { size_t act_size = sizeof(struct nfp_fl_set_ipv4_udp_tun); - struct ip_tunnel_info *ip_tun = tcf_tunnel_info(action); + const struct ip_tunnel_info *ip_tun = act->tunnel; struct nfp_flower_priv *priv = app->priv; u32 tmp_set_ip_tun_type_index = 0; /* Currently support one pre-tunnel so index is always 0. */ @@ -360,7 +344,7 @@ static void nfp_fl_set_helper32(u32 value, u32 mask, u8 *p_exact, u8 *p_mask) } static int -nfp_fl_set_eth(const struct tc_action *action, int idx, u32 off, +nfp_fl_set_eth(const struct flow_action_entry *act, u32 off, struct nfp_fl_set_eth *set_eth) { u32 exact, mask; @@ -368,8 +352,8 @@ nfp_fl_set_eth(const struct tc_action *action, int idx, u32 off, if (off + 4 > ETH_ALEN * 2) return -EOPNOTSUPP; - mask = ~tcf_pedit_mask(action, idx); - exact = tcf_pedit_val(action, idx); + mask = ~act->mangle.mask; + exact = act->mangle.val; if (exact & ~mask) return -EOPNOTSUPP; @@ -384,15 +368,26 @@ nfp_fl_set_eth(const struct tc_action *action, int idx, u32 off, return 0; } +struct ipv4_ttl_word { + __u8 ttl; + __u8 protocol; + __sum16 check; +}; + static int -nfp_fl_set_ip4(const struct tc_action *action, int idx, u32 off, - struct nfp_fl_set_ip4_addrs *set_ip_addr) +nfp_fl_set_ip4(const struct flow_action_entry *act, u32 off, + struct nfp_fl_set_ip4_addrs *set_ip_addr, + struct nfp_fl_set_ip4_ttl_tos *set_ip_ttl_tos) { + struct ipv4_ttl_word *ttl_word_mask; + struct ipv4_ttl_word *ttl_word; + struct iphdr *tos_word_mask; + struct iphdr *tos_word; __be32 exact, mask; /* We are expecting tcf_pedit to return a big endian value */ - mask = (__force __be32)~tcf_pedit_mask(action, idx); - exact = (__force __be32)tcf_pedit_val(action, idx); + mask = (__force __be32)~act->mangle.mask; + exact = (__force __be32)act->mangle.val; if (exact & ~mask) return -EOPNOTSUPP; @@ -402,20 +397,53 @@ nfp_fl_set_ip4(const struct tc_action *action, int idx, u32 off, set_ip_addr->ipv4_dst_mask |= mask; set_ip_addr->ipv4_dst &= ~mask; set_ip_addr->ipv4_dst |= exact & mask; + set_ip_addr->head.jump_id = NFP_FL_ACTION_OPCODE_SET_IPV4_ADDRS; + set_ip_addr->head.len_lw = sizeof(*set_ip_addr) >> + NFP_FL_LW_SIZ; break; case offsetof(struct iphdr, saddr): set_ip_addr->ipv4_src_mask |= mask; set_ip_addr->ipv4_src &= ~mask; set_ip_addr->ipv4_src |= exact & mask; + set_ip_addr->head.jump_id = NFP_FL_ACTION_OPCODE_SET_IPV4_ADDRS; + set_ip_addr->head.len_lw = sizeof(*set_ip_addr) >> + NFP_FL_LW_SIZ; + break; + case offsetof(struct iphdr, ttl): + ttl_word_mask = (struct ipv4_ttl_word *)&mask; + ttl_word = (struct ipv4_ttl_word *)&exact; + + if (ttl_word_mask->protocol || ttl_word_mask->check) + return -EOPNOTSUPP; + + set_ip_ttl_tos->ipv4_ttl_mask |= ttl_word_mask->ttl; + set_ip_ttl_tos->ipv4_ttl &= ~ttl_word_mask->ttl; + set_ip_ttl_tos->ipv4_ttl |= ttl_word->ttl & ttl_word_mask->ttl; + set_ip_ttl_tos->head.jump_id = + NFP_FL_ACTION_OPCODE_SET_IPV4_TTL_TOS; + set_ip_ttl_tos->head.len_lw = sizeof(*set_ip_ttl_tos) >> + NFP_FL_LW_SIZ; + break; + case round_down(offsetof(struct iphdr, tos), 4): + tos_word_mask = (struct iphdr *)&mask; + tos_word = (struct iphdr *)&exact; + + if (tos_word_mask->version || tos_word_mask->ihl || + tos_word_mask->tot_len) + return -EOPNOTSUPP; + + set_ip_ttl_tos->ipv4_tos_mask |= tos_word_mask->tos; + set_ip_ttl_tos->ipv4_tos &= ~tos_word_mask->tos; + set_ip_ttl_tos->ipv4_tos |= tos_word->tos & tos_word_mask->tos; + set_ip_ttl_tos->head.jump_id = + NFP_FL_ACTION_OPCODE_SET_IPV4_TTL_TOS; + set_ip_ttl_tos->head.len_lw = sizeof(*set_ip_ttl_tos) >> + NFP_FL_LW_SIZ; break; default: return -EOPNOTSUPP; } - set_ip_addr->reserved = cpu_to_be16(0); - set_ip_addr->head.jump_id = NFP_FL_ACTION_OPCODE_SET_IPV4_ADDRS; - set_ip_addr->head.len_lw = sizeof(*set_ip_addr) >> NFP_FL_LW_SIZ; - return 0; } @@ -432,23 +460,69 @@ nfp_fl_set_ip6_helper(int opcode_tag, u8 word, __be32 exact, __be32 mask, ip6->head.len_lw = sizeof(*ip6) >> NFP_FL_LW_SIZ; } +struct ipv6_hop_limit_word { + __be16 payload_len; + u8 nexthdr; + u8 hop_limit; +}; + static int -nfp_fl_set_ip6(const struct tc_action *action, int idx, u32 off, +nfp_fl_set_ip6_hop_limit_flow_label(u32 off, __be32 exact, __be32 mask, + struct nfp_fl_set_ipv6_tc_hl_fl *ip_hl_fl) +{ + struct ipv6_hop_limit_word *fl_hl_mask; + struct ipv6_hop_limit_word *fl_hl; + + switch (off) { + case offsetof(struct ipv6hdr, payload_len): + fl_hl_mask = (struct ipv6_hop_limit_word *)&mask; + fl_hl = (struct ipv6_hop_limit_word *)&exact; + + if (fl_hl_mask->nexthdr || fl_hl_mask->payload_len) + return -EOPNOTSUPP; + + ip_hl_fl->ipv6_hop_limit_mask |= fl_hl_mask->hop_limit; + ip_hl_fl->ipv6_hop_limit &= ~fl_hl_mask->hop_limit; + ip_hl_fl->ipv6_hop_limit |= fl_hl->hop_limit & + fl_hl_mask->hop_limit; + break; + case round_down(offsetof(struct ipv6hdr, flow_lbl), 4): + if (mask & ~IPV6_FLOW_LABEL_MASK || + exact & ~IPV6_FLOW_LABEL_MASK) + return -EOPNOTSUPP; + + ip_hl_fl->ipv6_label_mask |= mask; + ip_hl_fl->ipv6_label &= ~mask; + ip_hl_fl->ipv6_label |= exact & mask; + break; + } + + ip_hl_fl->head.jump_id = NFP_FL_ACTION_OPCODE_SET_IPV6_TC_HL_FL; + ip_hl_fl->head.len_lw = sizeof(*ip_hl_fl) >> NFP_FL_LW_SIZ; + + return 0; +} + +static int +nfp_fl_set_ip6(const struct flow_action_entry *act, u32 off, struct nfp_fl_set_ipv6_addr *ip_dst, - struct nfp_fl_set_ipv6_addr *ip_src) + struct nfp_fl_set_ipv6_addr *ip_src, + struct nfp_fl_set_ipv6_tc_hl_fl *ip_hl_fl) { __be32 exact, mask; + int err = 0; u8 word; /* We are expecting tcf_pedit to return a big endian value */ - mask = (__force __be32)~tcf_pedit_mask(action, idx); - exact = (__force __be32)tcf_pedit_val(action, idx); + mask = (__force __be32)~act->mangle.mask; + exact = (__force __be32)act->mangle.val; if (exact & ~mask) return -EOPNOTSUPP; if (off < offsetof(struct ipv6hdr, saddr)) { - return -EOPNOTSUPP; + err = nfp_fl_set_ip6_hop_limit_flow_label(off, exact, mask, + ip_hl_fl); } else if (off < offsetof(struct ipv6hdr, daddr)) { word = (off - offsetof(struct ipv6hdr, saddr)) / sizeof(exact); nfp_fl_set_ip6_helper(NFP_FL_ACTION_OPCODE_SET_IPV6_SRC, word, @@ -462,11 +536,11 @@ nfp_fl_set_ip6(const struct tc_action *action, int idx, u32 off, return -EOPNOTSUPP; } - return 0; + return err; } static int -nfp_fl_set_tport(const struct tc_action *action, int idx, u32 off, +nfp_fl_set_tport(const struct flow_action_entry *act, u32 off, struct nfp_fl_set_tport *set_tport, int opcode) { u32 exact, mask; @@ -474,8 +548,8 @@ nfp_fl_set_tport(const struct tc_action *action, int idx, u32 off, if (off) return -EOPNOTSUPP; - mask = ~tcf_pedit_mask(action, idx); - exact = tcf_pedit_val(action, idx); + mask = ~act->mangle.mask; + exact = act->mangle.val; if (exact & ~mask) return -EOPNOTSUPP; @@ -509,67 +583,65 @@ static u32 nfp_fl_csum_l4_to_flag(u8 ip_proto) } static int -nfp_fl_pedit(const struct tc_action *action, struct tc_cls_flower_offload *flow, +nfp_fl_pedit(const struct flow_action_entry *act, + struct tc_cls_flower_offload *flow, char *nfp_action, int *a_len, u32 *csum_updated) { + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); struct nfp_fl_set_ipv6_addr set_ip6_dst, set_ip6_src; + struct nfp_fl_set_ipv6_tc_hl_fl set_ip6_tc_hl_fl; + struct nfp_fl_set_ip4_ttl_tos set_ip_ttl_tos; struct nfp_fl_set_ip4_addrs set_ip_addr; + enum flow_action_mangle_base htype; struct nfp_fl_set_tport set_tport; struct nfp_fl_set_eth set_eth; - enum pedit_header_type htype; - int idx, nkeys, err; size_t act_size = 0; - u32 offset, cmd; u8 ip_proto = 0; + u32 offset; + int err; + memset(&set_ip6_tc_hl_fl, 0, sizeof(set_ip6_tc_hl_fl)); + memset(&set_ip_ttl_tos, 0, sizeof(set_ip_ttl_tos)); memset(&set_ip6_dst, 0, sizeof(set_ip6_dst)); memset(&set_ip6_src, 0, sizeof(set_ip6_src)); memset(&set_ip_addr, 0, sizeof(set_ip_addr)); memset(&set_tport, 0, sizeof(set_tport)); memset(&set_eth, 0, sizeof(set_eth)); - nkeys = tcf_pedit_nkeys(action); - for (idx = 0; idx < nkeys; idx++) { - cmd = tcf_pedit_cmd(action, idx); - htype = tcf_pedit_htype(action, idx); - offset = tcf_pedit_offset(action, idx); + htype = act->mangle.htype; + offset = act->mangle.offset; - if (cmd != TCA_PEDIT_KEY_EX_CMD_SET) - return -EOPNOTSUPP; - - switch (htype) { - case TCA_PEDIT_KEY_EX_HDR_TYPE_ETH: - err = nfp_fl_set_eth(action, idx, offset, &set_eth); - break; - case TCA_PEDIT_KEY_EX_HDR_TYPE_IP4: - err = nfp_fl_set_ip4(action, idx, offset, &set_ip_addr); - break; - case TCA_PEDIT_KEY_EX_HDR_TYPE_IP6: - err = nfp_fl_set_ip6(action, idx, offset, &set_ip6_dst, - &set_ip6_src); - break; - case TCA_PEDIT_KEY_EX_HDR_TYPE_TCP: - err = nfp_fl_set_tport(action, idx, offset, &set_tport, - NFP_FL_ACTION_OPCODE_SET_TCP); - break; - case TCA_PEDIT_KEY_EX_HDR_TYPE_UDP: - err = nfp_fl_set_tport(action, idx, offset, &set_tport, - NFP_FL_ACTION_OPCODE_SET_UDP); - break; - default: - return -EOPNOTSUPP; - } - if (err) - return err; + switch (htype) { + case TCA_PEDIT_KEY_EX_HDR_TYPE_ETH: + err = nfp_fl_set_eth(act, offset, &set_eth); + break; + case TCA_PEDIT_KEY_EX_HDR_TYPE_IP4: + err = nfp_fl_set_ip4(act, offset, &set_ip_addr, + &set_ip_ttl_tos); + break; + case TCA_PEDIT_KEY_EX_HDR_TYPE_IP6: + err = nfp_fl_set_ip6(act, offset, &set_ip6_dst, + &set_ip6_src, &set_ip6_tc_hl_fl); + break; + case TCA_PEDIT_KEY_EX_HDR_TYPE_TCP: + err = nfp_fl_set_tport(act, offset, &set_tport, + NFP_FL_ACTION_OPCODE_SET_TCP); + break; + case TCA_PEDIT_KEY_EX_HDR_TYPE_UDP: + err = nfp_fl_set_tport(act, offset, &set_tport, + NFP_FL_ACTION_OPCODE_SET_UDP); + break; + default: + return -EOPNOTSUPP; } + if (err) + return err; - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_BASIC)) { - struct flow_dissector_key_basic *basic; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) { + struct flow_match_basic match; - basic = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_BASIC, - flow->key); - ip_proto = basic->ip_proto; + flow_rule_match_basic(rule, &match); + ip_proto = match.key->ip_proto; } if (set_eth.head.len_lw) { @@ -577,6 +649,16 @@ nfp_fl_pedit(const struct tc_action *action, struct tc_cls_flower_offload *flow, memcpy(nfp_action, &set_eth, act_size); *a_len += act_size; } + if (set_ip_ttl_tos.head.len_lw) { + nfp_action += act_size; + act_size = sizeof(set_ip_ttl_tos); + memcpy(nfp_action, &set_ip_ttl_tos, act_size); + *a_len += act_size; + + /* Hardware will automatically fix IPv4 and TCP/UDP checksum. */ + *csum_updated |= TCA_CSUM_UPDATE_FLAG_IPV4HDR | + nfp_fl_csum_l4_to_flag(ip_proto); + } if (set_ip_addr.head.len_lw) { nfp_action += act_size; act_size = sizeof(set_ip_addr); @@ -587,6 +669,15 @@ nfp_fl_pedit(const struct tc_action *action, struct tc_cls_flower_offload *flow, *csum_updated |= TCA_CSUM_UPDATE_FLAG_IPV4HDR | nfp_fl_csum_l4_to_flag(ip_proto); } + if (set_ip6_tc_hl_fl.head.len_lw) { + nfp_action += act_size; + act_size = sizeof(set_ip6_tc_hl_fl); + memcpy(nfp_action, &set_ip6_tc_hl_fl, act_size); + *a_len += act_size; + + /* Hardware will automatically fix TCP/UDP checksum. */ + *csum_updated |= nfp_fl_csum_l4_to_flag(ip_proto); + } if (set_ip6_dst.head.len_lw && set_ip6_src.head.len_lw) { /* TC compiles set src and dst IPv6 address as a single action, * the hardware requires this to be 2 separate actions. @@ -634,7 +725,7 @@ nfp_fl_pedit(const struct tc_action *action, struct tc_cls_flower_offload *flow, } static int -nfp_flower_output_action(struct nfp_app *app, const struct tc_action *a, +nfp_flower_output_action(struct nfp_app *app, const struct flow_action_entry *act, struct nfp_fl_payload *nfp_fl, int *a_len, struct net_device *netdev, bool last, enum nfp_flower_tun_type *tun_type, int *tun_out_cnt, @@ -654,7 +745,7 @@ nfp_flower_output_action(struct nfp_app *app, const struct tc_action *a, return -EOPNOTSUPP; output = (struct nfp_fl_output *)&nfp_fl->action_data[*a_len]; - err = nfp_fl_output(app, output, a, nfp_fl, last, netdev, *tun_type, + err = nfp_fl_output(app, output, act, nfp_fl, last, netdev, *tun_type, tun_out_cnt); if (err) return err; @@ -665,7 +756,7 @@ nfp_flower_output_action(struct nfp_app *app, const struct tc_action *a, /* nfp_fl_pre_lag returns -err or size of prelag action added. * This will be 0 if it is not egressing to a lag dev. */ - prelag_size = nfp_fl_pre_lag(app, a, nfp_fl, *a_len); + prelag_size = nfp_fl_pre_lag(app, act, nfp_fl, *a_len); if (prelag_size < 0) return prelag_size; else if (prelag_size > 0 && (!last || *out_cnt)) @@ -679,7 +770,7 @@ nfp_flower_output_action(struct nfp_app *app, const struct tc_action *a, } static int -nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a, +nfp_flower_loop_action(struct nfp_app *app, const struct flow_action_entry *act, struct tc_cls_flower_offload *flow, struct nfp_fl_payload *nfp_fl, int *a_len, struct net_device *netdev, @@ -692,23 +783,25 @@ nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a, struct nfp_fl_pop_vlan *pop_v; int err; - if (is_tcf_gact_shot(a)) { + switch (act->id) { + case FLOW_ACTION_DROP: nfp_fl->meta.shortcut = cpu_to_be32(NFP_FL_SC_ACT_DROP); - } else if (is_tcf_mirred_egress_redirect(a)) { - err = nfp_flower_output_action(app, a, nfp_fl, a_len, netdev, + break; + case FLOW_ACTION_REDIRECT: + err = nfp_flower_output_action(app, act, nfp_fl, a_len, netdev, true, tun_type, tun_out_cnt, out_cnt, csum_updated); if (err) return err; - - } else if (is_tcf_mirred_egress_mirror(a)) { - err = nfp_flower_output_action(app, a, nfp_fl, a_len, netdev, + break; + case FLOW_ACTION_MIRRED: + err = nfp_flower_output_action(app, act, nfp_fl, a_len, netdev, false, tun_type, tun_out_cnt, out_cnt, csum_updated); if (err) return err; - - } else if (is_tcf_vlan(a) && tcf_vlan_action(a) == TCA_VLAN_ACT_POP) { + break; + case FLOW_ACTION_VLAN_POP: if (*a_len + sizeof(struct nfp_fl_pop_vlan) > NFP_FL_MAX_A_SIZ) return -EOPNOTSUPP; @@ -717,20 +810,21 @@ nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a, nfp_fl_pop_vlan(pop_v); *a_len += sizeof(struct nfp_fl_pop_vlan); - } else if (is_tcf_vlan(a) && tcf_vlan_action(a) == TCA_VLAN_ACT_PUSH) { + break; + case FLOW_ACTION_VLAN_PUSH: if (*a_len + sizeof(struct nfp_fl_push_vlan) > NFP_FL_MAX_A_SIZ) return -EOPNOTSUPP; psh_v = (struct nfp_fl_push_vlan *)&nfp_fl->action_data[*a_len]; nfp_fl->meta.shortcut = cpu_to_be32(NFP_FL_SC_ACT_NULL); - nfp_fl_push_vlan(psh_v, a); + nfp_fl_push_vlan(psh_v, act); *a_len += sizeof(struct nfp_fl_push_vlan); - } else if (is_tcf_tunnel_set(a)) { - struct ip_tunnel_info *ip_tun = tcf_tunnel_info(a); - struct nfp_repr *repr = netdev_priv(netdev); + break; + case FLOW_ACTION_TUNNEL_ENCAP: { + const struct ip_tunnel_info *ip_tun = act->tunnel; - *tun_type = nfp_fl_get_tun_from_act_l4_port(repr->app, a); + *tun_type = nfp_fl_get_tun_from_act_l4_port(app, act); if (*tun_type == NFP_FL_TUNNEL_NONE) return -EOPNOTSUPP; @@ -749,32 +843,36 @@ nfp_flower_loop_action(struct nfp_app *app, const struct tc_action *a, nfp_fl->meta.shortcut = cpu_to_be32(NFP_FL_SC_ACT_NULL); *a_len += sizeof(struct nfp_fl_pre_tunnel); - err = nfp_fl_push_geneve_options(nfp_fl, a_len, a); + err = nfp_fl_push_geneve_options(nfp_fl, a_len, act); if (err) return err; set_tun = (void *)&nfp_fl->action_data[*a_len]; - err = nfp_fl_set_ipv4_udp_tun(app, set_tun, a, pre_tun, + err = nfp_fl_set_ipv4_udp_tun(app, set_tun, act, pre_tun, *tun_type, netdev); if (err) return err; *a_len += sizeof(struct nfp_fl_set_ipv4_udp_tun); - } else if (is_tcf_tunnel_release(a)) { + } + break; + case FLOW_ACTION_TUNNEL_DECAP: /* Tunnel decap is handled by default so accept action. */ return 0; - } else if (is_tcf_pedit(a)) { - if (nfp_fl_pedit(a, flow, &nfp_fl->action_data[*a_len], + case FLOW_ACTION_MANGLE: + if (nfp_fl_pedit(act, flow, &nfp_fl->action_data[*a_len], a_len, csum_updated)) return -EOPNOTSUPP; - } else if (is_tcf_csum(a)) { + break; + case FLOW_ACTION_CSUM: /* csum action requests recalc of something we have not fixed */ - if (tcf_csum_update_flags(a) & ~*csum_updated) + if (act->csum_flags & ~*csum_updated) return -EOPNOTSUPP; /* If we will correctly fix the csum we can remove it from the * csum update list. Which will later be used to check support. */ - *csum_updated &= ~tcf_csum_update_flags(a); - } else { + *csum_updated &= ~act->csum_flags; + break; + default: /* Currently we do not handle any other actions. */ return -EOPNOTSUPP; } @@ -789,7 +887,7 @@ int nfp_flower_compile_action(struct nfp_app *app, { int act_len, act_cnt, err, tun_out_cnt, out_cnt, i; enum nfp_flower_tun_type tun_type; - const struct tc_action *a; + struct flow_action_entry *act; u32 csum_updated = 0; memset(nfp_flow->action_data, 0, NFP_FL_MAX_A_SIZ); @@ -800,8 +898,8 @@ int nfp_flower_compile_action(struct nfp_app *app, tun_out_cnt = 0; out_cnt = 0; - tcf_exts_for_each_action(i, a, flow->exts) { - err = nfp_flower_loop_action(app, a, flow, nfp_flow, &act_len, + flow_action_for_each(i, act, &flow->rule->action) { + err = nfp_flower_loop_action(app, act, flow, nfp_flow, &act_len, netdev, &tun_type, &tun_out_cnt, &out_cnt, &csum_updated); if (err) diff --git a/drivers/net/ethernet/netronome/nfp/flower/cmsg.c b/drivers/net/ethernet/netronome/nfp/flower/cmsg.c index 4c5eaf36d5bb..cf9e1118ee8f 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/cmsg.c +++ b/drivers/net/ethernet/netronome/nfp/flower/cmsg.c @@ -45,11 +45,9 @@ nfp_flower_cmsg_mac_repr_start(struct nfp_app *app, unsigned int num_ports) { struct nfp_flower_cmsg_mac_repr *msg; struct sk_buff *skb; - unsigned int size; - size = sizeof(*msg) + num_ports * sizeof(msg->ports[0]); - skb = nfp_flower_cmsg_alloc(app, size, NFP_FLOWER_CMSG_TYPE_MAC_REPR, - GFP_KERNEL); + skb = nfp_flower_cmsg_alloc(app, struct_size(msg, ports, num_ports), + NFP_FLOWER_CMSG_TYPE_MAC_REPR, GFP_KERNEL); if (!skb) return NULL; @@ -203,7 +201,7 @@ nfp_flower_cmsg_portreify_rx(struct nfp_app *app, struct sk_buff *skb) } atomic_inc(&priv->reify_replies); - wake_up_interruptible(&priv->reify_wait_queue); + wake_up(&priv->reify_wait_queue); } static void diff --git a/drivers/net/ethernet/netronome/nfp/flower/cmsg.h b/drivers/net/ethernet/netronome/nfp/flower/cmsg.h index 29d673aa5277..0ed51e79db00 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/cmsg.h +++ b/drivers/net/ethernet/netronome/nfp/flower/cmsg.h @@ -8,6 +8,7 @@ #include <linux/skbuff.h> #include <linux/types.h> #include <net/geneve.h> +#include <net/vxlan.h> #include "../nfp_app.h" #include "../nfpcore/nfp_cpp.h" @@ -25,7 +26,7 @@ #define NFP_FLOWER_LAYER2_GENEVE_OP BIT(6) #define NFP_FLOWER_MASK_VLAN_PRIO GENMASK(15, 13) -#define NFP_FLOWER_MASK_VLAN_CFI BIT(12) +#define NFP_FLOWER_MASK_VLAN_PRESENT BIT(12) #define NFP_FLOWER_MASK_VLAN_VID GENMASK(11, 0) #define NFP_FLOWER_MASK_MPLS_LB GENMASK(31, 12) @@ -65,8 +66,10 @@ #define NFP_FL_ACTION_OPCODE_SET_IPV4_TUNNEL 6 #define NFP_FL_ACTION_OPCODE_SET_ETHERNET 7 #define NFP_FL_ACTION_OPCODE_SET_IPV4_ADDRS 9 +#define NFP_FL_ACTION_OPCODE_SET_IPV4_TTL_TOS 10 #define NFP_FL_ACTION_OPCODE_SET_IPV6_SRC 11 #define NFP_FL_ACTION_OPCODE_SET_IPV6_DST 12 +#define NFP_FL_ACTION_OPCODE_SET_IPV6_TC_HL_FL 13 #define NFP_FL_ACTION_OPCODE_SET_UDP 14 #define NFP_FL_ACTION_OPCODE_SET_TCP 15 #define NFP_FL_ACTION_OPCODE_PRE_LAG 16 @@ -79,9 +82,10 @@ #define NFP_FL_OUT_FLAGS_TYPE_IDX GENMASK(2, 0) #define NFP_FL_PUSH_VLAN_PRIO GENMASK(15, 13) -#define NFP_FL_PUSH_VLAN_CFI BIT(12) #define NFP_FL_PUSH_VLAN_VID GENMASK(11, 0) +#define IPV6_FLOW_LABEL_MASK cpu_to_be32(0x000fffff) + /* LAG ports */ #define NFP_FL_LAG_OUT 0xC0DE0000 @@ -92,6 +96,9 @@ #define NFP_FLOWER_WORKQ_MAX_SKBS 30000 +/* Cmesg reply (empirical) timeout*/ +#define NFP_FL_REPLY_TIMEOUT msecs_to_jiffies(40) + #define nfp_flower_cmsg_warn(app, fmt, args...) \ do { \ if (net_ratelimit()) \ @@ -125,6 +132,26 @@ struct nfp_fl_set_ip4_addrs { __be32 ipv4_dst; }; +struct nfp_fl_set_ip4_ttl_tos { + struct nfp_fl_act_head head; + u8 ipv4_ttl_mask; + u8 ipv4_tos_mask; + u8 ipv4_ttl; + u8 ipv4_tos; + __be16 reserved; +}; + +struct nfp_fl_set_ipv6_tc_hl_fl { + struct nfp_fl_act_head head; + u8 ipv6_tc_mask; + u8 ipv6_hop_limit_mask; + __be16 reserved; + u8 ipv6_tc; + u8 ipv6_hop_limit; + __be32 ipv6_label_mask; + __be32 ipv6_label; +}; + struct nfp_fl_set_ipv6_addr { struct nfp_fl_act_head head; __be16 reserved; @@ -475,6 +502,32 @@ static inline int nfp_flower_cmsg_get_data_len(struct sk_buff *skb) return skb->len - NFP_FLOWER_CMSG_HLEN; } +static inline bool +nfp_fl_netdev_is_tunnel_type(struct net_device *netdev, + enum nfp_flower_tun_type tun_type) +{ + if (netif_is_vxlan(netdev)) + return tun_type == NFP_FL_TUNNEL_VXLAN; + if (netif_is_geneve(netdev)) + return tun_type == NFP_FL_TUNNEL_GENEVE; + + return false; +} + +static inline bool nfp_fl_is_netdev_to_offload(struct net_device *netdev) +{ + if (!netdev->rtnl_link_ops) + return false; + if (!strcmp(netdev->rtnl_link_ops->kind, "openvswitch")) + return true; + if (netif_is_vxlan(netdev)) + return true; + if (netif_is_geneve(netdev)) + return true; + + return false; +} + struct sk_buff * nfp_flower_cmsg_mac_repr_start(struct nfp_app *app, unsigned int num_ports); void diff --git a/drivers/net/ethernet/netronome/nfp/flower/lag_conf.c b/drivers/net/ethernet/netronome/nfp/flower/lag_conf.c index 81dcf5b318ba..5db838f45694 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/lag_conf.c +++ b/drivers/net/ethernet/netronome/nfp/flower/lag_conf.c @@ -472,17 +472,25 @@ nfp_fl_lag_schedule_group_remove(struct nfp_fl_lag *lag, schedule_delayed_work(&lag->work, NFP_FL_LAG_DELAY); } -static int +static void nfp_fl_lag_schedule_group_delete(struct nfp_fl_lag *lag, struct net_device *master) { struct nfp_fl_lag_group *group; + struct nfp_flower_priv *priv; + + priv = container_of(lag, struct nfp_flower_priv, nfp_lag); + + if (!netif_is_bond_master(master)) + return; mutex_lock(&lag->lock); group = nfp_fl_lag_find_group_for_master_with_lag(lag, master); if (!group) { mutex_unlock(&lag->lock); - return -ENOENT; + nfp_warn(priv->app->cpp, "untracked bond got unregistered %s\n", + netdev_name(master)); + return; } group->to_remove = true; @@ -490,7 +498,6 @@ nfp_fl_lag_schedule_group_delete(struct nfp_fl_lag *lag, mutex_unlock(&lag->lock); schedule_delayed_work(&lag->work, NFP_FL_LAG_DELAY); - return 0; } static int @@ -575,7 +582,7 @@ nfp_fl_lag_changeupper_event(struct nfp_fl_lag *lag, return 0; } -static int +static void nfp_fl_lag_changels_event(struct nfp_fl_lag *lag, struct net_device *netdev, struct netdev_notifier_changelowerstate_info *info) { @@ -586,18 +593,18 @@ nfp_fl_lag_changels_event(struct nfp_fl_lag *lag, struct net_device *netdev, unsigned long *flags; if (!netif_is_lag_port(netdev) || !nfp_netdev_is_nfp_repr(netdev)) - return 0; + return; lag_lower_info = info->lower_state_info; if (!lag_lower_info) - return 0; + return; priv = container_of(lag, struct nfp_flower_priv, nfp_lag); repr = netdev_priv(netdev); /* Verify that the repr is associated with this app. */ if (repr->app != priv->app) - return 0; + return; repr_priv = repr->app_priv; flags = &repr_priv->lag_port_flags; @@ -617,20 +624,15 @@ nfp_fl_lag_changels_event(struct nfp_fl_lag *lag, struct net_device *netdev, mutex_unlock(&lag->lock); schedule_delayed_work(&lag->work, NFP_FL_LAG_DELAY); - return 0; } -static int -nfp_fl_lag_netdev_event(struct notifier_block *nb, unsigned long event, - void *ptr) +int nfp_flower_lag_netdev_event(struct nfp_flower_priv *priv, + struct net_device *netdev, + unsigned long event, void *ptr) { - struct net_device *netdev; - struct nfp_fl_lag *lag; + struct nfp_fl_lag *lag = &priv->nfp_lag; int err; - netdev = netdev_notifier_info_to_dev(ptr); - lag = container_of(nb, struct nfp_fl_lag, lag_nb); - switch (event) { case NETDEV_CHANGEUPPER: err = nfp_fl_lag_changeupper_event(lag, ptr); @@ -638,17 +640,11 @@ nfp_fl_lag_netdev_event(struct notifier_block *nb, unsigned long event, return NOTIFY_BAD; return NOTIFY_OK; case NETDEV_CHANGELOWERSTATE: - err = nfp_fl_lag_changels_event(lag, netdev, ptr); - if (err) - return NOTIFY_BAD; + nfp_fl_lag_changels_event(lag, netdev, ptr); return NOTIFY_OK; case NETDEV_UNREGISTER: - if (netif_is_bond_master(netdev)) { - err = nfp_fl_lag_schedule_group_delete(lag, netdev); - if (err) - return NOTIFY_BAD; - return NOTIFY_OK; - } + nfp_fl_lag_schedule_group_delete(lag, netdev); + return NOTIFY_OK; } return NOTIFY_DONE; @@ -673,8 +669,6 @@ void nfp_flower_lag_init(struct nfp_fl_lag *lag) /* 0 is a reserved batch version so increment to first valid value. */ nfp_fl_increment_version(lag); - - lag->lag_nb.notifier_call = nfp_fl_lag_netdev_event; } void nfp_flower_lag_cleanup(struct nfp_fl_lag *lag) diff --git a/drivers/net/ethernet/netronome/nfp/flower/main.c b/drivers/net/ethernet/netronome/nfp/flower/main.c index 3a54728d2ea6..408089133599 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/main.c +++ b/drivers/net/ethernet/netronome/nfp/flower/main.c @@ -32,6 +32,71 @@ static enum devlink_eswitch_mode eswitch_mode_get(struct nfp_app *app) return DEVLINK_ESWITCH_MODE_SWITCHDEV; } +static struct nfp_flower_non_repr_priv * +nfp_flower_non_repr_priv_lookup(struct nfp_app *app, struct net_device *netdev) +{ + struct nfp_flower_priv *priv = app->priv; + struct nfp_flower_non_repr_priv *entry; + + ASSERT_RTNL(); + + list_for_each_entry(entry, &priv->non_repr_priv, list) + if (entry->netdev == netdev) + return entry; + + return NULL; +} + +void +__nfp_flower_non_repr_priv_get(struct nfp_flower_non_repr_priv *non_repr_priv) +{ + non_repr_priv->ref_count++; +} + +struct nfp_flower_non_repr_priv * +nfp_flower_non_repr_priv_get(struct nfp_app *app, struct net_device *netdev) +{ + struct nfp_flower_priv *priv = app->priv; + struct nfp_flower_non_repr_priv *entry; + + entry = nfp_flower_non_repr_priv_lookup(app, netdev); + if (entry) + goto inc_ref; + + entry = kzalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + return NULL; + + entry->netdev = netdev; + list_add(&entry->list, &priv->non_repr_priv); + +inc_ref: + __nfp_flower_non_repr_priv_get(entry); + return entry; +} + +void +__nfp_flower_non_repr_priv_put(struct nfp_flower_non_repr_priv *non_repr_priv) +{ + if (--non_repr_priv->ref_count) + return; + + list_del(&non_repr_priv->list); + kfree(non_repr_priv); +} + +void +nfp_flower_non_repr_priv_put(struct nfp_app *app, struct net_device *netdev) +{ + struct nfp_flower_non_repr_priv *entry; + + entry = nfp_flower_non_repr_priv_lookup(app, netdev); + if (!entry) + return; + + __nfp_flower_non_repr_priv_put(entry); +} + static enum nfp_repr_type nfp_flower_repr_get_type_and_port(struct nfp_app *app, u32 port_id, u8 *port) { @@ -107,16 +172,14 @@ static int nfp_flower_wait_repr_reify(struct nfp_app *app, atomic_t *replies, int tot_repl) { struct nfp_flower_priv *priv = app->priv; - int err; if (!tot_repl) return 0; lockdep_assert_held(&app->pf->lock); - err = wait_event_interruptible_timeout(priv->reify_wait_queue, - atomic_read(replies) >= tot_repl, - msecs_to_jiffies(10)); - if (err <= 0) { + if (!wait_event_timeout(priv->reify_wait_queue, + atomic_read(replies) >= tot_repl, + NFP_FL_REPLY_TIMEOUT)) { nfp_warn(app->cpp, "Not all reprs responded to reify\n"); return -EIO; } @@ -146,23 +209,12 @@ nfp_flower_repr_netdev_stop(struct nfp_app *app, struct nfp_repr *repr) return nfp_flower_cmsg_portmod(repr, false, repr->netdev->mtu, false); } -static int -nfp_flower_repr_netdev_init(struct nfp_app *app, struct net_device *netdev) -{ - return tc_setup_cb_egdev_register(netdev, - nfp_flower_setup_tc_egress_cb, - netdev_priv(netdev)); -} - static void nfp_flower_repr_netdev_clean(struct nfp_app *app, struct net_device *netdev) { struct nfp_repr *repr = netdev_priv(netdev); kfree(repr->app_priv); - - tc_setup_cb_egdev_unregister(netdev, nfp_flower_setup_tc_egress_cb, - netdev_priv(netdev)); } static void @@ -234,6 +286,7 @@ nfp_flower_spawn_vnic_reprs(struct nfp_app *app, nfp_repr = netdev_priv(repr); nfp_repr->app_priv = repr_priv; + repr_priv->nfp_repr = nfp_repr; /* For now we only support 1 PF */ WARN_ON(repr_type == NFP_REPR_TYPE_PF && i); @@ -348,6 +401,7 @@ nfp_flower_spawn_phy_reprs(struct nfp_app *app, struct nfp_flower_priv *priv) nfp_repr = netdev_priv(repr); nfp_repr->app_priv = repr_priv; + repr_priv->nfp_repr = nfp_repr; port = nfp_port_alloc(app, NFP_PORT_PHYS_PORT, repr); if (IS_ERR(port)) { @@ -487,8 +541,8 @@ err_clear_nn: static int nfp_flower_init(struct nfp_app *app) { + u64 version, features, ctx_count, num_mems; const struct nfp_pf *pf = app->pf; - u64 version, features, ctx_count; struct nfp_flower_priv *app_priv; int err; @@ -513,6 +567,23 @@ static int nfp_flower_init(struct nfp_app *app) return err; } + num_mems = nfp_rtsym_read_le(app->pf->rtbl, "CONFIG_FC_HOST_CTX_SPLIT", + &err); + if (err) { + nfp_warn(app->cpp, + "FlowerNIC: unsupported host context memory: %d\n", + err); + err = 0; + num_mems = 1; + } + + if (!FIELD_FIT(NFP_FL_STAT_ID_MU_NUM, num_mems) || !num_mems) { + nfp_warn(app->cpp, + "FlowerNIC: invalid host context memory: %llu\n", + num_mems); + return -EINVAL; + } + ctx_count = nfp_rtsym_read_le(app->pf->rtbl, "CONFIG_FC_HOST_CTX_COUNT", &err); if (err) { @@ -533,6 +604,8 @@ static int nfp_flower_init(struct nfp_app *app) if (!app_priv) return -ENOMEM; + app_priv->total_mem_units = num_mems; + app_priv->active_mem_unit = 0; app_priv->stats_ring_size = roundup_pow_of_two(ctx_count); app->priv = app_priv; app_priv->app = app; @@ -544,7 +617,7 @@ static int nfp_flower_init(struct nfp_app *app) init_waitqueue_head(&app_priv->mtu_conf.wait_q); spin_lock_init(&app_priv->mtu_conf.lock); - err = nfp_flower_metadata_init(app, ctx_count); + err = nfp_flower_metadata_init(app, ctx_count, num_mems); if (err) goto err_free_app_priv; @@ -568,6 +641,9 @@ static int nfp_flower_init(struct nfp_app *app) goto err_cleanup_metadata; } + INIT_LIST_HEAD(&app_priv->indr_block_cb_priv); + INIT_LIST_HEAD(&app_priv->non_repr_priv); + return 0; err_cleanup_metadata: @@ -610,7 +686,7 @@ nfp_flower_repr_change_mtu(struct nfp_app *app, struct net_device *netdev, { struct nfp_flower_priv *app_priv = app->priv; struct nfp_repr *repr = netdev_priv(netdev); - int err, ack; + int err; /* Only need to config FW for physical port MTU change. */ if (repr->port->type != NFP_PORT_PHYS_PORT) @@ -637,11 +713,9 @@ nfp_flower_repr_change_mtu(struct nfp_app *app, struct net_device *netdev, } /* Wait for fw to ack the change. */ - ack = wait_event_timeout(app_priv->mtu_conf.wait_q, - nfp_flower_check_ack(app_priv), - msecs_to_jiffies(10)); - - if (!ack) { + if (!wait_event_timeout(app_priv->mtu_conf.wait_q, + nfp_flower_check_ack(app_priv), + NFP_FL_REPLY_TIMEOUT)) { spin_lock_bh(&app_priv->mtu_conf.lock); app_priv->mtu_conf.requested_val = 0; spin_unlock_bh(&app_priv->mtu_conf.lock); @@ -661,10 +735,6 @@ static int nfp_flower_start(struct nfp_app *app) err = nfp_flower_lag_reset(&app_priv->nfp_lag); if (err) return err; - - err = register_netdevice_notifier(&app_priv->nfp_lag.lag_nb); - if (err) - return err; } return nfp_tunnel_config_start(app); @@ -672,12 +742,27 @@ static int nfp_flower_start(struct nfp_app *app) static void nfp_flower_stop(struct nfp_app *app) { + nfp_tunnel_config_stop(app); +} + +static int +nfp_flower_netdev_event(struct nfp_app *app, struct net_device *netdev, + unsigned long event, void *ptr) +{ struct nfp_flower_priv *app_priv = app->priv; + int ret; - if (app_priv->flower_ext_feats & NFP_FL_FEATS_LAG) - unregister_netdevice_notifier(&app_priv->nfp_lag.lag_nb); + if (app_priv->flower_ext_feats & NFP_FL_FEATS_LAG) { + ret = nfp_flower_lag_netdev_event(app_priv, netdev, event, ptr); + if (ret & NOTIFY_STOP_MASK) + return ret; + } - nfp_tunnel_config_stop(app); + ret = nfp_flower_reg_indir_block_handler(app, netdev, event); + if (ret & NOTIFY_STOP_MASK) + return ret; + + return nfp_tunnel_mac_event_handler(app, netdev, event, ptr); } const struct nfp_app_type app_flower = { @@ -698,7 +783,6 @@ const struct nfp_app_type app_flower = { .vnic_init = nfp_flower_vnic_init, .vnic_clean = nfp_flower_vnic_clean, - .repr_init = nfp_flower_repr_netdev_init, .repr_preclean = nfp_flower_repr_netdev_preclean, .repr_clean = nfp_flower_repr_netdev_clean, @@ -708,6 +792,8 @@ const struct nfp_app_type app_flower = { .start = nfp_flower_start, .stop = nfp_flower_stop, + .netdev_event = nfp_flower_netdev_event, + .ctrl_msg_rx = nfp_flower_cmsg_rx, .sriov_enable = nfp_flower_sriov_enable, diff --git a/drivers/net/ethernet/netronome/nfp/flower/main.h b/drivers/net/ethernet/netronome/nfp/flower/main.h index 90045bab95bf..c0945a5fd1a4 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/main.h +++ b/drivers/net/ethernet/netronome/nfp/flower/main.h @@ -20,7 +20,9 @@ struct nfp_fl_pre_lag; struct net_device; struct nfp_app; -#define NFP_FL_STATS_CTX_DONT_CARE cpu_to_be32(0xffffffff) +#define NFP_FL_STAT_ID_MU_NUM GENMASK(31, 22) +#define NFP_FL_STAT_ID_STAT GENMASK(21, 0) + #define NFP_FL_STATS_ELEM_RS FIELD_SIZEOF(struct nfp_fl_stats_id, \ init_unalloc) #define NFP_FLOWER_MASK_ENTRY_RS 256 @@ -55,6 +57,26 @@ struct nfp_fl_stats_id { }; /** + * struct nfp_fl_tunnel_offloads - priv data for tunnel offloads + * @offloaded_macs: Hashtable of the offloaded MAC addresses + * @ipv4_off_list: List of IPv4 addresses to offload + * @neigh_off_list: List of neighbour offloads + * @ipv4_off_lock: Lock for the IPv4 address list + * @neigh_off_lock: Lock for the neighbour address list + * @mac_off_ids: IDA to manage id assignment for offloaded MACs + * @neigh_nb: Notifier to monitor neighbour state + */ +struct nfp_fl_tunnel_offloads { + struct rhashtable offloaded_macs; + struct list_head ipv4_off_list; + struct list_head neigh_off_list; + struct mutex ipv4_off_lock; + spinlock_t neigh_off_lock; + struct ida mac_off_ids; + struct notifier_block neigh_nb; +}; + +/** * struct nfp_mtu_conf - manage MTU setting * @portnum: NFP port number of repr with requested MTU change * @requested_val: MTU value requested for repr @@ -72,7 +94,6 @@ struct nfp_mtu_conf { /** * struct nfp_fl_lag - Flower APP priv data for link aggregation - * @lag_nb: Notifier to track master/slave events * @work: Work queue for writing configs to the HW * @lock: Lock to protect lag_group_list * @group_list: List of all master/slave groups offloaded @@ -85,7 +106,6 @@ struct nfp_mtu_conf { * retransmission */ struct nfp_fl_lag { - struct notifier_block lag_nb; struct delayed_work work; struct mutex lock; struct list_head group_list; @@ -116,23 +136,16 @@ struct nfp_fl_lag { * processing * @cmsg_skbs_low: List of lower priority skbs for control message * processing - * @nfp_mac_off_list: List of MAC addresses to offload - * @nfp_mac_index_list: List of unique 8-bit indexes for non NFP netdevs - * @nfp_ipv4_off_list: List of IPv4 addresses to offload - * @nfp_neigh_off_list: List of neighbour offloads - * @nfp_mac_off_lock: Lock for the MAC address list - * @nfp_mac_index_lock: Lock for the MAC index list - * @nfp_ipv4_off_lock: Lock for the IPv4 address list - * @nfp_neigh_off_lock: Lock for the neighbour address list - * @nfp_mac_off_ids: IDA to manage id assignment for offloaded macs - * @nfp_mac_off_count: Number of MACs in address list - * @nfp_tun_mac_nb: Notifier to monitor link state - * @nfp_tun_neigh_nb: Notifier to monitor neighbour state + * @tun: Tunnel offload data * @reify_replies: atomically stores the number of replies received * from firmware for repr reify * @reify_wait_queue: wait queue for repr reify response counting * @mtu_conf: Configuration of repr MTU value * @nfp_lag: Link aggregation data block + * @indr_block_cb_priv: List of priv data passed to indirect block cbs + * @non_repr_priv: List of offloaded non-repr ports and their priv data + * @active_mem_unit: Current active memory unit for flower rules + * @total_mem_units: Total number of available memory units for flower rules */ struct nfp_flower_priv { struct nfp_app *app; @@ -150,30 +163,47 @@ struct nfp_flower_priv { struct work_struct cmsg_work; struct sk_buff_head cmsg_skbs_high; struct sk_buff_head cmsg_skbs_low; - struct list_head nfp_mac_off_list; - struct list_head nfp_mac_index_list; - struct list_head nfp_ipv4_off_list; - struct list_head nfp_neigh_off_list; - struct mutex nfp_mac_off_lock; - struct mutex nfp_mac_index_lock; - struct mutex nfp_ipv4_off_lock; - spinlock_t nfp_neigh_off_lock; - struct ida nfp_mac_off_ids; - int nfp_mac_off_count; - struct notifier_block nfp_tun_mac_nb; - struct notifier_block nfp_tun_neigh_nb; + struct nfp_fl_tunnel_offloads tun; atomic_t reify_replies; wait_queue_head_t reify_wait_queue; struct nfp_mtu_conf mtu_conf; struct nfp_fl_lag nfp_lag; + struct list_head indr_block_cb_priv; + struct list_head non_repr_priv; + unsigned int active_mem_unit; + unsigned int total_mem_units; }; /** * struct nfp_flower_repr_priv - Flower APP per-repr priv data + * @nfp_repr: Back pointer to nfp_repr * @lag_port_flags: Extended port flags to record lag state of repr + * @mac_offloaded: Flag indicating a MAC address is offloaded for repr + * @offloaded_mac_addr: MAC address that has been offloaded for repr + * @mac_list: List entry of reprs that share the same offloaded MAC */ struct nfp_flower_repr_priv { + struct nfp_repr *nfp_repr; unsigned long lag_port_flags; + bool mac_offloaded; + u8 offloaded_mac_addr[ETH_ALEN]; + struct list_head mac_list; +}; + +/** + * struct nfp_flower_non_repr_priv - Priv data for non-repr offloaded ports + * @list: List entry of offloaded reprs + * @netdev: Pointer to non-repr net_device + * @ref_count: Number of references held for this priv data + * @mac_offloaded: Flag indicating a MAC address is offloaded for device + * @offloaded_mac_addr: MAC address that has been offloaded for dev + */ +struct nfp_flower_non_repr_priv { + struct list_head list; + struct net_device *netdev; + int ref_count; + bool mac_offloaded; + u8 offloaded_mac_addr[ETH_ALEN]; }; struct nfp_fl_key_ls { @@ -209,7 +239,6 @@ struct nfp_fl_payload { char *unmasked_data; char *mask_data; char *action_data; - bool ingress_offload; }; extern const struct rhashtable_params nfp_flower_table_params; @@ -221,12 +250,14 @@ struct nfp_fl_stats_frame { __be64 stats_cookie; }; -int nfp_flower_metadata_init(struct nfp_app *app, u64 host_ctx_count); +int nfp_flower_metadata_init(struct nfp_app *app, u64 host_ctx_count, + unsigned int host_ctx_split); void nfp_flower_metadata_cleanup(struct nfp_app *app); int nfp_flower_setup_tc(struct nfp_app *app, struct net_device *netdev, enum tc_setup_type type, void *type_data); -int nfp_flower_compile_flow_match(struct tc_cls_flower_offload *flow, +int nfp_flower_compile_flow_match(struct nfp_app *app, + struct tc_cls_flower_offload *flow, struct nfp_fl_key_ls *key_ls, struct net_device *netdev, struct nfp_fl_payload *nfp_flow, @@ -244,7 +275,7 @@ int nfp_modify_flow_metadata(struct nfp_app *app, struct nfp_fl_payload * nfp_flower_search_fl_table(struct nfp_app *app, unsigned long tc_flower_cookie, - struct net_device *netdev, __be32 host_ctx); + struct net_device *netdev); struct nfp_fl_payload * nfp_flower_remove_fl_table(struct nfp_app *app, unsigned long tc_flower_cookie); @@ -252,21 +283,35 @@ void nfp_flower_rx_flow_stats(struct nfp_app *app, struct sk_buff *skb); int nfp_tunnel_config_start(struct nfp_app *app); void nfp_tunnel_config_stop(struct nfp_app *app); -void nfp_tunnel_write_macs(struct nfp_app *app); +int nfp_tunnel_mac_event_handler(struct nfp_app *app, + struct net_device *netdev, + unsigned long event, void *ptr); void nfp_tunnel_del_ipv4_off(struct nfp_app *app, __be32 ipv4); void nfp_tunnel_add_ipv4_off(struct nfp_app *app, __be32 ipv4); void nfp_tunnel_request_route(struct nfp_app *app, struct sk_buff *skb); void nfp_tunnel_keep_alive(struct nfp_app *app, struct sk_buff *skb); -int nfp_flower_setup_tc_egress_cb(enum tc_setup_type type, void *type_data, - void *cb_priv); void nfp_flower_lag_init(struct nfp_fl_lag *lag); void nfp_flower_lag_cleanup(struct nfp_fl_lag *lag); int nfp_flower_lag_reset(struct nfp_fl_lag *lag); +int nfp_flower_lag_netdev_event(struct nfp_flower_priv *priv, + struct net_device *netdev, + unsigned long event, void *ptr); bool nfp_flower_lag_unprocessed_msg(struct nfp_app *app, struct sk_buff *skb); int nfp_flower_lag_populate_pre_action(struct nfp_app *app, struct net_device *master, struct nfp_fl_pre_lag *pre_act); int nfp_flower_lag_get_output_id(struct nfp_app *app, struct net_device *master); +int nfp_flower_reg_indir_block_handler(struct nfp_app *app, + struct net_device *netdev, + unsigned long event); +void +__nfp_flower_non_repr_priv_get(struct nfp_flower_non_repr_priv *non_repr_priv); +struct nfp_flower_non_repr_priv * +nfp_flower_non_repr_priv_get(struct nfp_app *app, struct net_device *netdev); +void +__nfp_flower_non_repr_priv_put(struct nfp_flower_non_repr_priv *non_repr_priv); +void +nfp_flower_non_repr_priv_put(struct nfp_app *app, struct net_device *netdev); #endif diff --git a/drivers/net/ethernet/netronome/nfp/flower/match.c b/drivers/net/ethernet/netronome/nfp/flower/match.c index e54fb6034326..9b8b843d0340 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/match.c +++ b/drivers/net/ethernet/netronome/nfp/flower/match.c @@ -8,32 +8,41 @@ #include "main.h" static void -nfp_flower_compile_meta_tci(struct nfp_flower_meta_tci *frame, - struct tc_cls_flower_offload *flow, u8 key_type, - bool mask_version) +nfp_flower_compile_meta_tci(struct nfp_flower_meta_tci *ext, + struct nfp_flower_meta_tci *msk, + struct tc_cls_flower_offload *flow, u8 key_type) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_vlan *flow_vlan; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); u16 tmp_tci; - memset(frame, 0, sizeof(struct nfp_flower_meta_tci)); + memset(ext, 0, sizeof(struct nfp_flower_meta_tci)); + memset(msk, 0, sizeof(struct nfp_flower_meta_tci)); + /* Populate the metadata frame. */ - frame->nfp_flow_key_layer = key_type; - frame->mask_id = ~0; + ext->nfp_flow_key_layer = key_type; + ext->mask_id = ~0; + + msk->nfp_flow_key_layer = key_type; + msk->mask_id = ~0; - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_VLAN)) { - flow_vlan = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_VLAN, - target); + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_VLAN)) { + struct flow_match_vlan match; + + flow_rule_match_vlan(rule, &match); /* Populate the tci field. */ - if (flow_vlan->vlan_id || flow_vlan->vlan_priority) { - tmp_tci = FIELD_PREP(NFP_FLOWER_MASK_VLAN_PRIO, - flow_vlan->vlan_priority) | - FIELD_PREP(NFP_FLOWER_MASK_VLAN_VID, - flow_vlan->vlan_id) | - NFP_FLOWER_MASK_VLAN_CFI; - frame->tci = cpu_to_be16(tmp_tci); - } + tmp_tci = NFP_FLOWER_MASK_VLAN_PRESENT; + tmp_tci |= FIELD_PREP(NFP_FLOWER_MASK_VLAN_PRIO, + match.key->vlan_priority) | + FIELD_PREP(NFP_FLOWER_MASK_VLAN_VID, + match.key->vlan_id); + ext->tci = cpu_to_be16(tmp_tci); + + tmp_tci = NFP_FLOWER_MASK_VLAN_PRESENT; + tmp_tci |= FIELD_PREP(NFP_FLOWER_MASK_VLAN_PRIO, + match.mask->vlan_priority) | + FIELD_PREP(NFP_FLOWER_MASK_VLAN_VID, + match.mask->vlan_id); + msk->tci = cpu_to_be16(tmp_tci); } } @@ -52,266 +61,288 @@ nfp_flower_compile_port(struct nfp_flower_in_port *frame, u32 cmsg_port, return 0; } - if (tun_type) + if (tun_type) { frame->in_port = cpu_to_be32(NFP_FL_PORT_TYPE_TUN | tun_type); - else + } else { + if (!cmsg_port) + return -EOPNOTSUPP; frame->in_port = cpu_to_be32(cmsg_port); + } return 0; } static void -nfp_flower_compile_mac(struct nfp_flower_mac_mpls *frame, - struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_mac(struct nfp_flower_mac_mpls *ext, + struct nfp_flower_mac_mpls *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_eth_addrs *addr; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); + + memset(ext, 0, sizeof(struct nfp_flower_mac_mpls)); + memset(msk, 0, sizeof(struct nfp_flower_mac_mpls)); - memset(frame, 0, sizeof(struct nfp_flower_mac_mpls)); + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ETH_ADDRS)) { + struct flow_match_eth_addrs match; - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_ETH_ADDRS)) { - addr = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ETH_ADDRS, - target); + flow_rule_match_eth_addrs(rule, &match); /* Populate mac frame. */ - ether_addr_copy(frame->mac_dst, &addr->dst[0]); - ether_addr_copy(frame->mac_src, &addr->src[0]); + ether_addr_copy(ext->mac_dst, &match.key->dst[0]); + ether_addr_copy(ext->mac_src, &match.key->src[0]); + ether_addr_copy(msk->mac_dst, &match.mask->dst[0]); + ether_addr_copy(msk->mac_src, &match.mask->src[0]); } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_MPLS)) { - struct flow_dissector_key_mpls *mpls; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_MPLS)) { + struct flow_match_mpls match; u32 t_mpls; - mpls = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_MPLS, - target); - - t_mpls = FIELD_PREP(NFP_FLOWER_MASK_MPLS_LB, mpls->mpls_label) | - FIELD_PREP(NFP_FLOWER_MASK_MPLS_TC, mpls->mpls_tc) | - FIELD_PREP(NFP_FLOWER_MASK_MPLS_BOS, mpls->mpls_bos) | + flow_rule_match_mpls(rule, &match); + t_mpls = FIELD_PREP(NFP_FLOWER_MASK_MPLS_LB, match.key->mpls_label) | + FIELD_PREP(NFP_FLOWER_MASK_MPLS_TC, match.key->mpls_tc) | + FIELD_PREP(NFP_FLOWER_MASK_MPLS_BOS, match.key->mpls_bos) | NFP_FLOWER_MASK_MPLS_Q; - - frame->mpls_lse = cpu_to_be32(t_mpls); - } else if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_BASIC)) { + ext->mpls_lse = cpu_to_be32(t_mpls); + t_mpls = FIELD_PREP(NFP_FLOWER_MASK_MPLS_LB, match.mask->mpls_label) | + FIELD_PREP(NFP_FLOWER_MASK_MPLS_TC, match.mask->mpls_tc) | + FIELD_PREP(NFP_FLOWER_MASK_MPLS_BOS, match.mask->mpls_bos) | + NFP_FLOWER_MASK_MPLS_Q; + msk->mpls_lse = cpu_to_be32(t_mpls); + } else if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) { /* Check for mpls ether type and set NFP_FLOWER_MASK_MPLS_Q * bit, which indicates an mpls ether type but without any * mpls fields. */ - struct flow_dissector_key_basic *key_basic; - - key_basic = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_BASIC, - flow->key); - if (key_basic->n_proto == cpu_to_be16(ETH_P_MPLS_UC) || - key_basic->n_proto == cpu_to_be16(ETH_P_MPLS_MC)) - frame->mpls_lse = cpu_to_be32(NFP_FLOWER_MASK_MPLS_Q); + struct flow_match_basic match; + + flow_rule_match_basic(rule, &match); + if (match.key->n_proto == cpu_to_be16(ETH_P_MPLS_UC) || + match.key->n_proto == cpu_to_be16(ETH_P_MPLS_MC)) { + ext->mpls_lse = cpu_to_be32(NFP_FLOWER_MASK_MPLS_Q); + msk->mpls_lse = cpu_to_be32(NFP_FLOWER_MASK_MPLS_Q); + } } } static void -nfp_flower_compile_tport(struct nfp_flower_tp_ports *frame, - struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_tport(struct nfp_flower_tp_ports *ext, + struct nfp_flower_tp_ports *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_ports *tp; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); + + memset(ext, 0, sizeof(struct nfp_flower_tp_ports)); + memset(msk, 0, sizeof(struct nfp_flower_tp_ports)); - memset(frame, 0, sizeof(struct nfp_flower_tp_ports)); + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_PORTS)) { + struct flow_match_ports match; - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_PORTS)) { - tp = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_PORTS, - target); - frame->port_src = tp->src; - frame->port_dst = tp->dst; + flow_rule_match_ports(rule, &match); + ext->port_src = match.key->src; + ext->port_dst = match.key->dst; + msk->port_src = match.mask->src; + msk->port_dst = match.mask->dst; } } static void -nfp_flower_compile_ip_ext(struct nfp_flower_ip_ext *frame, - struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_ip_ext(struct nfp_flower_ip_ext *ext, + struct nfp_flower_ip_ext *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_BASIC)) { - struct flow_dissector_key_basic *basic; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) { + struct flow_match_basic match; - basic = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_BASIC, - target); - frame->proto = basic->ip_proto; + flow_rule_match_basic(rule, &match); + ext->proto = match.key->ip_proto; + msk->proto = match.mask->ip_proto; } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_IP)) { - struct flow_dissector_key_ip *flow_ip; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IP)) { + struct flow_match_ip match; - flow_ip = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_IP, - target); - frame->tos = flow_ip->tos; - frame->ttl = flow_ip->ttl; + flow_rule_match_ip(rule, &match); + ext->tos = match.key->tos; + ext->ttl = match.key->ttl; + msk->tos = match.mask->tos; + msk->ttl = match.mask->ttl; } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_TCP)) { - struct flow_dissector_key_tcp *tcp; - u32 tcp_flags; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_TCP)) { + u16 tcp_flags, tcp_flags_mask; + struct flow_match_tcp match; - tcp = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_TCP, target); - tcp_flags = be16_to_cpu(tcp->flags); + flow_rule_match_tcp(rule, &match); + tcp_flags = be16_to_cpu(match.key->flags); + tcp_flags_mask = be16_to_cpu(match.mask->flags); if (tcp_flags & TCPHDR_FIN) - frame->flags |= NFP_FL_TCP_FLAG_FIN; + ext->flags |= NFP_FL_TCP_FLAG_FIN; + if (tcp_flags_mask & TCPHDR_FIN) + msk->flags |= NFP_FL_TCP_FLAG_FIN; + if (tcp_flags & TCPHDR_SYN) - frame->flags |= NFP_FL_TCP_FLAG_SYN; + ext->flags |= NFP_FL_TCP_FLAG_SYN; + if (tcp_flags_mask & TCPHDR_SYN) + msk->flags |= NFP_FL_TCP_FLAG_SYN; + if (tcp_flags & TCPHDR_RST) - frame->flags |= NFP_FL_TCP_FLAG_RST; + ext->flags |= NFP_FL_TCP_FLAG_RST; + if (tcp_flags_mask & TCPHDR_RST) + msk->flags |= NFP_FL_TCP_FLAG_RST; + if (tcp_flags & TCPHDR_PSH) - frame->flags |= NFP_FL_TCP_FLAG_PSH; + ext->flags |= NFP_FL_TCP_FLAG_PSH; + if (tcp_flags_mask & TCPHDR_PSH) + msk->flags |= NFP_FL_TCP_FLAG_PSH; + if (tcp_flags & TCPHDR_URG) - frame->flags |= NFP_FL_TCP_FLAG_URG; + ext->flags |= NFP_FL_TCP_FLAG_URG; + if (tcp_flags_mask & TCPHDR_URG) + msk->flags |= NFP_FL_TCP_FLAG_URG; } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_CONTROL)) { - struct flow_dissector_key_control *key; - - key = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_CONTROL, - target); - if (key->flags & FLOW_DIS_IS_FRAGMENT) - frame->flags |= NFP_FL_IP_FRAGMENTED; - if (key->flags & FLOW_DIS_FIRST_FRAG) - frame->flags |= NFP_FL_IP_FRAG_FIRST; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_CONTROL)) { + struct flow_match_control match; + + flow_rule_match_control(rule, &match); + if (match.key->flags & FLOW_DIS_IS_FRAGMENT) + ext->flags |= NFP_FL_IP_FRAGMENTED; + if (match.mask->flags & FLOW_DIS_IS_FRAGMENT) + msk->flags |= NFP_FL_IP_FRAGMENTED; + if (match.key->flags & FLOW_DIS_FIRST_FRAG) + ext->flags |= NFP_FL_IP_FRAG_FIRST; + if (match.mask->flags & FLOW_DIS_FIRST_FRAG) + msk->flags |= NFP_FL_IP_FRAG_FIRST; } } static void -nfp_flower_compile_ipv4(struct nfp_flower_ipv4 *frame, - struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_ipv4(struct nfp_flower_ipv4 *ext, + struct nfp_flower_ipv4 *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_ipv4_addrs *addr; - - memset(frame, 0, sizeof(struct nfp_flower_ipv4)); - - if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_IPV4_ADDRS)) { - addr = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_IPV4_ADDRS, - target); - frame->ipv4_src = addr->src; - frame->ipv4_dst = addr->dst; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); + struct flow_match_ipv4_addrs match; + + memset(ext, 0, sizeof(struct nfp_flower_ipv4)); + memset(msk, 0, sizeof(struct nfp_flower_ipv4)); + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IPV4_ADDRS)) { + flow_rule_match_ipv4_addrs(rule, &match); + ext->ipv4_src = match.key->src; + ext->ipv4_dst = match.key->dst; + msk->ipv4_src = match.mask->src; + msk->ipv4_dst = match.mask->dst; } - nfp_flower_compile_ip_ext(&frame->ip_ext, flow, mask_version); + nfp_flower_compile_ip_ext(&ext->ip_ext, &msk->ip_ext, flow); } static void -nfp_flower_compile_ipv6(struct nfp_flower_ipv6 *frame, - struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_ipv6(struct nfp_flower_ipv6 *ext, + struct nfp_flower_ipv6 *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_ipv6_addrs *addr; - - memset(frame, 0, sizeof(struct nfp_flower_ipv6)); - - if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_IPV6_ADDRS)) { - addr = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_IPV6_ADDRS, - target); - frame->ipv6_src = addr->src; - frame->ipv6_dst = addr->dst; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); + + memset(ext, 0, sizeof(struct nfp_flower_ipv6)); + memset(msk, 0, sizeof(struct nfp_flower_ipv6)); + + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IPV6_ADDRS)) { + struct flow_match_ipv6_addrs match; + + flow_rule_match_ipv6_addrs(rule, &match); + ext->ipv6_src = match.key->src; + ext->ipv6_dst = match.key->dst; + msk->ipv6_src = match.mask->src; + msk->ipv6_dst = match.mask->dst; } - nfp_flower_compile_ip_ext(&frame->ip_ext, flow, mask_version); + nfp_flower_compile_ip_ext(&ext->ip_ext, &msk->ip_ext, flow); } static int -nfp_flower_compile_geneve_opt(void *key_buf, struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_geneve_opt(void *ext, void *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_enc_opts *opts; + struct flow_match_enc_opts match; - opts = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_OPTS, - target); - memcpy(key_buf, opts->data, opts->len); + flow_rule_match_enc_opts(flow->rule, &match); + memcpy(ext, match.key->data, match.key->len); + memcpy(msk, match.mask->data, match.mask->len); return 0; } static void -nfp_flower_compile_ipv4_udp_tun(struct nfp_flower_ipv4_udp_tun *frame, - struct tc_cls_flower_offload *flow, - bool mask_version) +nfp_flower_compile_ipv4_udp_tun(struct nfp_flower_ipv4_udp_tun *ext, + struct nfp_flower_ipv4_udp_tun *msk, + struct tc_cls_flower_offload *flow) { - struct fl_flow_key *target = mask_version ? flow->mask : flow->key; - struct flow_dissector_key_ipv4_addrs *tun_ips; - struct flow_dissector_key_keyid *vni; - struct flow_dissector_key_ip *ip; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); - memset(frame, 0, sizeof(struct nfp_flower_ipv4_udp_tun)); + memset(ext, 0, sizeof(struct nfp_flower_ipv4_udp_tun)); + memset(msk, 0, sizeof(struct nfp_flower_ipv4_udp_tun)); - if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_KEYID)) { + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_KEYID)) { + struct flow_match_enc_keyid match; u32 temp_vni; - vni = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_KEYID, - target); - temp_vni = be32_to_cpu(vni->keyid) << NFP_FL_TUN_VNI_OFFSET; - frame->tun_id = cpu_to_be32(temp_vni); + flow_rule_match_enc_keyid(rule, &match); + temp_vni = be32_to_cpu(match.key->keyid) << NFP_FL_TUN_VNI_OFFSET; + ext->tun_id = cpu_to_be32(temp_vni); + temp_vni = be32_to_cpu(match.mask->keyid) << NFP_FL_TUN_VNI_OFFSET; + msk->tun_id = cpu_to_be32(temp_vni); } - if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_IPV4_ADDRS)) { - tun_ips = - skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_IPV4_ADDRS, - target); - frame->ip_src = tun_ips->src; - frame->ip_dst = tun_ips->dst; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_IPV4_ADDRS)) { + struct flow_match_ipv4_addrs match; + + flow_rule_match_enc_ipv4_addrs(rule, &match); + ext->ip_src = match.key->src; + ext->ip_dst = match.key->dst; + msk->ip_src = match.mask->src; + msk->ip_dst = match.mask->dst; } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_ENC_IP)) { - ip = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_IP, - target); - frame->tos = ip->tos; - frame->ttl = ip->ttl; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_IP)) { + struct flow_match_ip match; + + flow_rule_match_enc_ip(rule, &match); + ext->tos = match.key->tos; + ext->ttl = match.key->ttl; + msk->tos = match.mask->tos; + msk->ttl = match.mask->ttl; } } -int nfp_flower_compile_flow_match(struct tc_cls_flower_offload *flow, +int nfp_flower_compile_flow_match(struct nfp_app *app, + struct tc_cls_flower_offload *flow, struct nfp_fl_key_ls *key_ls, struct net_device *netdev, struct nfp_fl_payload *nfp_flow, enum nfp_flower_tun_type tun_type) { - struct nfp_repr *netdev_repr; + u32 cmsg_port = 0; int err; u8 *ext; u8 *msk; + if (nfp_netdev_is_nfp_repr(netdev)) + cmsg_port = nfp_repr_get_port_id(netdev); + memset(nfp_flow->unmasked_data, 0, key_ls->key_size); memset(nfp_flow->mask_data, 0, key_ls->key_size); ext = nfp_flow->unmasked_data; msk = nfp_flow->mask_data; - /* Populate Exact Metadata. */ nfp_flower_compile_meta_tci((struct nfp_flower_meta_tci *)ext, - flow, key_ls->key_layer, false); - /* Populate Mask Metadata. */ - nfp_flower_compile_meta_tci((struct nfp_flower_meta_tci *)msk, - flow, key_ls->key_layer, true); + (struct nfp_flower_meta_tci *)msk, + flow, key_ls->key_layer); ext += sizeof(struct nfp_flower_meta_tci); msk += sizeof(struct nfp_flower_meta_tci); @@ -327,15 +358,13 @@ int nfp_flower_compile_flow_match(struct tc_cls_flower_offload *flow, /* Populate Exact Port data. */ err = nfp_flower_compile_port((struct nfp_flower_in_port *)ext, - nfp_repr_get_port_id(netdev), - false, tun_type); + cmsg_port, false, tun_type); if (err) return err; /* Populate Mask Port Data. */ err = nfp_flower_compile_port((struct nfp_flower_in_port *)msk, - nfp_repr_get_port_id(netdev), - true, tun_type); + cmsg_port, true, tun_type); if (err) return err; @@ -343,45 +372,33 @@ int nfp_flower_compile_flow_match(struct tc_cls_flower_offload *flow, msk += sizeof(struct nfp_flower_in_port); if (NFP_FLOWER_LAYER_MAC & key_ls->key_layer) { - /* Populate Exact MAC Data. */ nfp_flower_compile_mac((struct nfp_flower_mac_mpls *)ext, - flow, false); - /* Populate Mask MAC Data. */ - nfp_flower_compile_mac((struct nfp_flower_mac_mpls *)msk, - flow, true); + (struct nfp_flower_mac_mpls *)msk, + flow); ext += sizeof(struct nfp_flower_mac_mpls); msk += sizeof(struct nfp_flower_mac_mpls); } if (NFP_FLOWER_LAYER_TP & key_ls->key_layer) { - /* Populate Exact TP Data. */ nfp_flower_compile_tport((struct nfp_flower_tp_ports *)ext, - flow, false); - /* Populate Mask TP Data. */ - nfp_flower_compile_tport((struct nfp_flower_tp_ports *)msk, - flow, true); + (struct nfp_flower_tp_ports *)msk, + flow); ext += sizeof(struct nfp_flower_tp_ports); msk += sizeof(struct nfp_flower_tp_ports); } if (NFP_FLOWER_LAYER_IPV4 & key_ls->key_layer) { - /* Populate Exact IPv4 Data. */ nfp_flower_compile_ipv4((struct nfp_flower_ipv4 *)ext, - flow, false); - /* Populate Mask IPv4 Data. */ - nfp_flower_compile_ipv4((struct nfp_flower_ipv4 *)msk, - flow, true); + (struct nfp_flower_ipv4 *)msk, + flow); ext += sizeof(struct nfp_flower_ipv4); msk += sizeof(struct nfp_flower_ipv4); } if (NFP_FLOWER_LAYER_IPV6 & key_ls->key_layer) { - /* Populate Exact IPv4 Data. */ nfp_flower_compile_ipv6((struct nfp_flower_ipv6 *)ext, - flow, false); - /* Populate Mask IPv4 Data. */ - nfp_flower_compile_ipv6((struct nfp_flower_ipv6 *)msk, - flow, true); + (struct nfp_flower_ipv6 *)msk, + flow); ext += sizeof(struct nfp_flower_ipv6); msk += sizeof(struct nfp_flower_ipv6); } @@ -390,32 +407,19 @@ int nfp_flower_compile_flow_match(struct tc_cls_flower_offload *flow, key_ls->key_layer_two & NFP_FLOWER_LAYER2_GENEVE) { __be32 tun_dst; - /* Populate Exact VXLAN Data. */ - nfp_flower_compile_ipv4_udp_tun((void *)ext, flow, false); - /* Populate Mask VXLAN Data. */ - nfp_flower_compile_ipv4_udp_tun((void *)msk, flow, true); + nfp_flower_compile_ipv4_udp_tun((void *)ext, (void *)msk, flow); tun_dst = ((struct nfp_flower_ipv4_udp_tun *)ext)->ip_dst; ext += sizeof(struct nfp_flower_ipv4_udp_tun); msk += sizeof(struct nfp_flower_ipv4_udp_tun); - /* Configure tunnel end point MAC. */ - if (nfp_netdev_is_nfp_repr(netdev)) { - netdev_repr = netdev_priv(netdev); - nfp_tunnel_write_macs(netdev_repr->app); - - /* Store the tunnel destination in the rule data. - * This must be present and be an exact match. - */ - nfp_flow->nfp_tun_ipv4_addr = tun_dst; - nfp_tunnel_add_ipv4_off(netdev_repr->app, tun_dst); - } + /* Store the tunnel destination in the rule data. + * This must be present and be an exact match. + */ + nfp_flow->nfp_tun_ipv4_addr = tun_dst; + nfp_tunnel_add_ipv4_off(app, tun_dst); if (key_ls->key_layer_two & NFP_FLOWER_LAYER2_GENEVE_OP) { - err = nfp_flower_compile_geneve_opt(ext, flow, false); - if (err) - return err; - - err = nfp_flower_compile_geneve_opt(msk, flow, true); + err = nfp_flower_compile_geneve_opt(ext, msk, flow); if (err) return err; } diff --git a/drivers/net/ethernet/netronome/nfp/flower/metadata.c b/drivers/net/ethernet/netronome/nfp/flower/metadata.c index 48729bf171e0..492837b852b6 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/metadata.c +++ b/drivers/net/ethernet/netronome/nfp/flower/metadata.c @@ -4,6 +4,7 @@ #include <linux/hash.h> #include <linux/hashtable.h> #include <linux/jhash.h> +#include <linux/math64.h> #include <linux/vmalloc.h> #include <net/pkt_cls.h> @@ -21,7 +22,6 @@ struct nfp_mask_id_table { struct nfp_fl_flow_table_cmp_arg { struct net_device *netdev; unsigned long cookie; - __be32 host_ctx; }; static int nfp_release_stats_entry(struct nfp_app *app, u32 stats_context_id) @@ -53,8 +53,17 @@ static int nfp_get_stats_entry(struct nfp_app *app, u32 *stats_context_id) freed_stats_id = priv->stats_ring_size; /* Check for unallocated entries first. */ if (priv->stats_ids.init_unalloc > 0) { - *stats_context_id = priv->stats_ids.init_unalloc - 1; - priv->stats_ids.init_unalloc--; + if (priv->active_mem_unit == priv->total_mem_units) { + priv->stats_ids.init_unalloc--; + priv->active_mem_unit = 0; + } + + *stats_context_id = + FIELD_PREP(NFP_FL_STAT_ID_STAT, + priv->stats_ids.init_unalloc - 1) | + FIELD_PREP(NFP_FL_STAT_ID_MU_NUM, + priv->active_mem_unit); + priv->active_mem_unit++; return 0; } @@ -76,14 +85,13 @@ static int nfp_get_stats_entry(struct nfp_app *app, u32 *stats_context_id) /* Must be called with either RTNL or rcu_read_lock */ struct nfp_fl_payload * nfp_flower_search_fl_table(struct nfp_app *app, unsigned long tc_flower_cookie, - struct net_device *netdev, __be32 host_ctx) + struct net_device *netdev) { struct nfp_fl_flow_table_cmp_arg flower_cmp_arg; struct nfp_flower_priv *priv = app->priv; flower_cmp_arg.netdev = netdev; flower_cmp_arg.cookie = tc_flower_cookie; - flower_cmp_arg.host_ctx = host_ctx; return rhashtable_lookup_fast(&priv->flow_table, &flower_cmp_arg, nfp_flower_table_params); @@ -287,6 +295,7 @@ int nfp_compile_flow_metadata(struct nfp_app *app, nfp_flow->meta.host_ctx_id = cpu_to_be32(stats_cxt); nfp_flow->meta.host_cookie = cpu_to_be64(flow->cookie); + nfp_flow->ingress_dev = netdev; new_mask_id = 0; if (!nfp_check_mask_add(app, nfp_flow->mask_data, @@ -306,8 +315,7 @@ int nfp_compile_flow_metadata(struct nfp_app *app, priv->stats[stats_cxt].bytes = 0; priv->stats[stats_cxt].used = jiffies; - check_entry = nfp_flower_search_fl_table(app, flow->cookie, netdev, - NFP_FL_STATS_CTX_DONT_CARE); + check_entry = nfp_flower_search_fl_table(app, flow->cookie, netdev); if (check_entry) { if (nfp_release_stats_entry(app, stats_cxt)) return -EINVAL; @@ -352,9 +360,7 @@ static int nfp_fl_obj_cmpfn(struct rhashtable_compare_arg *arg, const struct nfp_fl_flow_table_cmp_arg *cmp_arg = arg->key; const struct nfp_fl_payload *flow_entry = obj; - if ((!cmp_arg->netdev || flow_entry->ingress_dev == cmp_arg->netdev) && - (cmp_arg->host_ctx == NFP_FL_STATS_CTX_DONT_CARE || - flow_entry->meta.host_ctx_id == cmp_arg->host_ctx)) + if (flow_entry->ingress_dev == cmp_arg->netdev) return flow_entry->tc_flower_cookie != cmp_arg->cookie; return 1; @@ -385,10 +391,11 @@ const struct rhashtable_params nfp_flower_table_params = { .automatic_shrinking = true, }; -int nfp_flower_metadata_init(struct nfp_app *app, u64 host_ctx_count) +int nfp_flower_metadata_init(struct nfp_app *app, u64 host_ctx_count, + unsigned int host_num_mems) { struct nfp_flower_priv *priv = app->priv; - int err; + int err, stats_size; hash_init(priv->mask_table); @@ -421,10 +428,12 @@ int nfp_flower_metadata_init(struct nfp_app *app, u64 host_ctx_count) if (!priv->stats_ids.free_list.buf) goto err_free_last_used; - priv->stats_ids.init_unalloc = host_ctx_count; + priv->stats_ids.init_unalloc = div_u64(host_ctx_count, host_num_mems); - priv->stats = kvmalloc_array(priv->stats_ring_size, - sizeof(struct nfp_fl_stats), GFP_KERNEL); + stats_size = FIELD_PREP(NFP_FL_STAT_ID_STAT, host_ctx_count) | + FIELD_PREP(NFP_FL_STAT_ID_MU_NUM, host_num_mems - 1); + priv->stats = kvmalloc_array(stats_size, sizeof(struct nfp_fl_stats), + GFP_KERNEL); if (!priv->stats) goto err_free_ring_buf; diff --git a/drivers/net/ethernet/netronome/nfp/flower/offload.c b/drivers/net/ethernet/netronome/nfp/flower/offload.c index 67e576fe7fc0..450d7296fd57 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/offload.c +++ b/drivers/net/ethernet/netronome/nfp/flower/offload.c @@ -56,11 +56,10 @@ BIT(FLOW_DISSECTOR_KEY_ENC_PORTS)) static int -nfp_flower_xmit_flow(struct net_device *netdev, - struct nfp_fl_payload *nfp_flow, u8 mtype) +nfp_flower_xmit_flow(struct nfp_app *app, struct nfp_fl_payload *nfp_flow, + u8 mtype) { u32 meta_len, key_len, mask_len, act_len, tot_len; - struct nfp_repr *priv = netdev_priv(netdev); struct sk_buff *skb; unsigned char *msg; @@ -78,7 +77,7 @@ nfp_flower_xmit_flow(struct net_device *netdev, nfp_flow->meta.mask_len >>= NFP_FL_LW_SIZ; nfp_flow->meta.act_len >>= NFP_FL_LW_SIZ; - skb = nfp_flower_cmsg_alloc(priv->app, tot_len, mtype, GFP_KERNEL); + skb = nfp_flower_cmsg_alloc(app, tot_len, mtype, GFP_KERNEL); if (!skb) return -ENOMEM; @@ -96,30 +95,29 @@ nfp_flower_xmit_flow(struct net_device *netdev, nfp_flow->meta.mask_len <<= NFP_FL_LW_SIZ; nfp_flow->meta.act_len <<= NFP_FL_LW_SIZ; - nfp_ctrl_tx(priv->app->ctrl, skb); + nfp_ctrl_tx(app->ctrl, skb); return 0; } static bool nfp_flower_check_higher_than_mac(struct tc_cls_flower_offload *f) { - return dissector_uses_key(f->dissector, - FLOW_DISSECTOR_KEY_IPV4_ADDRS) || - dissector_uses_key(f->dissector, - FLOW_DISSECTOR_KEY_IPV6_ADDRS) || - dissector_uses_key(f->dissector, - FLOW_DISSECTOR_KEY_PORTS) || - dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_ICMP); + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(f); + + return flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IPV4_ADDRS) || + flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IPV6_ADDRS) || + flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_PORTS) || + flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ICMP); } static int -nfp_flower_calc_opt_layer(struct flow_dissector_key_enc_opts *enc_opts, +nfp_flower_calc_opt_layer(struct flow_match_enc_opts *enc_opts, u32 *key_layer_two, int *key_size) { - if (enc_opts->len > NFP_FL_MAX_GENEVE_OPT_KEY) + if (enc_opts->key->len > NFP_FL_MAX_GENEVE_OPT_KEY) return -EOPNOTSUPP; - if (enc_opts->len > 0) { + if (enc_opts->key->len > 0) { *key_layer_two |= NFP_FLOWER_LAYER2_GENEVE_OP; *key_size += sizeof(struct nfp_flower_geneve_options); } @@ -129,25 +127,26 @@ nfp_flower_calc_opt_layer(struct flow_dissector_key_enc_opts *enc_opts, static int nfp_flower_calculate_key_layers(struct nfp_app *app, + struct net_device *netdev, struct nfp_fl_key_ls *ret_key_ls, struct tc_cls_flower_offload *flow, - bool egress, enum nfp_flower_tun_type *tun_type) { - struct flow_dissector_key_basic *mask_basic = NULL; - struct flow_dissector_key_basic *key_basic = NULL; + struct flow_rule *rule = tc_cls_flower_offload_flow_rule(flow); + struct flow_dissector *dissector = rule->match.dissector; + struct flow_match_basic basic = { NULL, NULL}; struct nfp_flower_priv *priv = app->priv; u32 key_layer_two; u8 key_layer; int key_size; int err; - if (flow->dissector->used_keys & ~NFP_FLOWER_WHITELIST_DISSECTOR) + if (dissector->used_keys & ~NFP_FLOWER_WHITELIST_DISSECTOR) return -EOPNOTSUPP; /* If any tun dissector is used then the required set must be used. */ - if (flow->dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR && - (flow->dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR_R) + if (dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR && + (dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR_R) != NFP_FLOWER_WHITELIST_TUN_DISSECTOR_R) return -EOPNOTSUPP; @@ -156,78 +155,52 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, key_size = sizeof(struct nfp_flower_meta_tci) + sizeof(struct nfp_flower_in_port); - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_ETH_ADDRS) || - dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_MPLS)) { + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ETH_ADDRS) || + flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_MPLS)) { key_layer |= NFP_FLOWER_LAYER_MAC; key_size += sizeof(struct nfp_flower_mac_mpls); } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_VLAN)) { - struct flow_dissector_key_vlan *flow_vlan; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_VLAN)) { + struct flow_match_vlan vlan; - flow_vlan = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_VLAN, - flow->mask); + flow_rule_match_vlan(rule, &vlan); if (!(priv->flower_ext_feats & NFP_FL_FEATS_VLAN_PCP) && - flow_vlan->vlan_priority) + vlan.key->vlan_priority) return -EOPNOTSUPP; } - if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_CONTROL)) { - struct flow_dissector_key_ipv4_addrs *mask_ipv4 = NULL; - struct flow_dissector_key_ports *mask_enc_ports = NULL; - struct flow_dissector_key_enc_opts *enc_op = NULL; - struct flow_dissector_key_ports *enc_ports = NULL; - struct flow_dissector_key_control *mask_enc_ctl = - skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_CONTROL, - flow->mask); - struct flow_dissector_key_control *enc_ctl = - skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_CONTROL, - flow->key); - if (!egress) - return -EOPNOTSUPP; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_CONTROL)) { + struct flow_match_enc_opts enc_op = { NULL, NULL }; + struct flow_match_ipv4_addrs ipv4_addrs; + struct flow_match_control enc_ctl; + struct flow_match_ports enc_ports; + + flow_rule_match_enc_control(rule, &enc_ctl); - if (mask_enc_ctl->addr_type != 0xffff || - enc_ctl->addr_type != FLOW_DISSECTOR_KEY_IPV4_ADDRS) + if (enc_ctl.mask->addr_type != 0xffff || + enc_ctl.key->addr_type != FLOW_DISSECTOR_KEY_IPV4_ADDRS) return -EOPNOTSUPP; /* These fields are already verified as used. */ - mask_ipv4 = - skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_IPV4_ADDRS, - flow->mask); - if (mask_ipv4->dst != cpu_to_be32(~0)) + flow_rule_match_enc_ipv4_addrs(rule, &ipv4_addrs); + if (ipv4_addrs.mask->dst != cpu_to_be32(~0)) return -EOPNOTSUPP; - mask_enc_ports = - skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_PORTS, - flow->mask); - enc_ports = - skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_PORTS, - flow->key); - - if (mask_enc_ports->dst != cpu_to_be16(~0)) + flow_rule_match_enc_ports(rule, &enc_ports); + if (enc_ports.mask->dst != cpu_to_be16(~0)) return -EOPNOTSUPP; - if (dissector_uses_key(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_OPTS)) { - enc_op = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_ENC_OPTS, - flow->key); - } + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_OPTS)) + flow_rule_match_enc_opts(rule, &enc_op); - switch (enc_ports->dst) { + switch (enc_ports.key->dst) { case htons(NFP_FL_VXLAN_PORT): *tun_type = NFP_FL_TUNNEL_VXLAN; key_layer |= NFP_FLOWER_LAYER_VXLAN; key_size += sizeof(struct nfp_flower_ipv4_udp_tun); - if (enc_op) + if (enc_op.key) return -EOPNOTSUPP; break; case htons(NFP_FL_GENEVE_PORT): @@ -239,11 +212,11 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, key_layer_two |= NFP_FLOWER_LAYER2_GENEVE; key_size += sizeof(struct nfp_flower_ipv4_udp_tun); - if (!enc_op) + if (!enc_op.key) break; if (!(priv->flower_ext_feats & NFP_FL_FEATS_GENEVE_OPT)) return -EOPNOTSUPP; - err = nfp_flower_calc_opt_layer(enc_op, &key_layer_two, + err = nfp_flower_calc_opt_layer(&enc_op, &key_layer_two, &key_size); if (err) return err; @@ -251,24 +224,18 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, default: return -EOPNOTSUPP; } - } else if (egress) { - /* Reject non tunnel matches offloaded to egress repr. */ - return -EOPNOTSUPP; - } - - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_BASIC)) { - mask_basic = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_BASIC, - flow->mask); - key_basic = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_BASIC, - flow->key); + /* Ensure the ingress netdev matches the expected tun type. */ + if (!nfp_fl_netdev_is_tunnel_type(netdev, *tun_type)) + return -EOPNOTSUPP; } - if (mask_basic && mask_basic->n_proto) { + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) + flow_rule_match_basic(rule, &basic); + + if (basic.mask && basic.mask->n_proto) { /* Ethernet type is present in the key. */ - switch (key_basic->n_proto) { + switch (basic.key->n_proto) { case cpu_to_be16(ETH_P_IP): key_layer |= NFP_FLOWER_LAYER_IPV4; key_size += sizeof(struct nfp_flower_ipv4); @@ -307,9 +274,9 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, } } - if (mask_basic && mask_basic->ip_proto) { + if (basic.mask && basic.mask->ip_proto) { /* Ethernet type is present in the key. */ - switch (key_basic->ip_proto) { + switch (basic.key->ip_proto) { case IPPROTO_TCP: case IPPROTO_UDP: case IPPROTO_SCTP: @@ -326,14 +293,12 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, } } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_TCP)) { - struct flow_dissector_key_tcp *tcp; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_TCP)) { + struct flow_match_tcp tcp; u32 tcp_flags; - tcp = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_TCP, - flow->key); - tcp_flags = be16_to_cpu(tcp->flags); + flow_rule_match_tcp(rule, &tcp); + tcp_flags = be16_to_cpu(tcp.key->flags); if (tcp_flags & ~NFP_FLOWER_SUPPORTED_TCPFLAGS) return -EOPNOTSUPP; @@ -349,12 +314,12 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, * space, thus we need to ensure we include a IPv4/IPv6 key * layer if we have not done so already. */ - if (!key_basic) + if (!basic.key) return -EOPNOTSUPP; if (!(key_layer & NFP_FLOWER_LAYER_IPV4) && !(key_layer & NFP_FLOWER_LAYER_IPV6)) { - switch (key_basic->n_proto) { + switch (basic.key->n_proto) { case cpu_to_be16(ETH_P_IP): key_layer |= NFP_FLOWER_LAYER_IPV4; key_size += sizeof(struct nfp_flower_ipv4); @@ -371,14 +336,11 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, } } - if (dissector_uses_key(flow->dissector, FLOW_DISSECTOR_KEY_CONTROL)) { - struct flow_dissector_key_control *key_ctl; - - key_ctl = skb_flow_dissector_target(flow->dissector, - FLOW_DISSECTOR_KEY_CONTROL, - flow->key); + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_CONTROL)) { + struct flow_match_control ctl; - if (key_ctl->flags & ~NFP_FLOWER_SUPPORTED_CTLFLAGS) + flow_rule_match_control(rule, &ctl); + if (ctl.key->flags & ~NFP_FLOWER_SUPPORTED_CTLFLAGS) return -EOPNOTSUPP; } @@ -390,7 +352,7 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, } static struct nfp_fl_payload * -nfp_flower_allocate_new(struct nfp_fl_key_ls *key_layer, bool egress) +nfp_flower_allocate_new(struct nfp_fl_key_ls *key_layer) { struct nfp_fl_payload *flow_pay; @@ -414,7 +376,6 @@ nfp_flower_allocate_new(struct nfp_fl_key_ls *key_layer, bool egress) flow_pay->nfp_tun_ipv4_addr = 0; flow_pay->meta.flags = 0; - flow_pay->ingress_offload = !egress; return flow_pay; @@ -432,7 +393,6 @@ err_free_flow: * @app: Pointer to the APP handle * @netdev: netdev structure. * @flow: TC flower classifier offload structure. - * @egress: NFP netdev is the egress. * * Adds a new flow to the repeated hash structure and action payload. * @@ -440,46 +400,35 @@ err_free_flow: */ static int nfp_flower_add_offload(struct nfp_app *app, struct net_device *netdev, - struct tc_cls_flower_offload *flow, bool egress) + struct tc_cls_flower_offload *flow) { enum nfp_flower_tun_type tun_type = NFP_FL_TUNNEL_NONE; - struct nfp_port *port = nfp_port_from_netdev(netdev); struct nfp_flower_priv *priv = app->priv; struct nfp_fl_payload *flow_pay; struct nfp_fl_key_ls *key_layer; - struct net_device *ingr_dev; + struct nfp_port *port = NULL; int err; - ingr_dev = egress ? NULL : netdev; - flow_pay = nfp_flower_search_fl_table(app, flow->cookie, ingr_dev, - NFP_FL_STATS_CTX_DONT_CARE); - if (flow_pay) { - /* Ignore as duplicate if it has been added by different cb. */ - if (flow_pay->ingress_offload && egress) - return 0; - else - return -EOPNOTSUPP; - } + if (nfp_netdev_is_nfp_repr(netdev)) + port = nfp_port_from_netdev(netdev); key_layer = kmalloc(sizeof(*key_layer), GFP_KERNEL); if (!key_layer) return -ENOMEM; - err = nfp_flower_calculate_key_layers(app, key_layer, flow, egress, + err = nfp_flower_calculate_key_layers(app, netdev, key_layer, flow, &tun_type); if (err) goto err_free_key_ls; - flow_pay = nfp_flower_allocate_new(key_layer, egress); + flow_pay = nfp_flower_allocate_new(key_layer); if (!flow_pay) { err = -ENOMEM; goto err_free_key_ls; } - flow_pay->ingress_dev = egress ? NULL : netdev; - - err = nfp_flower_compile_flow_match(flow, key_layer, netdev, flow_pay, - tun_type); + err = nfp_flower_compile_flow_match(app, flow, key_layer, netdev, + flow_pay, tun_type); if (err) goto err_destroy_flow; @@ -487,8 +436,7 @@ nfp_flower_add_offload(struct nfp_app *app, struct net_device *netdev, if (err) goto err_destroy_flow; - err = nfp_compile_flow_metadata(app, flow, flow_pay, - flow_pay->ingress_dev); + err = nfp_compile_flow_metadata(app, flow, flow_pay, netdev); if (err) goto err_destroy_flow; @@ -498,12 +446,13 @@ nfp_flower_add_offload(struct nfp_app *app, struct net_device *netdev, if (err) goto err_release_metadata; - err = nfp_flower_xmit_flow(netdev, flow_pay, + err = nfp_flower_xmit_flow(app, flow_pay, NFP_FLOWER_CMSG_TYPE_FLOW_ADD); if (err) goto err_remove_rhash; - port->tc_offload_cnt++; + if (port) + port->tc_offload_cnt++; /* Deallocate flow payload when flower rule has been destroyed. */ kfree(key_layer); @@ -531,7 +480,6 @@ err_free_key_ls: * @app: Pointer to the APP handle * @netdev: netdev structure. * @flow: TC flower classifier offload structure - * @egress: Netdev is the egress dev. * * Removes a flow from the repeated hash structure and clears the * action payload. @@ -540,19 +488,19 @@ err_free_key_ls: */ static int nfp_flower_del_offload(struct nfp_app *app, struct net_device *netdev, - struct tc_cls_flower_offload *flow, bool egress) + struct tc_cls_flower_offload *flow) { - struct nfp_port *port = nfp_port_from_netdev(netdev); struct nfp_flower_priv *priv = app->priv; struct nfp_fl_payload *nfp_flow; - struct net_device *ingr_dev; + struct nfp_port *port = NULL; int err; - ingr_dev = egress ? NULL : netdev; - nfp_flow = nfp_flower_search_fl_table(app, flow->cookie, ingr_dev, - NFP_FL_STATS_CTX_DONT_CARE); + if (nfp_netdev_is_nfp_repr(netdev)) + port = nfp_port_from_netdev(netdev); + + nfp_flow = nfp_flower_search_fl_table(app, flow->cookie, netdev); if (!nfp_flow) - return egress ? 0 : -ENOENT; + return -ENOENT; err = nfp_modify_flow_metadata(app, nfp_flow); if (err) @@ -561,13 +509,14 @@ nfp_flower_del_offload(struct nfp_app *app, struct net_device *netdev, if (nfp_flow->nfp_tun_ipv4_addr) nfp_tunnel_del_ipv4_off(app, nfp_flow->nfp_tun_ipv4_addr); - err = nfp_flower_xmit_flow(netdev, nfp_flow, + err = nfp_flower_xmit_flow(app, nfp_flow, NFP_FLOWER_CMSG_TYPE_FLOW_DEL); if (err) goto err_free_flow; err_free_flow: - port->tc_offload_cnt--; + if (port) + port->tc_offload_cnt--; kfree(nfp_flow->action_data); kfree(nfp_flow->mask_data); kfree(nfp_flow->unmasked_data); @@ -583,7 +532,6 @@ err_free_flow: * @app: Pointer to the APP handle * @netdev: Netdev structure. * @flow: TC flower classifier offload structure - * @egress: Netdev is the egress dev. * * Populates a flow statistics structure which which corresponds to a * specific flow. @@ -592,28 +540,21 @@ err_free_flow: */ static int nfp_flower_get_stats(struct nfp_app *app, struct net_device *netdev, - struct tc_cls_flower_offload *flow, bool egress) + struct tc_cls_flower_offload *flow) { struct nfp_flower_priv *priv = app->priv; struct nfp_fl_payload *nfp_flow; - struct net_device *ingr_dev; u32 ctx_id; - ingr_dev = egress ? NULL : netdev; - nfp_flow = nfp_flower_search_fl_table(app, flow->cookie, ingr_dev, - NFP_FL_STATS_CTX_DONT_CARE); + nfp_flow = nfp_flower_search_fl_table(app, flow->cookie, netdev); if (!nfp_flow) return -EINVAL; - if (nfp_flow->ingress_offload && egress) - return 0; - ctx_id = be32_to_cpu(nfp_flow->meta.host_ctx_id); spin_lock_bh(&priv->stats_lock); - tcf_exts_stats_update(flow->exts, priv->stats[ctx_id].bytes, - priv->stats[ctx_id].pkts, - priv->stats[ctx_id].used); + flow_stats_update(&flow->stats, priv->stats[ctx_id].bytes, + priv->stats[ctx_id].pkts, priv->stats[ctx_id].used); priv->stats[ctx_id].pkts = 0; priv->stats[ctx_id].bytes = 0; @@ -624,35 +565,18 @@ nfp_flower_get_stats(struct nfp_app *app, struct net_device *netdev, static int nfp_flower_repr_offload(struct nfp_app *app, struct net_device *netdev, - struct tc_cls_flower_offload *flower, bool egress) + struct tc_cls_flower_offload *flower) { if (!eth_proto_is_802_3(flower->common.protocol)) return -EOPNOTSUPP; switch (flower->command) { case TC_CLSFLOWER_REPLACE: - return nfp_flower_add_offload(app, netdev, flower, egress); + return nfp_flower_add_offload(app, netdev, flower); case TC_CLSFLOWER_DESTROY: - return nfp_flower_del_offload(app, netdev, flower, egress); + return nfp_flower_del_offload(app, netdev, flower); case TC_CLSFLOWER_STATS: - return nfp_flower_get_stats(app, netdev, flower, egress); - default: - return -EOPNOTSUPP; - } -} - -int nfp_flower_setup_tc_egress_cb(enum tc_setup_type type, void *type_data, - void *cb_priv) -{ - struct nfp_repr *repr = cb_priv; - - if (!tc_cls_can_offload_and_chain0(repr->netdev, type_data)) - return -EOPNOTSUPP; - - switch (type) { - case TC_SETUP_CLSFLOWER: - return nfp_flower_repr_offload(repr->app, repr->netdev, - type_data, true); + return nfp_flower_get_stats(app, netdev, flower); default: return -EOPNOTSUPP; } @@ -669,7 +593,7 @@ static int nfp_flower_setup_tc_block_cb(enum tc_setup_type type, switch (type) { case TC_SETUP_CLSFLOWER: return nfp_flower_repr_offload(repr->app, repr->netdev, - type_data, false); + type_data); default: return -EOPNOTSUPP; } @@ -708,3 +632,130 @@ int nfp_flower_setup_tc(struct nfp_app *app, struct net_device *netdev, return -EOPNOTSUPP; } } + +struct nfp_flower_indr_block_cb_priv { + struct net_device *netdev; + struct nfp_app *app; + struct list_head list; +}; + +static struct nfp_flower_indr_block_cb_priv * +nfp_flower_indr_block_cb_priv_lookup(struct nfp_app *app, + struct net_device *netdev) +{ + struct nfp_flower_indr_block_cb_priv *cb_priv; + struct nfp_flower_priv *priv = app->priv; + + /* All callback list access should be protected by RTNL. */ + ASSERT_RTNL(); + + list_for_each_entry(cb_priv, &priv->indr_block_cb_priv, list) + if (cb_priv->netdev == netdev) + return cb_priv; + + return NULL; +} + +static int nfp_flower_setup_indr_block_cb(enum tc_setup_type type, + void *type_data, void *cb_priv) +{ + struct nfp_flower_indr_block_cb_priv *priv = cb_priv; + struct tc_cls_flower_offload *flower = type_data; + + if (flower->common.chain_index) + return -EOPNOTSUPP; + + switch (type) { + case TC_SETUP_CLSFLOWER: + return nfp_flower_repr_offload(priv->app, priv->netdev, + type_data); + default: + return -EOPNOTSUPP; + } +} + +static int +nfp_flower_setup_indr_tc_block(struct net_device *netdev, struct nfp_app *app, + struct tc_block_offload *f) +{ + struct nfp_flower_indr_block_cb_priv *cb_priv; + struct nfp_flower_priv *priv = app->priv; + int err; + + if (f->binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) + return -EOPNOTSUPP; + + switch (f->command) { + case TC_BLOCK_BIND: + cb_priv = kmalloc(sizeof(*cb_priv), GFP_KERNEL); + if (!cb_priv) + return -ENOMEM; + + cb_priv->netdev = netdev; + cb_priv->app = app; + list_add(&cb_priv->list, &priv->indr_block_cb_priv); + + err = tcf_block_cb_register(f->block, + nfp_flower_setup_indr_block_cb, + cb_priv, cb_priv, f->extack); + if (err) { + list_del(&cb_priv->list); + kfree(cb_priv); + } + + return err; + case TC_BLOCK_UNBIND: + cb_priv = nfp_flower_indr_block_cb_priv_lookup(app, netdev); + if (!cb_priv) + return -ENOENT; + + tcf_block_cb_unregister(f->block, + nfp_flower_setup_indr_block_cb, + cb_priv); + list_del(&cb_priv->list); + kfree(cb_priv); + + return 0; + default: + return -EOPNOTSUPP; + } + return 0; +} + +static int +nfp_flower_indr_setup_tc_cb(struct net_device *netdev, void *cb_priv, + enum tc_setup_type type, void *type_data) +{ + switch (type) { + case TC_SETUP_BLOCK: + return nfp_flower_setup_indr_tc_block(netdev, cb_priv, + type_data); + default: + return -EOPNOTSUPP; + } +} + +int nfp_flower_reg_indir_block_handler(struct nfp_app *app, + struct net_device *netdev, + unsigned long event) +{ + int err; + + if (!nfp_fl_is_netdev_to_offload(netdev)) + return NOTIFY_OK; + + if (event == NETDEV_REGISTER) { + err = __tc_indr_block_cb_register(netdev, app, + nfp_flower_indr_setup_tc_cb, + app); + if (err) + nfp_flower_cmsg_warn(app, + "Indirect block reg failed - %s\n", + netdev->name); + } else if (event == NETDEV_UNREGISTER) { + __tc_indr_block_cb_unregister(netdev, + nfp_flower_indr_setup_tc_cb, app); + } + + return NOTIFY_OK; +} diff --git a/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c b/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c index 8e5bec04d1f9..4d78be4ec4e9 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c +++ b/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c @@ -4,7 +4,6 @@ #include <linux/etherdevice.h> #include <linux/inetdevice.h> #include <net/netevent.h> -#include <net/vxlan.h> #include <linux/idr.h> #include <net/dst_metadata.h> #include <net/arp.h> @@ -99,47 +98,51 @@ struct nfp_ipv4_addr_entry { struct list_head list; }; -/** - * struct nfp_tun_mac_addr - configure MAC address of tunnel EP on NFP - * @reserved: reserved for future use - * @count: number of MAC addresses in the message - * @addresses.index: index of MAC address in the lookup table - * @addresses.addr: interface MAC address - * @addresses: series of MACs to offload - */ -struct nfp_tun_mac_addr { - __be16 reserved; - __be16 count; - struct index_mac_addr { - __be16 index; - u8 addr[ETH_ALEN]; - } addresses[]; -}; +#define NFP_TUN_MAC_OFFLOAD_DEL_FLAG 0x2 /** - * struct nfp_tun_mac_offload_entry - list of MACs to offload - * @index: index of MAC address for offloading + * struct nfp_tun_mac_addr_offload - configure MAC address of tunnel EP on NFP + * @flags: MAC address offload options + * @count: number of MAC addresses in the message (should be 1) + * @index: index of MAC address in the lookup table * @addr: interface MAC address - * @list: list pointer */ -struct nfp_tun_mac_offload_entry { +struct nfp_tun_mac_addr_offload { + __be16 flags; + __be16 count; __be16 index; u8 addr[ETH_ALEN]; - struct list_head list; +}; + +enum nfp_flower_mac_offload_cmd { + NFP_TUNNEL_MAC_OFFLOAD_ADD = 0, + NFP_TUNNEL_MAC_OFFLOAD_DEL = 1, + NFP_TUNNEL_MAC_OFFLOAD_MOD = 2, }; #define NFP_MAX_MAC_INDEX 0xff /** - * struct nfp_tun_mac_non_nfp_idx - converts non NFP netdev ifindex to 8-bit id - * @ifindex: netdev ifindex of the device - * @index: index of netdevs mac on NFP - * @list: list pointer + * struct nfp_tun_offloaded_mac - hashtable entry for an offloaded MAC + * @ht_node: Hashtable entry + * @addr: Offloaded MAC address + * @index: Offloaded index for given MAC address + * @ref_count: Number of devs using this MAC address + * @repr_list: List of reprs sharing this MAC address */ -struct nfp_tun_mac_non_nfp_idx { - int ifindex; - u8 index; - struct list_head list; +struct nfp_tun_offloaded_mac { + struct rhash_head ht_node; + u8 addr[ETH_ALEN]; + u16 index; + int ref_count; + struct list_head repr_list; +}; + +static const struct rhashtable_params offloaded_macs_params = { + .key_offset = offsetof(struct nfp_tun_offloaded_mac, addr), + .head_offset = offsetof(struct nfp_tun_offloaded_mac, ht_node), + .key_len = ETH_ALEN, + .automatic_shrinking = true, }; void nfp_tunnel_keep_alive(struct nfp_app *app, struct sk_buff *skb) @@ -182,18 +185,6 @@ void nfp_tunnel_keep_alive(struct nfp_app *app, struct sk_buff *skb) } } -static bool nfp_tun_is_netdev_to_offload(struct net_device *netdev) -{ - if (!netdev->rtnl_link_ops) - return false; - if (!strcmp(netdev->rtnl_link_ops->kind, "openvswitch")) - return true; - if (netif_is_vxlan(netdev)) - return true; - - return false; -} - static int nfp_flower_xmit_tun_conf(struct nfp_app *app, u8 mtype, u16 plen, void *pdata, gfp_t flag) @@ -218,15 +209,15 @@ static bool nfp_tun_has_route(struct nfp_app *app, __be32 ipv4_addr) struct nfp_ipv4_route_entry *entry; struct list_head *ptr, *storage; - spin_lock_bh(&priv->nfp_neigh_off_lock); - list_for_each_safe(ptr, storage, &priv->nfp_neigh_off_list) { + spin_lock_bh(&priv->tun.neigh_off_lock); + list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); if (entry->ipv4_addr == ipv4_addr) { - spin_unlock_bh(&priv->nfp_neigh_off_lock); + spin_unlock_bh(&priv->tun.neigh_off_lock); return true; } } - spin_unlock_bh(&priv->nfp_neigh_off_lock); + spin_unlock_bh(&priv->tun.neigh_off_lock); return false; } @@ -236,24 +227,24 @@ static void nfp_tun_add_route_to_cache(struct nfp_app *app, __be32 ipv4_addr) struct nfp_ipv4_route_entry *entry; struct list_head *ptr, *storage; - spin_lock_bh(&priv->nfp_neigh_off_lock); - list_for_each_safe(ptr, storage, &priv->nfp_neigh_off_list) { + spin_lock_bh(&priv->tun.neigh_off_lock); + list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); if (entry->ipv4_addr == ipv4_addr) { - spin_unlock_bh(&priv->nfp_neigh_off_lock); + spin_unlock_bh(&priv->tun.neigh_off_lock); return; } } entry = kmalloc(sizeof(*entry), GFP_ATOMIC); if (!entry) { - spin_unlock_bh(&priv->nfp_neigh_off_lock); + spin_unlock_bh(&priv->tun.neigh_off_lock); nfp_flower_cmsg_warn(app, "Mem error when storing new route.\n"); return; } entry->ipv4_addr = ipv4_addr; - list_add_tail(&entry->list, &priv->nfp_neigh_off_list); - spin_unlock_bh(&priv->nfp_neigh_off_lock); + list_add_tail(&entry->list, &priv->tun.neigh_off_list); + spin_unlock_bh(&priv->tun.neigh_off_lock); } static void nfp_tun_del_route_from_cache(struct nfp_app *app, __be32 ipv4_addr) @@ -262,8 +253,8 @@ static void nfp_tun_del_route_from_cache(struct nfp_app *app, __be32 ipv4_addr) struct nfp_ipv4_route_entry *entry; struct list_head *ptr, *storage; - spin_lock_bh(&priv->nfp_neigh_off_lock); - list_for_each_safe(ptr, storage, &priv->nfp_neigh_off_list) { + spin_lock_bh(&priv->tun.neigh_off_lock); + list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); if (entry->ipv4_addr == ipv4_addr) { list_del(&entry->list); @@ -271,7 +262,7 @@ static void nfp_tun_del_route_from_cache(struct nfp_app *app, __be32 ipv4_addr) break; } } - spin_unlock_bh(&priv->nfp_neigh_off_lock); + spin_unlock_bh(&priv->tun.neigh_off_lock); } static void @@ -339,7 +330,7 @@ nfp_tun_neigh_event_handler(struct notifier_block *nb, unsigned long event, if (!nfp_netdev_is_nfp_repr(n->dev)) return NOTIFY_DONE; - app_priv = container_of(nb, struct nfp_flower_priv, nfp_tun_neigh_nb); + app_priv = container_of(nb, struct nfp_flower_priv, tun.neigh_nb); app = app_priv->app; /* Only concerned with changes to routes already added to NFP. */ @@ -414,11 +405,11 @@ static void nfp_tun_write_ipv4_list(struct nfp_app *app) int count; memset(&payload, 0, sizeof(struct nfp_tun_ipv4_addr)); - mutex_lock(&priv->nfp_ipv4_off_lock); + mutex_lock(&priv->tun.ipv4_off_lock); count = 0; - list_for_each_safe(ptr, storage, &priv->nfp_ipv4_off_list) { + list_for_each_safe(ptr, storage, &priv->tun.ipv4_off_list) { if (count >= NFP_FL_IPV4_ADDRS_MAX) { - mutex_unlock(&priv->nfp_ipv4_off_lock); + mutex_unlock(&priv->tun.ipv4_off_lock); nfp_flower_cmsg_warn(app, "IPv4 offload exceeds limit.\n"); return; } @@ -426,7 +417,7 @@ static void nfp_tun_write_ipv4_list(struct nfp_app *app) payload.ipv4_addr[count++] = entry->ipv4_addr; } payload.count = cpu_to_be32(count); - mutex_unlock(&priv->nfp_ipv4_off_lock); + mutex_unlock(&priv->tun.ipv4_off_lock); nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_IPS, sizeof(struct nfp_tun_ipv4_addr), @@ -439,26 +430,26 @@ void nfp_tunnel_add_ipv4_off(struct nfp_app *app, __be32 ipv4) struct nfp_ipv4_addr_entry *entry; struct list_head *ptr, *storage; - mutex_lock(&priv->nfp_ipv4_off_lock); - list_for_each_safe(ptr, storage, &priv->nfp_ipv4_off_list) { + mutex_lock(&priv->tun.ipv4_off_lock); + list_for_each_safe(ptr, storage, &priv->tun.ipv4_off_list) { entry = list_entry(ptr, struct nfp_ipv4_addr_entry, list); if (entry->ipv4_addr == ipv4) { entry->ref_count++; - mutex_unlock(&priv->nfp_ipv4_off_lock); + mutex_unlock(&priv->tun.ipv4_off_lock); return; } } entry = kmalloc(sizeof(*entry), GFP_KERNEL); if (!entry) { - mutex_unlock(&priv->nfp_ipv4_off_lock); + mutex_unlock(&priv->tun.ipv4_off_lock); nfp_flower_cmsg_warn(app, "Mem error when offloading IP address.\n"); return; } entry->ipv4_addr = ipv4; entry->ref_count = 1; - list_add_tail(&entry->list, &priv->nfp_ipv4_off_list); - mutex_unlock(&priv->nfp_ipv4_off_lock); + list_add_tail(&entry->list, &priv->tun.ipv4_off_list); + mutex_unlock(&priv->tun.ipv4_off_lock); nfp_tun_write_ipv4_list(app); } @@ -469,8 +460,8 @@ void nfp_tunnel_del_ipv4_off(struct nfp_app *app, __be32 ipv4) struct nfp_ipv4_addr_entry *entry; struct list_head *ptr, *storage; - mutex_lock(&priv->nfp_ipv4_off_lock); - list_for_each_safe(ptr, storage, &priv->nfp_ipv4_off_list) { + mutex_lock(&priv->tun.ipv4_off_lock); + list_for_each_safe(ptr, storage, &priv->tun.ipv4_off_list) { entry = list_entry(ptr, struct nfp_ipv4_addr_entry, list); if (entry->ipv4_addr == ipv4) { entry->ref_count--; @@ -481,204 +472,357 @@ void nfp_tunnel_del_ipv4_off(struct nfp_app *app, __be32 ipv4) break; } } - mutex_unlock(&priv->nfp_ipv4_off_lock); + mutex_unlock(&priv->tun.ipv4_off_lock); nfp_tun_write_ipv4_list(app); } -void nfp_tunnel_write_macs(struct nfp_app *app) +static int +__nfp_tunnel_offload_mac(struct nfp_app *app, u8 *mac, u16 idx, bool del) { - struct nfp_flower_priv *priv = app->priv; - struct nfp_tun_mac_offload_entry *entry; - struct nfp_tun_mac_addr *payload; - struct list_head *ptr, *storage; - int mac_count, err, pay_size; + struct nfp_tun_mac_addr_offload payload; - mutex_lock(&priv->nfp_mac_off_lock); - if (!priv->nfp_mac_off_count) { - mutex_unlock(&priv->nfp_mac_off_lock); - return; - } + memset(&payload, 0, sizeof(payload)); - pay_size = sizeof(struct nfp_tun_mac_addr) + - sizeof(struct index_mac_addr) * priv->nfp_mac_off_count; + if (del) + payload.flags = cpu_to_be16(NFP_TUN_MAC_OFFLOAD_DEL_FLAG); - payload = kzalloc(pay_size, GFP_KERNEL); - if (!payload) { - mutex_unlock(&priv->nfp_mac_off_lock); - return; - } + /* FW supports multiple MACs per cmsg but restrict to single. */ + payload.count = cpu_to_be16(1); + payload.index = cpu_to_be16(idx); + ether_addr_copy(payload.addr, mac); - payload->count = cpu_to_be16(priv->nfp_mac_off_count); + return nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_MAC, + sizeof(struct nfp_tun_mac_addr_offload), + &payload, GFP_KERNEL); +} - mac_count = 0; - list_for_each_safe(ptr, storage, &priv->nfp_mac_off_list) { - entry = list_entry(ptr, struct nfp_tun_mac_offload_entry, - list); - payload->addresses[mac_count].index = entry->index; - ether_addr_copy(payload->addresses[mac_count].addr, - entry->addr); - mac_count++; - } +static bool nfp_tunnel_port_is_phy_repr(int port) +{ + if (FIELD_GET(NFP_FLOWER_CMSG_PORT_TYPE, port) == + NFP_FLOWER_CMSG_PORT_TYPE_PHYS_PORT) + return true; - err = nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_MAC, - pay_size, payload, GFP_KERNEL); + return false; +} - kfree(payload); +static u16 nfp_tunnel_get_mac_idx_from_phy_port_id(int port) +{ + return port << 8 | NFP_FLOWER_CMSG_PORT_TYPE_PHYS_PORT; +} - if (err) { - mutex_unlock(&priv->nfp_mac_off_lock); - /* Write failed so retain list for future retry. */ - return; - } +static u16 nfp_tunnel_get_global_mac_idx_from_ida(int id) +{ + return id << 8 | NFP_FLOWER_CMSG_PORT_TYPE_OTHER_PORT; +} + +static int nfp_tunnel_get_ida_from_global_mac_idx(u16 nfp_mac_idx) +{ + return nfp_mac_idx >> 8; +} + +static bool nfp_tunnel_is_mac_idx_global(u16 nfp_mac_idx) +{ + return (nfp_mac_idx & 0xff) == NFP_FLOWER_CMSG_PORT_TYPE_OTHER_PORT; +} + +static struct nfp_tun_offloaded_mac * +nfp_tunnel_lookup_offloaded_macs(struct nfp_app *app, u8 *mac) +{ + struct nfp_flower_priv *priv = app->priv; + + return rhashtable_lookup_fast(&priv->tun.offloaded_macs, mac, + offloaded_macs_params); +} + +static void +nfp_tunnel_offloaded_macs_inc_ref_and_link(struct nfp_tun_offloaded_mac *entry, + struct net_device *netdev, bool mod) +{ + if (nfp_netdev_is_nfp_repr(netdev)) { + struct nfp_flower_repr_priv *repr_priv; + struct nfp_repr *repr; + + repr = netdev_priv(netdev); + repr_priv = repr->app_priv; - /* If list was successfully offloaded, flush it. */ - list_for_each_safe(ptr, storage, &priv->nfp_mac_off_list) { - entry = list_entry(ptr, struct nfp_tun_mac_offload_entry, - list); - list_del(&entry->list); - kfree(entry); + /* If modifing MAC, remove repr from old list first. */ + if (mod) + list_del(&repr_priv->mac_list); + + list_add_tail(&repr_priv->mac_list, &entry->repr_list); } - priv->nfp_mac_off_count = 0; - mutex_unlock(&priv->nfp_mac_off_lock); + entry->ref_count++; } -static int nfp_tun_get_mac_idx(struct nfp_app *app, int ifindex) +static int +nfp_tunnel_add_shared_mac(struct nfp_app *app, struct net_device *netdev, + int port, bool mod) { struct nfp_flower_priv *priv = app->priv; - struct nfp_tun_mac_non_nfp_idx *entry; - struct list_head *ptr, *storage; - int idx; - - mutex_lock(&priv->nfp_mac_index_lock); - list_for_each_safe(ptr, storage, &priv->nfp_mac_index_list) { - entry = list_entry(ptr, struct nfp_tun_mac_non_nfp_idx, list); - if (entry->ifindex == ifindex) { - idx = entry->index; - mutex_unlock(&priv->nfp_mac_index_lock); - return idx; - } + int ida_idx = NFP_MAX_MAC_INDEX, err; + struct nfp_tun_offloaded_mac *entry; + u16 nfp_mac_idx = 0; + + entry = nfp_tunnel_lookup_offloaded_macs(app, netdev->dev_addr); + if (entry && nfp_tunnel_is_mac_idx_global(entry->index)) { + nfp_tunnel_offloaded_macs_inc_ref_and_link(entry, netdev, mod); + return 0; } - idx = ida_simple_get(&priv->nfp_mac_off_ids, 0, - NFP_MAX_MAC_INDEX, GFP_KERNEL); - if (idx < 0) { - mutex_unlock(&priv->nfp_mac_index_lock); - return idx; + /* Assign a global index if non-repr or MAC address is now shared. */ + if (entry || !port) { + ida_idx = ida_simple_get(&priv->tun.mac_off_ids, 0, + NFP_MAX_MAC_INDEX, GFP_KERNEL); + if (ida_idx < 0) + return ida_idx; + + nfp_mac_idx = nfp_tunnel_get_global_mac_idx_from_ida(ida_idx); + } else { + nfp_mac_idx = nfp_tunnel_get_mac_idx_from_phy_port_id(port); } - entry = kmalloc(sizeof(*entry), GFP_KERNEL); if (!entry) { - mutex_unlock(&priv->nfp_mac_index_lock); - return -ENOMEM; + entry = kzalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) { + err = -ENOMEM; + goto err_free_ida; + } + + ether_addr_copy(entry->addr, netdev->dev_addr); + INIT_LIST_HEAD(&entry->repr_list); + + if (rhashtable_insert_fast(&priv->tun.offloaded_macs, + &entry->ht_node, + offloaded_macs_params)) { + err = -ENOMEM; + goto err_free_entry; + } + } + + err = __nfp_tunnel_offload_mac(app, netdev->dev_addr, + nfp_mac_idx, false); + if (err) { + /* If not shared then free. */ + if (!entry->ref_count) + goto err_remove_hash; + goto err_free_ida; } - entry->ifindex = ifindex; - entry->index = idx; - list_add_tail(&entry->list, &priv->nfp_mac_index_list); - mutex_unlock(&priv->nfp_mac_index_lock); - return idx; + entry->index = nfp_mac_idx; + nfp_tunnel_offloaded_macs_inc_ref_and_link(entry, netdev, mod); + + return 0; + +err_remove_hash: + rhashtable_remove_fast(&priv->tun.offloaded_macs, &entry->ht_node, + offloaded_macs_params); +err_free_entry: + kfree(entry); +err_free_ida: + if (ida_idx != NFP_MAX_MAC_INDEX) + ida_simple_remove(&priv->tun.mac_off_ids, ida_idx); + + return err; } -static void nfp_tun_del_mac_idx(struct nfp_app *app, int ifindex) +static int +nfp_tunnel_del_shared_mac(struct nfp_app *app, struct net_device *netdev, + u8 *mac, bool mod) { struct nfp_flower_priv *priv = app->priv; - struct nfp_tun_mac_non_nfp_idx *entry; - struct list_head *ptr, *storage; + struct nfp_flower_repr_priv *repr_priv; + struct nfp_tun_offloaded_mac *entry; + struct nfp_repr *repr; + int ida_idx; + + entry = nfp_tunnel_lookup_offloaded_macs(app, mac); + if (!entry) + return 0; + + entry->ref_count--; + /* If del is part of a mod then mac_list is still in use elsewheree. */ + if (nfp_netdev_is_nfp_repr(netdev) && !mod) { + repr = netdev_priv(netdev); + repr_priv = repr->app_priv; + list_del(&repr_priv->mac_list); + } - mutex_lock(&priv->nfp_mac_index_lock); - list_for_each_safe(ptr, storage, &priv->nfp_mac_index_list) { - entry = list_entry(ptr, struct nfp_tun_mac_non_nfp_idx, list); - if (entry->ifindex == ifindex) { - ida_simple_remove(&priv->nfp_mac_off_ids, - entry->index); - list_del(&entry->list); - kfree(entry); - break; + /* If MAC is now used by 1 repr set the offloaded MAC index to port. */ + if (entry->ref_count == 1 && list_is_singular(&entry->repr_list)) { + u16 nfp_mac_idx; + int port, err; + + repr_priv = list_first_entry(&entry->repr_list, + struct nfp_flower_repr_priv, + mac_list); + repr = repr_priv->nfp_repr; + port = nfp_repr_get_port_id(repr->netdev); + nfp_mac_idx = nfp_tunnel_get_mac_idx_from_phy_port_id(port); + err = __nfp_tunnel_offload_mac(app, mac, nfp_mac_idx, false); + if (err) { + nfp_flower_cmsg_warn(app, "MAC offload index revert failed on %s.\n", + netdev_name(netdev)); + return 0; } + + ida_idx = nfp_tunnel_get_ida_from_global_mac_idx(entry->index); + ida_simple_remove(&priv->tun.mac_off_ids, ida_idx); + entry->index = nfp_mac_idx; + return 0; } - mutex_unlock(&priv->nfp_mac_index_lock); -} -static void nfp_tun_add_to_mac_offload_list(struct net_device *netdev, - struct nfp_app *app) -{ - struct nfp_flower_priv *priv = app->priv; - struct nfp_tun_mac_offload_entry *entry; - u16 nfp_mac_idx; - int port = 0; + if (entry->ref_count) + return 0; - /* Check if MAC should be offloaded. */ - if (!is_valid_ether_addr(netdev->dev_addr)) - return; + WARN_ON_ONCE(rhashtable_remove_fast(&priv->tun.offloaded_macs, + &entry->ht_node, + offloaded_macs_params)); + /* If MAC has global ID then extract and free the ida entry. */ + if (nfp_tunnel_is_mac_idx_global(entry->index)) { + ida_idx = nfp_tunnel_get_ida_from_global_mac_idx(entry->index); + ida_simple_remove(&priv->tun.mac_off_ids, ida_idx); + } + + kfree(entry); - if (nfp_netdev_is_nfp_repr(netdev)) + return __nfp_tunnel_offload_mac(app, mac, 0, true); +} + +static int +nfp_tunnel_offload_mac(struct nfp_app *app, struct net_device *netdev, + enum nfp_flower_mac_offload_cmd cmd) +{ + struct nfp_flower_non_repr_priv *nr_priv = NULL; + bool non_repr = false, *mac_offloaded; + u8 *off_mac = NULL; + int err, port = 0; + + if (nfp_netdev_is_nfp_repr(netdev)) { + struct nfp_flower_repr_priv *repr_priv; + struct nfp_repr *repr; + + repr = netdev_priv(netdev); + if (repr->app != app) + return 0; + + repr_priv = repr->app_priv; + mac_offloaded = &repr_priv->mac_offloaded; + off_mac = &repr_priv->offloaded_mac_addr[0]; port = nfp_repr_get_port_id(netdev); - else if (!nfp_tun_is_netdev_to_offload(netdev)) - return; + if (!nfp_tunnel_port_is_phy_repr(port)) + return 0; + } else if (nfp_fl_is_netdev_to_offload(netdev)) { + nr_priv = nfp_flower_non_repr_priv_get(app, netdev); + if (!nr_priv) + return -ENOMEM; + + mac_offloaded = &nr_priv->mac_offloaded; + off_mac = &nr_priv->offloaded_mac_addr[0]; + non_repr = true; + } else { + return 0; + } - entry = kmalloc(sizeof(*entry), GFP_KERNEL); - if (!entry) { - nfp_flower_cmsg_warn(app, "Mem fail when offloading MAC.\n"); - return; + if (!is_valid_ether_addr(netdev->dev_addr)) { + err = -EINVAL; + goto err_put_non_repr_priv; } - if (FIELD_GET(NFP_FLOWER_CMSG_PORT_TYPE, port) == - NFP_FLOWER_CMSG_PORT_TYPE_PHYS_PORT) { - nfp_mac_idx = port << 8 | NFP_FLOWER_CMSG_PORT_TYPE_PHYS_PORT; - } else if (FIELD_GET(NFP_FLOWER_CMSG_PORT_TYPE, port) == - NFP_FLOWER_CMSG_PORT_TYPE_PCIE_PORT) { - port = FIELD_GET(NFP_FLOWER_CMSG_PORT_VNIC, port); - nfp_mac_idx = port << 8 | NFP_FLOWER_CMSG_PORT_TYPE_PCIE_PORT; - } else { - /* Must assign our own unique 8-bit index. */ - int idx = nfp_tun_get_mac_idx(app, netdev->ifindex); + if (cmd == NFP_TUNNEL_MAC_OFFLOAD_MOD && !*mac_offloaded) + cmd = NFP_TUNNEL_MAC_OFFLOAD_ADD; - if (idx < 0) { - nfp_flower_cmsg_warn(app, "Can't assign non-repr MAC index.\n"); - kfree(entry); - return; - } - nfp_mac_idx = idx << 8 | NFP_FLOWER_CMSG_PORT_TYPE_OTHER_PORT; + switch (cmd) { + case NFP_TUNNEL_MAC_OFFLOAD_ADD: + err = nfp_tunnel_add_shared_mac(app, netdev, port, false); + if (err) + goto err_put_non_repr_priv; + + if (non_repr) + __nfp_flower_non_repr_priv_get(nr_priv); + + *mac_offloaded = true; + ether_addr_copy(off_mac, netdev->dev_addr); + break; + case NFP_TUNNEL_MAC_OFFLOAD_DEL: + /* Only attempt delete if add was successful. */ + if (!*mac_offloaded) + break; + + if (non_repr) + __nfp_flower_non_repr_priv_put(nr_priv); + + *mac_offloaded = false; + + err = nfp_tunnel_del_shared_mac(app, netdev, netdev->dev_addr, + false); + if (err) + goto err_put_non_repr_priv; + + break; + case NFP_TUNNEL_MAC_OFFLOAD_MOD: + /* Ignore if changing to the same address. */ + if (ether_addr_equal(netdev->dev_addr, off_mac)) + break; + + err = nfp_tunnel_add_shared_mac(app, netdev, port, true); + if (err) + goto err_put_non_repr_priv; + + /* Delete the previous MAC address. */ + err = nfp_tunnel_del_shared_mac(app, netdev, off_mac, true); + if (err) + nfp_flower_cmsg_warn(app, "Failed to remove offload of replaced MAC addr on %s.\n", + netdev_name(netdev)); + + ether_addr_copy(off_mac, netdev->dev_addr); + break; + default: + err = -EINVAL; + goto err_put_non_repr_priv; } - entry->index = cpu_to_be16(nfp_mac_idx); - ether_addr_copy(entry->addr, netdev->dev_addr); + if (non_repr) + __nfp_flower_non_repr_priv_put(nr_priv); + + return 0; - mutex_lock(&priv->nfp_mac_off_lock); - priv->nfp_mac_off_count++; - list_add_tail(&entry->list, &priv->nfp_mac_off_list); - mutex_unlock(&priv->nfp_mac_off_lock); +err_put_non_repr_priv: + if (non_repr) + __nfp_flower_non_repr_priv_put(nr_priv); + + return err; } -static int nfp_tun_mac_event_handler(struct notifier_block *nb, - unsigned long event, void *ptr) +int nfp_tunnel_mac_event_handler(struct nfp_app *app, + struct net_device *netdev, + unsigned long event, void *ptr) { - struct nfp_flower_priv *app_priv; - struct net_device *netdev; - struct nfp_app *app; + int err; - if (event == NETDEV_DOWN || event == NETDEV_UNREGISTER) { - app_priv = container_of(nb, struct nfp_flower_priv, - nfp_tun_mac_nb); - app = app_priv->app; - netdev = netdev_notifier_info_to_dev(ptr); - - /* If non-nfp netdev then free its offload index. */ - if (nfp_tun_is_netdev_to_offload(netdev)) - nfp_tun_del_mac_idx(app, netdev->ifindex); - } else if (event == NETDEV_UP || event == NETDEV_CHANGEADDR || - event == NETDEV_REGISTER) { - app_priv = container_of(nb, struct nfp_flower_priv, - nfp_tun_mac_nb); - app = app_priv->app; - netdev = netdev_notifier_info_to_dev(ptr); - - nfp_tun_add_to_mac_offload_list(netdev, app); - - /* Force a list write to keep NFP up to date. */ - nfp_tunnel_write_macs(app); + if (event == NETDEV_DOWN) { + err = nfp_tunnel_offload_mac(app, netdev, + NFP_TUNNEL_MAC_OFFLOAD_DEL); + if (err) + nfp_flower_cmsg_warn(app, "Failed to delete offload MAC on %s.\n", + netdev_name(netdev)); + } else if (event == NETDEV_UP) { + err = nfp_tunnel_offload_mac(app, netdev, + NFP_TUNNEL_MAC_OFFLOAD_ADD); + if (err) + nfp_flower_cmsg_warn(app, "Failed to offload MAC on %s.\n", + netdev_name(netdev)); + } else if (event == NETDEV_CHANGEADDR) { + /* Only offload addr change if netdev is already up. */ + if (!(netdev->flags & IFF_UP)) + return NOTIFY_OK; + + err = nfp_tunnel_offload_mac(app, netdev, + NFP_TUNNEL_MAC_OFFLOAD_MOD); + if (err) + nfp_flower_cmsg_warn(app, "Failed to offload MAC change on %s.\n", + netdev_name(netdev)); } return NOTIFY_OK; } @@ -686,92 +830,62 @@ static int nfp_tun_mac_event_handler(struct notifier_block *nb, int nfp_tunnel_config_start(struct nfp_app *app) { struct nfp_flower_priv *priv = app->priv; - struct net_device *netdev; int err; - /* Initialise priv data for MAC offloading. */ - priv->nfp_mac_off_count = 0; - mutex_init(&priv->nfp_mac_off_lock); - INIT_LIST_HEAD(&priv->nfp_mac_off_list); - priv->nfp_tun_mac_nb.notifier_call = nfp_tun_mac_event_handler; - mutex_init(&priv->nfp_mac_index_lock); - INIT_LIST_HEAD(&priv->nfp_mac_index_list); - ida_init(&priv->nfp_mac_off_ids); + /* Initialise rhash for MAC offload tracking. */ + err = rhashtable_init(&priv->tun.offloaded_macs, + &offloaded_macs_params); + if (err) + return err; + + ida_init(&priv->tun.mac_off_ids); /* Initialise priv data for IPv4 offloading. */ - mutex_init(&priv->nfp_ipv4_off_lock); - INIT_LIST_HEAD(&priv->nfp_ipv4_off_list); + mutex_init(&priv->tun.ipv4_off_lock); + INIT_LIST_HEAD(&priv->tun.ipv4_off_list); /* Initialise priv data for neighbour offloading. */ - spin_lock_init(&priv->nfp_neigh_off_lock); - INIT_LIST_HEAD(&priv->nfp_neigh_off_list); - priv->nfp_tun_neigh_nb.notifier_call = nfp_tun_neigh_event_handler; - - err = register_netdevice_notifier(&priv->nfp_tun_mac_nb); - if (err) - goto err_free_mac_ida; - - err = register_netevent_notifier(&priv->nfp_tun_neigh_nb); - if (err) - goto err_unreg_mac_nb; + spin_lock_init(&priv->tun.neigh_off_lock); + INIT_LIST_HEAD(&priv->tun.neigh_off_list); + priv->tun.neigh_nb.notifier_call = nfp_tun_neigh_event_handler; - /* Parse netdevs already registered for MACs that need offloaded. */ - rtnl_lock(); - for_each_netdev(&init_net, netdev) - nfp_tun_add_to_mac_offload_list(netdev, app); - rtnl_unlock(); + err = register_netevent_notifier(&priv->tun.neigh_nb); + if (err) { + rhashtable_free_and_destroy(&priv->tun.offloaded_macs, + nfp_check_rhashtable_empty, NULL); + return err; + } return 0; - -err_unreg_mac_nb: - unregister_netdevice_notifier(&priv->nfp_tun_mac_nb); -err_free_mac_ida: - ida_destroy(&priv->nfp_mac_off_ids); - return err; } void nfp_tunnel_config_stop(struct nfp_app *app) { - struct nfp_tun_mac_offload_entry *mac_entry; struct nfp_flower_priv *priv = app->priv; struct nfp_ipv4_route_entry *route_entry; - struct nfp_tun_mac_non_nfp_idx *mac_idx; struct nfp_ipv4_addr_entry *ip_entry; struct list_head *ptr, *storage; - unregister_netdevice_notifier(&priv->nfp_tun_mac_nb); - unregister_netevent_notifier(&priv->nfp_tun_neigh_nb); + unregister_netevent_notifier(&priv->tun.neigh_nb); - /* Free any memory that may be occupied by MAC list. */ - list_for_each_safe(ptr, storage, &priv->nfp_mac_off_list) { - mac_entry = list_entry(ptr, struct nfp_tun_mac_offload_entry, - list); - list_del(&mac_entry->list); - kfree(mac_entry); - } - - /* Free any memory that may be occupied by MAC index list. */ - list_for_each_safe(ptr, storage, &priv->nfp_mac_index_list) { - mac_idx = list_entry(ptr, struct nfp_tun_mac_non_nfp_idx, - list); - list_del(&mac_idx->list); - kfree(mac_idx); - } - - ida_destroy(&priv->nfp_mac_off_ids); + ida_destroy(&priv->tun.mac_off_ids); /* Free any memory that may be occupied by ipv4 list. */ - list_for_each_safe(ptr, storage, &priv->nfp_ipv4_off_list) { + list_for_each_safe(ptr, storage, &priv->tun.ipv4_off_list) { ip_entry = list_entry(ptr, struct nfp_ipv4_addr_entry, list); list_del(&ip_entry->list); kfree(ip_entry); } /* Free any memory that may be occupied by the route list. */ - list_for_each_safe(ptr, storage, &priv->nfp_neigh_off_list) { + list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { route_entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); list_del(&route_entry->list); kfree(route_entry); } + + /* Destroy rhash. Entries should be cleaned on netdev notifier unreg. */ + rhashtable_free_and_destroy(&priv->tun.offloaded_macs, + nfp_check_rhashtable_empty, NULL); } diff --git a/drivers/net/ethernet/netronome/nfp/nfp_app.c b/drivers/net/ethernet/netronome/nfp/nfp_app.c index 68a0991aac22..3a973282b2bb 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_app.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_app.c @@ -131,11 +131,100 @@ nfp_app_reprs_set(struct nfp_app *app, enum nfp_repr_type type, struct nfp_reprs *old; old = nfp_reprs_get_locked(app, type); + rtnl_lock(); rcu_assign_pointer(app->reprs[type], reprs); + rtnl_unlock(); return old; } +static void +nfp_app_netdev_feat_change(struct nfp_app *app, struct net_device *netdev) +{ + struct nfp_net *nn; + unsigned int type; + + if (!nfp_netdev_is_nfp_net(netdev)) + return; + nn = netdev_priv(netdev); + if (nn->app != app) + return; + + for (type = 0; type < __NFP_REPR_TYPE_MAX; type++) { + struct nfp_reprs *reprs; + unsigned int i; + + reprs = rtnl_dereference(app->reprs[type]); + if (!reprs) + continue; + + for (i = 0; i < reprs->num_reprs; i++) { + struct net_device *repr; + + repr = rtnl_dereference(reprs->reprs[i]); + if (!repr) + continue; + + nfp_repr_transfer_features(repr, netdev); + } + } +} + +static int +nfp_app_netdev_event(struct notifier_block *nb, unsigned long event, void *ptr) +{ + struct net_device *netdev; + struct nfp_app *app; + + netdev = netdev_notifier_info_to_dev(ptr); + app = container_of(nb, struct nfp_app, netdev_nb); + + /* Handle events common code is interested in */ + switch (event) { + case NETDEV_FEAT_CHANGE: + nfp_app_netdev_feat_change(app, netdev); + break; + } + + /* Call offload specific handlers */ + if (app->type->netdev_event) + return app->type->netdev_event(app, netdev, event, ptr); + return NOTIFY_DONE; +} + +int nfp_app_start(struct nfp_app *app, struct nfp_net *ctrl) +{ + int err; + + app->ctrl = ctrl; + + if (app->type->start) { + err = app->type->start(app); + if (err) + return err; + } + + app->netdev_nb.notifier_call = nfp_app_netdev_event; + err = register_netdevice_notifier(&app->netdev_nb); + if (err) + goto err_app_stop; + + return 0; + +err_app_stop: + if (app->type->stop) + app->type->stop(app); + return err; +} + +void nfp_app_stop(struct nfp_app *app) +{ + unregister_netdevice_notifier(&app->netdev_nb); + + if (app->type->stop) + app->type->stop(app); +} + struct nfp_app *nfp_app_alloc(struct nfp_pf *pf, enum nfp_app_id id) { struct nfp_app *app; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_app.h b/drivers/net/ethernet/netronome/nfp/nfp_app.h index 4d6ecf99b1cc..f8d422713705 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_app.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_app.h @@ -69,6 +69,7 @@ extern const struct nfp_app_type app_abm; * @port_get_stats_strings: get strings for extra statistics * @start: start application logic * @stop: stop application logic + * @netdev_event: Netdevice notifier event * @ctrl_msg_rx: control message handler * @ctrl_msg_rx_raw: handler for control messages from data queues * @setup_tc: setup TC ndo @@ -122,6 +123,9 @@ struct nfp_app_type { int (*start)(struct nfp_app *app); void (*stop)(struct nfp_app *app); + int (*netdev_event)(struct nfp_app *app, struct net_device *netdev, + unsigned long event, void *ptr); + void (*ctrl_msg_rx)(struct nfp_app *app, struct sk_buff *skb); void (*ctrl_msg_rx_raw)(struct nfp_app *app, const void *data, unsigned int len); @@ -151,6 +155,7 @@ struct nfp_app_type { * @reprs: array of pointers to representors * @type: pointer to const application ops and info * @ctrl_mtu: MTU to set on the control vNIC (set in .init()) + * @netdev_nb: Netdevice notifier block * @priv: app-specific priv data */ struct nfp_app { @@ -163,6 +168,9 @@ struct nfp_app { const struct nfp_app_type *type; unsigned int ctrl_mtu; + + struct notifier_block netdev_nb; + void *priv; }; @@ -264,21 +272,6 @@ nfp_app_repr_change_mtu(struct nfp_app *app, struct net_device *netdev, return app->type->repr_change_mtu(app, netdev, new_mtu); } -static inline int nfp_app_start(struct nfp_app *app, struct nfp_net *ctrl) -{ - app->ctrl = ctrl; - if (!app->type->start) - return 0; - return app->type->start(app); -} - -static inline void nfp_app_stop(struct nfp_app *app) -{ - if (!app->type->stop) - return; - app->type->stop(app); -} - static inline const char *nfp_app_name(struct nfp_app *app) { if (!app) @@ -430,6 +423,8 @@ nfp_app_ctrl_msg_alloc(struct nfp_app *app, unsigned int size, gfp_t priority); struct nfp_app *nfp_app_alloc(struct nfp_pf *pf, enum nfp_app_id id); void nfp_app_free(struct nfp_app *app); +int nfp_app_start(struct nfp_app *app, struct nfp_net *ctrl); +void nfp_app_stop(struct nfp_app *app); /* Callbacks shared between apps */ @@ -438,4 +433,6 @@ int nfp_app_nic_vnic_alloc(struct nfp_app *app, struct nfp_net *nn, int nfp_app_nic_vnic_init_phy_port(struct nfp_pf *pf, struct nfp_app *app, struct nfp_net *nn, unsigned int id); +struct devlink *nfp_devlink_get_devlink(struct net_device *netdev); + #endif diff --git a/drivers/net/ethernet/netronome/nfp/nfp_devlink.c b/drivers/net/ethernet/netronome/nfp/nfp_devlink.c index 808647ec3573..e9eca99cf493 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_devlink.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_devlink.c @@ -4,6 +4,7 @@ #include <linux/rtnetlink.h> #include <net/devlink.h> +#include "nfpcore/nfp.h" #include "nfpcore/nfp_nsp.h" #include "nfp_app.h" #include "nfp_main.h" @@ -171,6 +172,173 @@ static int nfp_devlink_eswitch_mode_set(struct devlink *devlink, u16 mode, return ret; } +static const struct nfp_devlink_versions_simple { + const char *key; + const char *hwinfo; +} nfp_devlink_versions_hwinfo[] = { + { DEVLINK_INFO_VERSION_GENERIC_BOARD_ID, "assembly.partno", }, + { DEVLINK_INFO_VERSION_GENERIC_BOARD_REV, "assembly.revision", }, + { DEVLINK_INFO_VERSION_GENERIC_BOARD_MANUFACTURE, "assembly.vendor", }, + { "board.model", /* code name */ "assembly.model", }, +}; + +static int +nfp_devlink_versions_get_hwinfo(struct nfp_pf *pf, struct devlink_info_req *req) +{ + unsigned int i; + int err; + + for (i = 0; i < ARRAY_SIZE(nfp_devlink_versions_hwinfo); i++) { + const struct nfp_devlink_versions_simple *info; + const char *val; + + info = &nfp_devlink_versions_hwinfo[i]; + + val = nfp_hwinfo_lookup(pf->hwinfo, info->hwinfo); + if (!val) + continue; + + err = devlink_info_version_fixed_put(req, info->key, val); + if (err) + return err; + } + + return 0; +} + +static const struct nfp_devlink_versions { + enum nfp_nsp_versions id; + const char *key; +} nfp_devlink_versions_nsp[] = { + { NFP_VERSIONS_BUNDLE, "fw.bundle_id", }, + { NFP_VERSIONS_BSP, DEVLINK_INFO_VERSION_GENERIC_FW_MGMT, }, + { NFP_VERSIONS_CPLD, "fw.cpld", }, + { NFP_VERSIONS_APP, DEVLINK_INFO_VERSION_GENERIC_FW_APP, }, + { NFP_VERSIONS_UNDI, DEVLINK_INFO_VERSION_GENERIC_FW_UNDI, }, + { NFP_VERSIONS_NCSI, DEVLINK_INFO_VERSION_GENERIC_FW_NCSI, }, + { NFP_VERSIONS_CFGR, "chip.init", }, +}; + +static int +nfp_devlink_versions_get_nsp(struct devlink_info_req *req, bool flash, + const u8 *buf, unsigned int size) +{ + unsigned int i; + int err; + + for (i = 0; i < ARRAY_SIZE(nfp_devlink_versions_nsp); i++) { + const struct nfp_devlink_versions *info; + const char *version; + + info = &nfp_devlink_versions_nsp[i]; + + version = nfp_nsp_versions_get(info->id, flash, buf, size); + if (IS_ERR(version)) { + if (PTR_ERR(version) == -ENOENT) + continue; + else + return PTR_ERR(version); + } + + if (flash) + err = devlink_info_version_stored_put(req, info->key, + version); + else + err = devlink_info_version_running_put(req, info->key, + version); + if (err) + return err; + } + + return 0; +} + +static int +nfp_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req, + struct netlink_ext_ack *extack) +{ + struct nfp_pf *pf = devlink_priv(devlink); + const char *sn, *vendor, *part; + struct nfp_nsp *nsp; + char *buf = NULL; + int err; + + err = devlink_info_driver_name_put(req, "nfp"); + if (err) + return err; + + vendor = nfp_hwinfo_lookup(pf->hwinfo, "assembly.vendor"); + part = nfp_hwinfo_lookup(pf->hwinfo, "assembly.partno"); + sn = nfp_hwinfo_lookup(pf->hwinfo, "assembly.serial"); + if (vendor && part && sn) { + char *buf; + + buf = kmalloc(strlen(vendor) + strlen(part) + strlen(sn) + 1, + GFP_KERNEL); + if (!buf) + return -ENOMEM; + + buf[0] = '\0'; + strcat(buf, vendor); + strcat(buf, part); + strcat(buf, sn); + + err = devlink_info_serial_number_put(req, buf); + kfree(buf); + if (err) + return err; + } + + nsp = nfp_nsp_open(pf->cpp); + if (IS_ERR(nsp)) { + NL_SET_ERR_MSG_MOD(extack, "can't access NSP"); + return PTR_ERR(nsp); + } + + if (nfp_nsp_has_versions(nsp)) { + buf = kzalloc(NFP_NSP_VERSION_BUFSZ, GFP_KERNEL); + if (!buf) { + err = -ENOMEM; + goto err_close_nsp; + } + + err = nfp_nsp_versions(nsp, buf, NFP_NSP_VERSION_BUFSZ); + if (err) + goto err_free_buf; + + err = nfp_devlink_versions_get_nsp(req, false, + buf, NFP_NSP_VERSION_BUFSZ); + if (err) + goto err_free_buf; + + err = nfp_devlink_versions_get_nsp(req, true, + buf, NFP_NSP_VERSION_BUFSZ); + if (err) + goto err_free_buf; + + kfree(buf); + } + + nfp_nsp_close(nsp); + + return nfp_devlink_versions_get_hwinfo(pf, req); + +err_free_buf: + kfree(buf); +err_close_nsp: + nfp_nsp_close(nsp); + return err; +} + +static int +nfp_devlink_flash_update(struct devlink *devlink, const char *path, + const char *component, struct netlink_ext_ack *extack) +{ + if (component) + return -EOPNOTSUPP; + return nfp_flash_update_common(devlink_priv(devlink), path, extack); +} + const struct devlink_ops nfp_devlink_ops = { .port_split = nfp_devlink_port_split, .port_unsplit = nfp_devlink_port_unsplit, @@ -178,6 +346,8 @@ const struct devlink_ops nfp_devlink_ops = { .sb_pool_set = nfp_devlink_sb_pool_set, .eswitch_mode_get = nfp_devlink_eswitch_mode_get, .eswitch_mode_set = nfp_devlink_eswitch_mode_set, + .info_get = nfp_devlink_info_get, + .flash_update = nfp_devlink_flash_update, }; int nfp_devlink_port_register(struct nfp_app *app, struct nfp_port *port) @@ -206,3 +376,14 @@ void nfp_devlink_port_unregister(struct nfp_port *port) { devlink_port_unregister(&port->dl_port); } + +struct devlink *nfp_devlink_get_devlink(struct net_device *netdev) +{ + struct nfp_app *app; + + app = nfp_app_from_netdev(netdev); + if (!app) + return NULL; + + return priv_to_devlink(app->pf); +} diff --git a/drivers/net/ethernet/netronome/nfp/nfp_main.c b/drivers/net/ethernet/netronome/nfp/nfp_main.c index 6c10e8d119e4..f4c8776e42b6 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_main.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_main.c @@ -300,6 +300,47 @@ static int nfp_pcie_sriov_configure(struct pci_dev *pdev, int num_vfs) return nfp_pcie_sriov_enable(pdev, num_vfs); } +int nfp_flash_update_common(struct nfp_pf *pf, const char *path, + struct netlink_ext_ack *extack) +{ + struct device *dev = &pf->pdev->dev; + const struct firmware *fw; + struct nfp_nsp *nsp; + int err; + + nsp = nfp_nsp_open(pf->cpp); + if (IS_ERR(nsp)) { + err = PTR_ERR(nsp); + if (extack) + NL_SET_ERR_MSG_MOD(extack, "can't access NSP"); + else + dev_err(dev, "Failed to access the NSP: %d\n", err); + return err; + } + + err = request_firmware_direct(&fw, path, dev); + if (err) { + NL_SET_ERR_MSG_MOD(extack, + "unable to read flash file from disk"); + goto exit_close_nsp; + } + + dev_info(dev, "Please be patient while writing flash image: %s\n", + path); + + err = nfp_nsp_write_flash(nsp, fw); + if (err < 0) + goto exit_release_fw; + dev_info(dev, "Finished writing flash image\n"); + err = 0; + +exit_release_fw: + release_firmware(fw); +exit_close_nsp: + nfp_nsp_close(nsp); + return err; +} + static const struct firmware * nfp_net_fw_request(struct pci_dev *pdev, struct nfp_pf *pf, const char *name) { diff --git a/drivers/net/ethernet/netronome/nfp/nfp_main.h b/drivers/net/ethernet/netronome/nfp/nfp_main.h index a3613a2e0aa5..b7211f200d22 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_main.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_main.h @@ -164,6 +164,8 @@ nfp_pf_map_rtsym(struct nfp_pf *pf, const char *name, const char *sym_fmt, unsigned int min_size, struct nfp_cpp_area **area); int nfp_mbox_cmd(struct nfp_pf *pf, u32 cmd, void *in_data, u64 in_length, void *out_data, u64 out_length); +int nfp_flash_update_common(struct nfp_pf *pf, const char *path, + struct netlink_ext_ack *extack); enum nfp_dump_diag { NFP_DUMP_NSP_DIAG = 0, diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net.h b/drivers/net/ethernet/netronome/nfp/nfp_net.h index 6f0c37d09256..be37c2d6151c 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_net.h @@ -158,6 +158,7 @@ struct nfp_net_tx_desc { __le16 data_len; /* Length of frame + meta data */ } __packed; __le32 vals[4]; + __le64 vals8[2]; }; }; @@ -543,6 +544,7 @@ struct nfp_net_dp { * @reconfig_timer_active: Timer for reading reconfiguration results is pending * @reconfig_sync_present: Some thread is performing synchronous reconfig * @reconfig_timer: Timer for async reading of reconfig results + * @reconfig_in_progress_update: Update FW is processing now (debug only) * @link_up: Is the link up? * @link_status_lock: Protects @link_* and ensures atomicity with BAR reading * @rx_coalesce_usecs: RX interrupt moderation usecs delay parameter @@ -611,6 +613,7 @@ struct nfp_net { bool reconfig_timer_active; bool reconfig_sync_present; struct timer_list reconfig_timer; + u32 reconfig_in_progress_update; u32 rx_coalesce_usecs; u32 rx_coalesce_max_frames; @@ -851,7 +854,7 @@ void nfp_net_get_fw_version(struct nfp_net_fw_version *fw_ver, void __iomem *ctrl_bar); struct nfp_net * -nfp_net_alloc(struct pci_dev *pdev, bool needs_netdev, +nfp_net_alloc(struct pci_dev *pdev, void __iomem *ctrl_bar, bool needs_netdev, unsigned int max_tx_rings, unsigned int max_rx_rings); void nfp_net_free(struct nfp_net *nn); @@ -868,6 +871,7 @@ unsigned int nfp_net_rss_key_sz(struct nfp_net *nn); void nfp_net_rss_write_itbl(struct nfp_net *nn); void nfp_net_rss_write_key(struct nfp_net *nn); void nfp_net_coalesce_write_cfg(struct nfp_net *nn); +int nfp_net_reconfig_mbox(struct nfp_net *nn, u32 mbox_cmd); unsigned int nfp_net_irqs_alloc(struct pci_dev *pdev, struct msix_entry *irq_entries, diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c index 6bddfcfdec34..6d1b8816552e 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c @@ -36,7 +36,6 @@ #include <linux/vmalloc.h> #include <linux/ktime.h> -#include <net/switchdev.h> #include <net/vxlan.h> #include "nfpcore/nfp_nsp.h" @@ -101,6 +100,7 @@ static void nfp_net_reconfig_start(struct nfp_net *nn, u32 update) /* ensure update is written before pinging HW */ nn_pci_flush(nn); nfp_qcp_wr_ptr_add(nn->qcp_cfg, 1); + nn->reconfig_in_progress_update = update; } /* Pass 0 as update to run posted reconfigs. */ @@ -123,10 +123,14 @@ static bool nfp_net_reconfig_check_done(struct nfp_net *nn, bool last_check) if (reg == 0) return true; if (reg & NFP_NET_CFG_UPDATE_ERR) { - nn_err(nn, "Reconfig error: 0x%08x\n", reg); + nn_err(nn, "Reconfig error (status: 0x%08x update: 0x%08x ctrl: 0x%08x)\n", + reg, nn->reconfig_in_progress_update, + nn_readl(nn, NFP_NET_CFG_CTRL)); return true; } else if (last_check) { - nn_err(nn, "Reconfig timeout: 0x%08x\n", reg); + nn_err(nn, "Reconfig timeout (status: 0x%08x update: 0x%08x ctrl: 0x%08x)\n", + reg, nn->reconfig_in_progress_update, + nn_readl(nn, NFP_NET_CFG_CTRL)); return true; } @@ -279,7 +283,7 @@ int nfp_net_reconfig(struct nfp_net *nn, u32 update) * * Return: Negative errno on error, 0 on success */ -static int nfp_net_reconfig_mbox(struct nfp_net *nn, u32 mbox_cmd) +int nfp_net_reconfig_mbox(struct nfp_net *nn, u32 mbox_cmd) { u32 mbox = nn->tlv_caps.mbox_off; int ret; @@ -647,27 +651,29 @@ static void nfp_net_tx_ring_stop(struct netdev_queue *nd_q, * @txbuf: Pointer to driver soft TX descriptor * @txd: Pointer to HW TX descriptor * @skb: Pointer to SKB + * @md_bytes: Prepend length * * Set up Tx descriptor for LSO, do nothing for non-LSO skbs. * Return error on packet header greater than maximum supported LSO header size. */ static void nfp_net_tx_tso(struct nfp_net_r_vector *r_vec, struct nfp_net_tx_buf *txbuf, - struct nfp_net_tx_desc *txd, struct sk_buff *skb) + struct nfp_net_tx_desc *txd, struct sk_buff *skb, + u32 md_bytes) { - u32 hdrlen; + u32 l3_offset, l4_offset, hdrlen; u16 mss; if (!skb_is_gso(skb)) return; if (!skb->encapsulation) { - txd->l3_offset = skb_network_offset(skb); - txd->l4_offset = skb_transport_offset(skb); + l3_offset = skb_network_offset(skb); + l4_offset = skb_transport_offset(skb); hdrlen = skb_transport_offset(skb) + tcp_hdrlen(skb); } else { - txd->l3_offset = skb_inner_network_offset(skb); - txd->l4_offset = skb_inner_transport_offset(skb); + l3_offset = skb_inner_network_offset(skb); + l4_offset = skb_inner_transport_offset(skb); hdrlen = skb_inner_transport_header(skb) - skb->data + inner_tcp_hdrlen(skb); } @@ -676,7 +682,9 @@ static void nfp_net_tx_tso(struct nfp_net_r_vector *r_vec, txbuf->real_len += hdrlen * (txbuf->pkt_cnt - 1); mss = skb_shinfo(skb)->gso_size & PCIE_DESC_TX_MSS_MASK; - txd->lso_hdrlen = hdrlen; + txd->l3_offset = l3_offset - md_bytes; + txd->l4_offset = l4_offset - md_bytes; + txd->lso_hdrlen = hdrlen - md_bytes; txd->mss = cpu_to_le16(mss); txd->flags |= PCIE_DESC_TX_LSO; @@ -786,11 +794,11 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev) { struct nfp_net *nn = netdev_priv(netdev); const struct skb_frag_struct *frag; - struct nfp_net_tx_desc *txd, txdg; int f, nr_frags, wr_idx, md_bytes; struct nfp_net_tx_ring *tx_ring; struct nfp_net_r_vector *r_vec; struct nfp_net_tx_buf *txbuf; + struct nfp_net_tx_desc *txd; struct netdev_queue *nd_q; struct nfp_net_dp *dp; dma_addr_t dma_addr; @@ -801,13 +809,13 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev) qidx = skb_get_queue_mapping(skb); tx_ring = &dp->tx_rings[qidx]; r_vec = tx_ring->r_vec; - nd_q = netdev_get_tx_queue(dp->netdev, qidx); nr_frags = skb_shinfo(skb)->nr_frags; if (unlikely(nfp_net_tx_full(tx_ring, nr_frags + 1))) { nn_dp_warn(dp, "TX ring %d busy. wrp=%u rdp=%u\n", qidx, tx_ring->wr_p, tx_ring->rd_p); + nd_q = netdev_get_tx_queue(dp->netdev, qidx); netif_tx_stop_queue(nd_q); nfp_net_tx_xmit_more_flush(tx_ring); u64_stats_update_begin(&r_vec->tx_sync); @@ -851,7 +859,7 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev) txd->lso_hdrlen = 0; /* Do not reorder - tso may adjust pkt cnt, vlan may override fields */ - nfp_net_tx_tso(r_vec, txbuf, txd, skb); + nfp_net_tx_tso(r_vec, txbuf, txd, skb, md_bytes); nfp_net_tx_csum(dp, r_vec, txbuf, txd, skb); if (skb_vlan_tag_present(skb) && dp->ctrl & NFP_NET_CFG_CTRL_TXVLAN) { txd->flags |= PCIE_DESC_TX_VLAN; @@ -860,8 +868,10 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev) /* Gather DMA */ if (nr_frags > 0) { + __le64 second_half; + /* all descs must match except for in addr, length and eop */ - txdg = *txd; + second_half = txd->vals8[1]; for (f = 0; f < nr_frags; f++) { frag = &skb_shinfo(skb)->frags[f]; @@ -878,11 +888,11 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev) tx_ring->txbufs[wr_idx].fidx = f; txd = &tx_ring->txds[wr_idx]; - *txd = txdg; txd->dma_len = cpu_to_le16(fsize); nfp_desc_set_dma_addr(txd, dma_addr); - txd->offset_eop |= - (f == nr_frags - 1) ? PCIE_DESC_TX_EOP : 0; + txd->offset_eop = md_bytes | + ((f == nr_frags - 1) ? PCIE_DESC_TX_EOP : 0); + txd->vals8[1] = second_half; } u64_stats_update_begin(&r_vec->tx_sync); @@ -890,16 +900,16 @@ static int nfp_net_tx(struct sk_buff *skb, struct net_device *netdev) u64_stats_update_end(&r_vec->tx_sync); } - netdev_tx_sent_queue(nd_q, txbuf->real_len); - skb_tx_timestamp(skb); + nd_q = netdev_get_tx_queue(dp->netdev, tx_ring->idx); + tx_ring->wr_p += nr_frags + 1; if (nfp_net_tx_ring_should_stop(tx_ring)) nfp_net_tx_ring_stop(nd_q, tx_ring); tx_ring->wr_ptr_add += nr_frags + 1; - if (!skb->xmit_more || netif_xmit_stopped(nd_q)) + if (__netdev_tx_sent_queue(nd_q, txbuf->real_len, skb->xmit_more)) nfp_net_tx_xmit_more_flush(tx_ring); return NETDEV_TX_OK; @@ -940,14 +950,10 @@ static void nfp_net_tx_complete(struct nfp_net_tx_ring *tx_ring, int budget) { struct nfp_net_r_vector *r_vec = tx_ring->r_vec; struct nfp_net_dp *dp = &r_vec->nfp_net->dp; - const struct skb_frag_struct *frag; struct netdev_queue *nd_q; u32 done_pkts = 0, done_bytes = 0; - struct sk_buff *skb; - int todo, nr_frags; u32 qcp_rd_p; - int fidx; - int idx; + int todo; if (tx_ring->wr_p == tx_ring->rd_p) return; @@ -961,26 +967,33 @@ static void nfp_net_tx_complete(struct nfp_net_tx_ring *tx_ring, int budget) todo = D_IDX(tx_ring, qcp_rd_p - tx_ring->qcp_rd_p); while (todo--) { + const struct skb_frag_struct *frag; + struct nfp_net_tx_buf *tx_buf; + struct sk_buff *skb; + int fidx, nr_frags; + int idx; + idx = D_IDX(tx_ring, tx_ring->rd_p++); + tx_buf = &tx_ring->txbufs[idx]; - skb = tx_ring->txbufs[idx].skb; + skb = tx_buf->skb; if (!skb) continue; nr_frags = skb_shinfo(skb)->nr_frags; - fidx = tx_ring->txbufs[idx].fidx; + fidx = tx_buf->fidx; if (fidx == -1) { /* unmap head */ - dma_unmap_single(dp->dev, tx_ring->txbufs[idx].dma_addr, + dma_unmap_single(dp->dev, tx_buf->dma_addr, skb_headlen(skb), DMA_TO_DEVICE); - done_pkts += tx_ring->txbufs[idx].pkt_cnt; - done_bytes += tx_ring->txbufs[idx].real_len; + done_pkts += tx_buf->pkt_cnt; + done_bytes += tx_buf->real_len; } else { /* unmap fragment */ frag = &skb_shinfo(skb)->frags[fidx]; - dma_unmap_page(dp->dev, tx_ring->txbufs[idx].dma_addr, + dma_unmap_page(dp->dev, tx_buf->dma_addr, skb_frag_size(frag), DMA_TO_DEVICE); } @@ -988,9 +1001,9 @@ static void nfp_net_tx_complete(struct nfp_net_tx_ring *tx_ring, int budget) if (fidx == nr_frags - 1) napi_consume_skb(skb, budget); - tx_ring->txbufs[idx].dma_addr = 0; - tx_ring->txbufs[idx].skb = NULL; - tx_ring->txbufs[idx].fidx = -2; + tx_buf->dma_addr = 0; + tx_buf->skb = NULL; + tx_buf->fidx = -2; } tx_ring->qcp_rd_p = qcp_rd_p; @@ -2156,9 +2169,9 @@ nfp_net_tx_ring_alloc(struct nfp_net_dp *dp, struct nfp_net_tx_ring *tx_ring) tx_ring->cnt = dp->txd_cnt; tx_ring->size = array_size(tx_ring->cnt, sizeof(*tx_ring->txds)); - tx_ring->txds = dma_zalloc_coherent(dp->dev, tx_ring->size, - &tx_ring->dma, - GFP_KERNEL | __GFP_NOWARN); + tx_ring->txds = dma_alloc_coherent(dp->dev, tx_ring->size, + &tx_ring->dma, + GFP_KERNEL | __GFP_NOWARN); if (!tx_ring->txds) { netdev_warn(dp->netdev, "failed to allocate TX descriptor ring memory, requested descriptor count: %d, consider lowering descriptor count\n", tx_ring->cnt); @@ -2314,9 +2327,9 @@ nfp_net_rx_ring_alloc(struct nfp_net_dp *dp, struct nfp_net_rx_ring *rx_ring) rx_ring->cnt = dp->rxd_cnt; rx_ring->size = array_size(rx_ring->cnt, sizeof(*rx_ring->rxds)); - rx_ring->rxds = dma_zalloc_coherent(dp->dev, rx_ring->size, - &rx_ring->dma, - GFP_KERNEL | __GFP_NOWARN); + rx_ring->rxds = dma_alloc_coherent(dp->dev, rx_ring->size, + &rx_ring->dma, + GFP_KERNEL | __GFP_NOWARN); if (!rx_ring->rxds) { netdev_warn(dp->netdev, "failed to allocate RX descriptor ring memory, requested descriptor count: %d, consider lowering descriptor count\n", rx_ring->cnt); @@ -3275,7 +3288,10 @@ nfp_net_features_check(struct sk_buff *skb, struct net_device *dev, hdrlen = skb_inner_transport_header(skb) - skb->data + inner_tcp_hdrlen(skb); - if (unlikely(hdrlen > NFP_NET_LSO_MAX_HDR_SZ)) + /* Assume worst case scenario of having longest possible + * metadata prepend - 8B + */ + if (unlikely(hdrlen > NFP_NET_LSO_MAX_HDR_SZ - 8)) features &= ~NETIF_F_GSO_MASK; } @@ -3514,6 +3530,8 @@ const struct net_device_ops nfp_net_netdev_ops = { .ndo_udp_tunnel_add = nfp_net_add_vxlan_port, .ndo_udp_tunnel_del = nfp_net_del_vxlan_port, .ndo_bpf = nfp_net_xdp, + .ndo_get_port_parent_id = nfp_port_get_port_parent_id, + .ndo_get_devlink = nfp_devlink_get_devlink, }; /** @@ -3560,6 +3578,7 @@ void nfp_net_info(struct nfp_net *nn) /** * nfp_net_alloc() - Allocate netdev and related structure * @pdev: PCI device + * @ctrl_bar: PCI IOMEM with vNIC config memory * @needs_netdev: Whether to allocate a netdev for this vNIC * @max_tx_rings: Maximum number of TX rings supported by device * @max_rx_rings: Maximum number of RX rings supported by device @@ -3570,11 +3589,12 @@ void nfp_net_info(struct nfp_net *nn) * * Return: NFP Net device structure, or ERR_PTR on error. */ -struct nfp_net *nfp_net_alloc(struct pci_dev *pdev, bool needs_netdev, - unsigned int max_tx_rings, - unsigned int max_rx_rings) +struct nfp_net * +nfp_net_alloc(struct pci_dev *pdev, void __iomem *ctrl_bar, bool needs_netdev, + unsigned int max_tx_rings, unsigned int max_rx_rings) { struct nfp_net *nn; + int err; if (needs_netdev) { struct net_device *netdev; @@ -3594,6 +3614,7 @@ struct nfp_net *nfp_net_alloc(struct pci_dev *pdev, bool needs_netdev, } nn->dp.dev = &pdev->dev; + nn->dp.ctrl_bar = ctrl_bar; nn->pdev = pdev; nn->max_tx_rings = max_tx_rings; @@ -3616,7 +3637,19 @@ struct nfp_net *nfp_net_alloc(struct pci_dev *pdev, bool needs_netdev, timer_setup(&nn->reconfig_timer, nfp_net_reconfig_timer, 0); + err = nfp_net_tlv_caps_parse(&nn->pdev->dev, nn->dp.ctrl_bar, + &nn->tlv_caps); + if (err) + goto err_free_nn; + return nn; + +err_free_nn: + if (nn->dp.netdev) + free_netdev(nn->dp.netdev); + else + vfree(nn); + return ERR_PTR(err); } /** @@ -3783,8 +3816,6 @@ static void nfp_net_netdev_init(struct nfp_net *nn) netdev->netdev_ops = &nfp_net_netdev_ops; netdev->watchdog_timeo = msecs_to_jiffies(5 * 1000); - SWITCHDEV_SET_OPS(netdev, &nfp_port_switchdev_ops); - /* MTU range: 68 - hw-specific max */ netdev->min_mtu = ETH_MIN_MTU; netdev->max_mtu = nn->max_mtu; @@ -3889,11 +3920,6 @@ int nfp_net_init(struct nfp_net *nn) nn->dp.ctrl |= NFP_NET_CFG_CTRL_IRQMOD; } - err = nfp_net_tlv_caps_parse(&nn->pdev->dev, nn->dp.ctrl_bar, - &nn->tlv_caps); - if (err) - return err; - if (nn->dp.netdev) nfp_net_netdev_init(nn); diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c index f2aaef976c7d..6d5213b5bcb0 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c @@ -41,8 +41,8 @@ int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, data += 4; if (length % NFP_NET_CFG_TLV_LENGTH_INC) { - dev_err(dev, "TLV size not multiple of %u len:%u\n", - NFP_NET_CFG_TLV_LENGTH_INC, length); + dev_err(dev, "TLV size not multiple of %u offset:%u len:%u\n", + NFP_NET_CFG_TLV_LENGTH_INC, offset, length); return -EINVAL; } if (data + length > end) { @@ -61,14 +61,14 @@ int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, if (!length) return 0; - dev_err(dev, "END TLV should be empty, has len:%d\n", - length); + dev_err(dev, "END TLV should be empty, has offset:%u len:%d\n", + offset, length); return -EINVAL; case NFP_NET_CFG_TLV_TYPE_ME_FREQ: if (length != 4) { dev_err(dev, - "ME FREQ TLV should be 4B, is %dB\n", - length); + "ME FREQ TLV should be 4B, is %dB offset:%u\n", + length, offset); return -EINVAL; } @@ -90,6 +90,15 @@ int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, FIELD_GET(NFP_NET_CFG_TLV_HEADER_TYPE, hdr), offset, length); break; + case NFP_NET_CFG_TLV_TYPE_REPR_CAP: + if (length < 4) { + dev_err(dev, "REPR CAP TLV short %dB < 4B offset:%u\n", + length, offset); + return -EINVAL; + } + + caps->repr_cap = readl(data); + break; default: if (!FIELD_GET(NFP_NET_CFG_TLV_HEADER_REQUIRED, hdr)) break; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h index d7c8518ac952..372adea10e14 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h @@ -392,11 +392,13 @@ #define NFP_NET_CFG_MBOX_SIMPLE_CMD 0x0 #define NFP_NET_CFG_MBOX_SIMPLE_RET 0x4 #define NFP_NET_CFG_MBOX_SIMPLE_VAL 0x8 -#define NFP_NET_CFG_MBOX_SIMPLE_LEN 0x12 +#define NFP_NET_CFG_MBOX_SIMPLE_LEN 12 #define NFP_NET_CFG_MBOX_CMD_CTAG_FILTER_ADD 1 #define NFP_NET_CFG_MBOX_CMD_CTAG_FILTER_KILL 2 +#define NFP_NET_CFG_MBOX_CMD_PCI_DSCP_PRIOMAP_SET 5 + /** * VLAN filtering using general use mailbox * %NFP_NET_CFG_VLAN_FILTER: Base address of VLAN filter mailbox @@ -464,6 +466,10 @@ * Variable, experimental IDs. IDs designated for internal development and * experiments before a stable TLV ID has been allocated to a feature. Should * never be present in production firmware. + * + * %NFP_NET_CFG_TLV_TYPE_REPR_CAP: + * Single word, equivalent of %NFP_NET_CFG_CAP for representors, features which + * can be used on representors. */ #define NFP_NET_CFG_TLV_TYPE_UNKNOWN 0 #define NFP_NET_CFG_TLV_TYPE_RESERVED 1 @@ -472,6 +478,7 @@ #define NFP_NET_CFG_TLV_TYPE_MBOX 4 #define NFP_NET_CFG_TLV_TYPE_EXPERIMENTAL0 5 #define NFP_NET_CFG_TLV_TYPE_EXPERIMENTAL1 6 +#define NFP_NET_CFG_TLV_TYPE_REPR_CAP 7 struct device; @@ -480,11 +487,13 @@ struct device; * @me_freq_mhz: ME clock_freq (MHz) * @mbox_off: vNIC mailbox area offset * @mbox_len: vNIC mailbox area length + * @repr_cap: capabilities for representors */ struct nfp_net_tlv_caps { u32 me_freq_mhz; unsigned int mbox_off; unsigned int mbox_len; + u32 repr_cap; }; int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_debugfs.c b/drivers/net/ethernet/netronome/nfp/nfp_net_debugfs.c index 69b1c9b62e3d..ab7f2498e1c4 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_debugfs.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_debugfs.c @@ -8,7 +8,7 @@ static struct dentry *nfp_dir; -static int nfp_net_debugfs_rx_q_read(struct seq_file *file, void *data) +static int nfp_rx_q_show(struct seq_file *file, void *data) { struct nfp_net_r_vector *r_vec = file->private; struct nfp_net_rx_ring *rx_ring; @@ -65,31 +65,12 @@ out: rtnl_unlock(); return 0; } +DEFINE_SHOW_ATTRIBUTE(nfp_rx_q); -static int nfp_net_debugfs_rx_q_open(struct inode *inode, struct file *f) -{ - return single_open(f, nfp_net_debugfs_rx_q_read, inode->i_private); -} +static int nfp_tx_q_show(struct seq_file *file, void *data); +DEFINE_SHOW_ATTRIBUTE(nfp_tx_q); -static const struct file_operations nfp_rx_q_fops = { - .owner = THIS_MODULE, - .open = nfp_net_debugfs_rx_q_open, - .release = single_release, - .read = seq_read, - .llseek = seq_lseek -}; - -static int nfp_net_debugfs_tx_q_open(struct inode *inode, struct file *f); - -static const struct file_operations nfp_tx_q_fops = { - .owner = THIS_MODULE, - .open = nfp_net_debugfs_tx_q_open, - .release = single_release, - .read = seq_read, - .llseek = seq_lseek -}; - -static int nfp_net_debugfs_tx_q_read(struct seq_file *file, void *data) +static int nfp_tx_q_show(struct seq_file *file, void *data) { struct nfp_net_r_vector *r_vec = file->private; struct nfp_net_tx_ring *tx_ring; @@ -158,18 +139,11 @@ out: return 0; } -static int nfp_net_debugfs_tx_q_open(struct inode *inode, struct file *f) +static int nfp_xdp_q_show(struct seq_file *file, void *data) { - return single_open(f, nfp_net_debugfs_tx_q_read, inode->i_private); + return nfp_tx_q_show(file, data); } - -static const struct file_operations nfp_xdp_q_fops = { - .owner = THIS_MODULE, - .open = nfp_net_debugfs_tx_q_open, - .release = single_release, - .read = seq_read, - .llseek = seq_lseek -}; +DEFINE_SHOW_ATTRIBUTE(nfp_xdp_q); void nfp_net_debugfs_vnic_add(struct nfp_net *nn, struct dentry *ddir) { diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c b/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c index cb9c512abc76..690b62718dbb 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c @@ -1234,57 +1234,6 @@ static int nfp_net_set_channels(struct net_device *netdev, return nfp_net_set_num_rings(nn, total_rx, total_tx); } -static int -nfp_net_flash_device(struct net_device *netdev, struct ethtool_flash *flash) -{ - const struct firmware *fw; - struct nfp_app *app; - struct nfp_nsp *nsp; - struct device *dev; - int err; - - if (flash->region != ETHTOOL_FLASH_ALL_REGIONS) - return -EOPNOTSUPP; - - app = nfp_app_from_netdev(netdev); - if (!app) - return -EOPNOTSUPP; - - dev = &app->pdev->dev; - - nsp = nfp_nsp_open(app->cpp); - if (IS_ERR(nsp)) { - err = PTR_ERR(nsp); - dev_err(dev, "Failed to access the NSP: %d\n", err); - return err; - } - - err = request_firmware_direct(&fw, flash->data, dev); - if (err) - goto exit_close_nsp; - - dev_info(dev, "Please be patient while writing flash image: %s\n", - flash->data); - dev_hold(netdev); - rtnl_unlock(); - - err = nfp_nsp_write_flash(nsp, fw); - if (err < 0) { - dev_err(dev, "Flash write failed: %d\n", err); - goto exit_rtnl_lock; - } - dev_info(dev, "Finished writing flash image\n"); - -exit_rtnl_lock: - rtnl_lock(); - dev_put(netdev); - release_firmware(fw); - -exit_close_nsp: - nfp_nsp_close(nsp); - return err; -} - static const struct ethtool_ops nfp_net_ethtool_ops = { .get_drvinfo = nfp_net_get_drvinfo, .get_link = ethtool_op_get_link, @@ -1295,7 +1244,6 @@ static const struct ethtool_ops nfp_net_ethtool_ops = { .get_sset_count = nfp_net_get_sset_count, .get_rxnfc = nfp_net_get_rxnfc, .set_rxnfc = nfp_net_set_rxnfc, - .flash_device = nfp_net_flash_device, .get_rxfh_indir_size = nfp_net_get_rxfh_indir_size, .get_rxfh_key_size = nfp_net_get_rxfh_key_size, .get_rxfh = nfp_net_get_rxfh, @@ -1321,7 +1269,6 @@ const struct ethtool_ops nfp_port_ethtool_ops = { .get_strings = nfp_port_get_strings, .get_ethtool_stats = nfp_port_get_stats, .get_sset_count = nfp_port_get_sset_count, - .flash_device = nfp_net_flash_device, .set_dump = nfp_app_set_dump, .get_dump_flag = nfp_app_get_dump_flag, .get_dump_data = nfp_app_get_dump_data, diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_main.c b/drivers/net/ethernet/netronome/nfp/nfp_net_main.c index 1e7d20468a34..08f5fdbd8e41 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_main.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_main.c @@ -116,13 +116,13 @@ nfp_net_pf_alloc_vnic(struct nfp_pf *pf, bool needs_netdev, n_rx_rings = readl(ctrl_bar + NFP_NET_CFG_MAX_RXRINGS); /* Allocate and initialise the vNIC */ - nn = nfp_net_alloc(pf->pdev, needs_netdev, n_tx_rings, n_rx_rings); + nn = nfp_net_alloc(pf->pdev, ctrl_bar, needs_netdev, + n_tx_rings, n_rx_rings); if (IS_ERR(nn)) return nn; nn->app = pf->app; nfp_net_get_fw_version(&nn->fw_ver, ctrl_bar); - nn->dp.ctrl_bar = ctrl_bar; nn->tx_bar = qc_bar + tx_base * NFP_QCP_QUEUE_ADDR_SZ; nn->rx_bar = qc_bar + rx_base * NFP_QCP_QUEUE_ADDR_SZ; nn->dp.is_vf = 0; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_repr.c b/drivers/net/ethernet/netronome/nfp/nfp_net_repr.c index c09b893c30dd..94d228c04496 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_repr.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_repr.c @@ -5,12 +5,12 @@ #include <linux/io-64-nonatomic-hi-lo.h> #include <linux/lockdep.h> #include <net/dst_metadata.h> -#include <net/switchdev.h> #include "nfpcore/nfp_cpp.h" #include "nfpcore/nfp_nsp.h" #include "nfp_app.h" #include "nfp_main.h" +#include "nfp_net.h" #include "nfp_net_ctrl.h" #include "nfp_net_repr.h" #include "nfp_net_sriov.h" @@ -195,7 +195,7 @@ static netdev_tx_t nfp_repr_xmit(struct sk_buff *skb, struct net_device *netdev) ret = dev_queue_xmit(skb); nfp_repr_inc_tx_stats(netdev, len, ret); - return ret; + return NETDEV_TX_OK; } static int nfp_repr_stop(struct net_device *netdev) @@ -231,6 +231,27 @@ err_port_disable: return err; } +static netdev_features_t +nfp_repr_fix_features(struct net_device *netdev, netdev_features_t features) +{ + struct nfp_repr *repr = netdev_priv(netdev); + netdev_features_t old_features = features; + netdev_features_t lower_features; + struct net_device *lower_dev; + + lower_dev = repr->dst->u.port_info.lower_dev; + + lower_features = lower_dev->features; + if (lower_features & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)) + lower_features |= NETIF_F_HW_CSUM; + + features = netdev_intersect_features(features, lower_features); + features |= old_features & (NETIF_F_SOFT_FEATURES | NETIF_F_HW_TC); + features |= NETIF_F_LLTX; + + return features; +} + const struct net_device_ops nfp_repr_netdev_ops = { .ndo_init = nfp_app_ndo_init, .ndo_uninit = nfp_app_ndo_uninit, @@ -248,10 +269,27 @@ const struct net_device_ops nfp_repr_netdev_ops = { .ndo_set_vf_spoofchk = nfp_app_set_vf_spoofchk, .ndo_get_vf_config = nfp_app_get_vf_config, .ndo_set_vf_link_state = nfp_app_set_vf_link_state, + .ndo_fix_features = nfp_repr_fix_features, .ndo_set_features = nfp_port_set_features, .ndo_set_mac_address = eth_mac_addr, + .ndo_get_port_parent_id = nfp_port_get_port_parent_id, + .ndo_get_devlink = nfp_devlink_get_devlink, }; +void +nfp_repr_transfer_features(struct net_device *netdev, struct net_device *lower) +{ + struct nfp_repr *repr = netdev_priv(netdev); + + if (repr->dst->u.port_info.lower_dev != lower) + return; + + netdev->gso_max_size = lower->gso_max_size; + netdev->gso_max_segs = lower->gso_max_segs; + + netdev_update_features(netdev); +} + static void nfp_repr_clean(struct nfp_repr *repr) { unregister_netdev(repr->netdev); @@ -281,6 +319,8 @@ int nfp_repr_init(struct nfp_app *app, struct net_device *netdev, struct net_device *pf_netdev) { struct nfp_repr *repr = netdev_priv(netdev); + struct nfp_net *nn = netdev_priv(pf_netdev); + u32 repr_cap = nn->tlv_caps.repr_cap; int err; nfp_repr_set_lockdep_class(netdev); @@ -297,7 +337,54 @@ int nfp_repr_init(struct nfp_app *app, struct net_device *netdev, netdev->max_mtu = pf_netdev->max_mtu; - SWITCHDEV_SET_OPS(netdev, &nfp_port_switchdev_ops); + /* Set features the lower device can support with representors */ + if (repr_cap & NFP_NET_CFG_CTRL_LIVE_ADDR) + netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE; + + netdev->hw_features = NETIF_F_HIGHDMA; + if (repr_cap & NFP_NET_CFG_CTRL_RXCSUM_ANY) + netdev->hw_features |= NETIF_F_RXCSUM; + if (repr_cap & NFP_NET_CFG_CTRL_TXCSUM) + netdev->hw_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM; + if (repr_cap & NFP_NET_CFG_CTRL_GATHER) + netdev->hw_features |= NETIF_F_SG; + if ((repr_cap & NFP_NET_CFG_CTRL_LSO && nn->fw_ver.major > 2) || + repr_cap & NFP_NET_CFG_CTRL_LSO2) + netdev->hw_features |= NETIF_F_TSO | NETIF_F_TSO6; + if (repr_cap & NFP_NET_CFG_CTRL_RSS_ANY) + netdev->hw_features |= NETIF_F_RXHASH; + if (repr_cap & NFP_NET_CFG_CTRL_VXLAN) { + if (repr_cap & NFP_NET_CFG_CTRL_LSO) + netdev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; + } + if (repr_cap & NFP_NET_CFG_CTRL_NVGRE) { + if (repr_cap & NFP_NET_CFG_CTRL_LSO) + netdev->hw_features |= NETIF_F_GSO_GRE; + } + if (repr_cap & (NFP_NET_CFG_CTRL_VXLAN | NFP_NET_CFG_CTRL_NVGRE)) + netdev->hw_enc_features = netdev->hw_features; + + netdev->vlan_features = netdev->hw_features; + + if (repr_cap & NFP_NET_CFG_CTRL_RXVLAN) + netdev->hw_features |= NETIF_F_HW_VLAN_CTAG_RX; + if (repr_cap & NFP_NET_CFG_CTRL_TXVLAN) { + if (repr_cap & NFP_NET_CFG_CTRL_LSO2) + netdev_warn(netdev, "Device advertises both TSO2 and TXVLAN. Refusing to enable TXVLAN.\n"); + else + netdev->hw_features |= NETIF_F_HW_VLAN_CTAG_TX; + } + if (repr_cap & NFP_NET_CFG_CTRL_CTAG_FILTER) + netdev->hw_features |= NETIF_F_HW_VLAN_CTAG_FILTER; + + netdev->features = netdev->hw_features; + + /* Advertise but disable TSO by default. */ + netdev->features &= ~(NETIF_F_TSO | NETIF_F_TSO6); + netdev->gso_max_segs = NFP_NET_LSO_MAX_SEGS; + + netdev->priv_flags |= IFF_NO_QUEUE | IFF_DISABLE_NETPOLL; + netdev->features |= NETIF_F_LLTX; if (nfp_app_has_tc(app)) { netdev->features |= NETIF_F_HW_TC; @@ -442,7 +529,9 @@ int nfp_reprs_resync_phys_ports(struct nfp_app *app) continue; nfp_app_repr_preclean(app, netdev); + rtnl_lock(); rcu_assign_pointer(reprs->reprs[i], NULL); + rtnl_unlock(); synchronize_rcu(); nfp_repr_clean(repr); } diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_repr.h b/drivers/net/ethernet/netronome/nfp/nfp_net_repr.h index c412b94bfb97..e0f13dfe1f39 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_repr.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_repr.h @@ -92,6 +92,8 @@ nfp_repr_get_locked(struct nfp_app *app, struct nfp_reprs *set, unsigned int id); void nfp_repr_inc_rx_stats(struct net_device *netdev, unsigned int len); +void +nfp_repr_transfer_features(struct net_device *netdev, struct net_device *lower); int nfp_repr_init(struct nfp_app *app, struct net_device *netdev, u32 cmsg_port_id, struct nfp_port *port, struct net_device *pf_netdev); diff --git a/drivers/net/ethernet/netronome/nfp/nfp_netvf_main.c b/drivers/net/ethernet/netronome/nfp/nfp_netvf_main.c index d2c1e9ea5668..1145849ca7ba 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_netvf_main.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_netvf_main.c @@ -172,7 +172,7 @@ static int nfp_netvf_pci_probe(struct pci_dev *pdev, rx_bar_off = NFP_PCIE_QUEUE(startq); /* Allocate and initialise the netdev */ - nn = nfp_net_alloc(pdev, true, max_tx_rings, max_rx_rings); + nn = nfp_net_alloc(pdev, ctrl_bar, true, max_tx_rings, max_rx_rings); if (IS_ERR(nn)) { err = PTR_ERR(nn); goto err_ctrl_unmap; @@ -180,7 +180,6 @@ static int nfp_netvf_pci_probe(struct pci_dev *pdev, vf->nn = nn; nn->fw_ver = fw_ver; - nn->dp.ctrl_bar = ctrl_bar; nn->dp.is_vf = 1; nn->stride_tx = stride; nn->stride_rx = stride; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_port.c b/drivers/net/ethernet/netronome/nfp/nfp_port.c index 86bc149ca231..93c5bfc0510b 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_port.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_port.c @@ -3,7 +3,6 @@ #include <linux/lockdep.h> #include <linux/netdevice.h> -#include <net/switchdev.h> #include "nfpcore/nfp_cpp.h" #include "nfpcore/nfp_nsp.h" @@ -31,34 +30,22 @@ struct nfp_port *nfp_port_from_netdev(struct net_device *netdev) return NULL; } -static int -nfp_port_attr_get(struct net_device *netdev, struct switchdev_attr *attr) +int nfp_port_get_port_parent_id(struct net_device *netdev, + struct netdev_phys_item_id *ppid) { struct nfp_port *port; + const u8 *serial; port = nfp_port_from_netdev(netdev); if (!port) return -EOPNOTSUPP; - switch (attr->id) { - case SWITCHDEV_ATTR_ID_PORT_PARENT_ID: { - const u8 *serial; - /* N.B: attr->u.ppid.id is binary data */ - attr->u.ppid.id_len = nfp_cpp_serial(port->app->cpp, &serial); - memcpy(&attr->u.ppid.id, serial, attr->u.ppid.id_len); - break; - } - default: - return -EOPNOTSUPP; - } + ppid->id_len = nfp_cpp_serial(port->app->cpp, &serial); + memcpy(&ppid->id, serial, ppid->id_len); return 0; } -const struct switchdev_ops nfp_port_switchdev_ops = { - .switchdev_port_attr_get = nfp_port_attr_get, -}; - int nfp_port_setup_tc(struct net_device *netdev, enum tc_setup_type type, void *type_data) { diff --git a/drivers/net/ethernet/netronome/nfp/nfp_port.h b/drivers/net/ethernet/netronome/nfp/nfp_port.h index b2479a2a49e5..90ae053f5c07 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_port.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_port.h @@ -7,6 +7,7 @@ #include <net/devlink.h> struct net_device; +struct netdev_phys_item_id; struct nfp_app; struct nfp_pf; struct nfp_port; @@ -90,7 +91,6 @@ struct nfp_port { }; extern const struct ethtool_ops nfp_port_ethtool_ops; -extern const struct switchdev_ops nfp_port_switchdev_ops; __printf(2, 3) u8 *nfp_pr_et(u8 *data, const char *fmt, ...); @@ -106,6 +106,8 @@ int nfp_port_set_features(struct net_device *netdev, netdev_features_t features); struct nfp_port *nfp_port_from_netdev(struct net_device *netdev); +int nfp_port_get_port_parent_id(struct net_device *netdev, + struct netdev_phys_item_id *ppid); struct nfp_port * nfp_port_from_id(struct nfp_pf *pf, enum nfp_port_type type, unsigned int id); struct nfp_eth_table_port *__nfp_port_get_eth_port(struct nfp_port *port); diff --git a/drivers/net/ethernet/netronome/nfp/nfp_shared_buf.c b/drivers/net/ethernet/netronome/nfp/nfp_shared_buf.c index 814360ed3a20..ea2e3f829aba 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_shared_buf.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_shared_buf.c @@ -48,6 +48,7 @@ int nfp_shared_buf_pool_get(struct nfp_pf *pf, unsigned int sb, u16 pool_index, pool_info->pool_type = le32_to_cpu(get_data.pool_type); pool_info->threshold_type = le32_to_cpu(get_data.threshold_type); pool_info->size = le32_to_cpu(get_data.size) * unit_size; + pool_info->cell_size = unit_size; return 0; } diff --git a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.c b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.c index ce1577bbbd2a..3a4e224a64b7 100644 --- a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.c +++ b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.c @@ -7,11 +7,13 @@ * Jason McMullan <jason.mcmullan@netronome.com> */ +#include <asm/unaligned.h> #include <linux/bitfield.h> #include <linux/delay.h> #include <linux/firmware.h> #include <linux/kernel.h> #include <linux/kthread.h> +#include <linux/overflow.h> #include <linux/sizes.h> #include <linux/slab.h> @@ -36,6 +38,7 @@ #define NSP_COMMAND 0x08 #define NSP_COMMAND_OPTION GENMASK_ULL(63, 32) #define NSP_COMMAND_CODE GENMASK_ULL(31, 16) +#define NSP_COMMAND_DMA_BUF BIT_ULL(1) #define NSP_COMMAND_START BIT_ULL(0) /* CPP address to retrieve the data from */ @@ -48,8 +51,12 @@ #define NSP_DFLT_BUFFER_ADDRESS GENMASK_ULL(39, 0) #define NSP_DFLT_BUFFER_CONFIG 0x20 +#define NSP_DFLT_BUFFER_DMA_CHUNK_ORDER GENMASK_ULL(63, 58) +#define NSP_DFLT_BUFFER_SIZE_4KB GENMASK_ULL(15, 8) #define NSP_DFLT_BUFFER_SIZE_MB GENMASK_ULL(7, 0) +#define NFP_CAP_CMD_DMA_SG 0x28 + #define NSP_MAGIC 0xab10 #define NSP_MAJOR 0 #define NSP_MINOR 8 @@ -62,6 +69,16 @@ #define NFP_HWINFO_LOOKUP_SIZE GENMASK(11, 0) +#define NFP_VERSIONS_SIZE GENMASK(11, 0) +#define NFP_VERSIONS_CNT_OFF 0 +#define NFP_VERSIONS_BSP_OFF 2 +#define NFP_VERSIONS_CPLD_OFF 6 +#define NFP_VERSIONS_APP_OFF 10 +#define NFP_VERSIONS_BUNDLE_OFF 14 +#define NFP_VERSIONS_UNDI_OFF 18 +#define NFP_VERSIONS_NCSI_OFF 22 +#define NFP_VERSIONS_CFGR_OFF 26 + enum nfp_nsp_cmd { SPCODE_NOOP = 0, /* No operation */ SPCODE_SOFT_RESET = 1, /* Soft reset the NFP */ @@ -77,6 +94,17 @@ enum nfp_nsp_cmd { SPCODE_NSP_IDENTIFY = 13, /* Read NSP version */ SPCODE_FW_STORED = 16, /* If no FW loaded, load flash app FW */ SPCODE_HWINFO_LOOKUP = 17, /* Lookup HWinfo with overwrites etc. */ + SPCODE_VERSIONS = 21, /* Report FW versions */ +}; + +struct nfp_nsp_dma_buf { + __le32 chunk_cnt; + __le32 reserved[3]; + struct { + __le32 size; + __le32 reserved; + __le64 addr; + } descs[]; }; static const struct { @@ -107,18 +135,18 @@ struct nfp_nsp { /** * struct nfp_nsp_command_arg - NFP command argument structure * @code: NFP SP Command Code + * @dma: @buf points to a host buffer, not NSP buffer * @timeout_sec:Timeout value to wait for completion in seconds * @option: NFP SP Command Argument - * @buff_cpp: NFP SP Buffer CPP Address info - * @buff_addr: NFP SP Buffer Host address + * @buf: NFP SP Buffer Address * @error_cb: Callback for interpreting option if error occurred */ struct nfp_nsp_command_arg { u16 code; + bool dma; unsigned int timeout_sec; u32 option; - u32 buff_cpp; - u64 buff_addr; + u64 buf; void (*error_cb)(struct nfp_nsp *state, u32 ret_val); }; @@ -332,22 +360,14 @@ __nfp_nsp_command(struct nfp_nsp *state, const struct nfp_nsp_command_arg *arg) if (err) return err; - if (!FIELD_FIT(NSP_BUFFER_CPP, arg->buff_cpp >> 8) || - !FIELD_FIT(NSP_BUFFER_ADDRESS, arg->buff_addr)) { - nfp_err(cpp, "Host buffer out of reach %08x %016llx\n", - arg->buff_cpp, arg->buff_addr); - return -EINVAL; - } - - err = nfp_cpp_writeq(cpp, nsp_cpp, nsp_buffer, - FIELD_PREP(NSP_BUFFER_CPP, arg->buff_cpp >> 8) | - FIELD_PREP(NSP_BUFFER_ADDRESS, arg->buff_addr)); + err = nfp_cpp_writeq(cpp, nsp_cpp, nsp_buffer, arg->buf); if (err < 0) return err; err = nfp_cpp_writeq(cpp, nsp_cpp, nsp_command, FIELD_PREP(NSP_COMMAND_OPTION, arg->option) | FIELD_PREP(NSP_COMMAND_CODE, arg->code) | + FIELD_PREP(NSP_COMMAND_DMA_BUF, arg->dma) | FIELD_PREP(NSP_COMMAND_START, 1)); if (err < 0) return err; @@ -399,36 +419,14 @@ static int nfp_nsp_command(struct nfp_nsp *state, u16 code) } static int -nfp_nsp_command_buf(struct nfp_nsp *nsp, struct nfp_nsp_command_buf_arg *arg) +nfp_nsp_command_buf_def(struct nfp_nsp *nsp, + struct nfp_nsp_command_buf_arg *arg) { struct nfp_cpp *cpp = nsp->cpp; - unsigned int max_size; u64 reg, cpp_buf; - int ret, err; + int err, ret; u32 cpp_id; - if (nsp->ver.minor < 13) { - nfp_err(cpp, "NSP: Code 0x%04x with buffer not supported (ABI %hu.%hu)\n", - arg->arg.code, nsp->ver.major, nsp->ver.minor); - return -EOPNOTSUPP; - } - - err = nfp_cpp_readq(cpp, nfp_resource_cpp_id(nsp->res), - nfp_resource_address(nsp->res) + - NSP_DFLT_BUFFER_CONFIG, - ®); - if (err < 0) - return err; - - max_size = max(arg->in_size, arg->out_size); - if (FIELD_GET(NSP_DFLT_BUFFER_SIZE_MB, reg) * SZ_1M < max_size) { - nfp_err(cpp, "NSP: default buffer too small for command 0x%04x (%llu < %u)\n", - arg->arg.code, - FIELD_GET(NSP_DFLT_BUFFER_SIZE_MB, reg) * SZ_1M, - max_size); - return -EINVAL; - } - err = nfp_cpp_readq(cpp, nfp_resource_cpp_id(nsp->res), nfp_resource_address(nsp->res) + NSP_DFLT_BUFFER, @@ -447,15 +445,21 @@ nfp_nsp_command_buf(struct nfp_nsp *nsp, struct nfp_nsp_command_buf_arg *arg) } /* Zero out remaining part of the buffer */ if (arg->out_buf && arg->out_size && arg->out_size > arg->in_size) { - memset(arg->out_buf, 0, arg->out_size - arg->in_size); err = nfp_cpp_write(cpp, cpp_id, cpp_buf + arg->in_size, arg->out_buf, arg->out_size - arg->in_size); if (err < 0) return err; } - arg->arg.buff_cpp = cpp_id; - arg->arg.buff_addr = cpp_buf; + if (!FIELD_FIT(NSP_BUFFER_CPP, cpp_id >> 8) || + !FIELD_FIT(NSP_BUFFER_ADDRESS, cpp_buf)) { + nfp_err(cpp, "Buffer out of reach %08x %016llx\n", + cpp_id, cpp_buf); + return -EINVAL; + } + + arg->arg.buf = FIELD_PREP(NSP_BUFFER_CPP, cpp_id >> 8) | + FIELD_PREP(NSP_BUFFER_ADDRESS, cpp_buf); ret = __nfp_nsp_command(nsp, &arg->arg); if (ret < 0) return ret; @@ -470,6 +474,210 @@ nfp_nsp_command_buf(struct nfp_nsp *nsp, struct nfp_nsp_command_buf_arg *arg) return ret; } +static int +nfp_nsp_command_buf_dma_sg(struct nfp_nsp *nsp, + struct nfp_nsp_command_buf_arg *arg, + unsigned int max_size, unsigned int chunk_order, + unsigned int dma_order) +{ + struct nfp_cpp *cpp = nsp->cpp; + struct nfp_nsp_dma_buf *desc; + struct { + dma_addr_t dma_addr; + unsigned long len; + void *chunk; + } *chunks; + size_t chunk_size, dma_size; + dma_addr_t dma_desc; + struct device *dev; + unsigned long off; + int i, ret, nseg; + size_t desc_sz; + + chunk_size = BIT_ULL(chunk_order); + dma_size = BIT_ULL(dma_order); + nseg = DIV_ROUND_UP(max_size, chunk_size); + + chunks = kzalloc(array_size(sizeof(*chunks), nseg), GFP_KERNEL); + if (!chunks) + return -ENOMEM; + + off = 0; + ret = -ENOMEM; + for (i = 0; i < nseg; i++) { + unsigned long coff; + + chunks[i].chunk = kmalloc(chunk_size, + GFP_KERNEL | __GFP_NOWARN); + if (!chunks[i].chunk) + goto exit_free_prev; + + chunks[i].len = min_t(u64, chunk_size, max_size - off); + + coff = 0; + if (arg->in_size > off) { + coff = min_t(u64, arg->in_size - off, chunk_size); + memcpy(chunks[i].chunk, arg->in_buf + off, coff); + } + memset(chunks[i].chunk + coff, 0, chunk_size - coff); + + off += chunks[i].len; + } + + dev = nfp_cpp_device(cpp)->parent; + + for (i = 0; i < nseg; i++) { + dma_addr_t addr; + + addr = dma_map_single(dev, chunks[i].chunk, chunks[i].len, + DMA_BIDIRECTIONAL); + chunks[i].dma_addr = addr; + + ret = dma_mapping_error(dev, addr); + if (ret) + goto exit_unmap_prev; + + if (WARN_ONCE(round_down(addr, dma_size) != + round_down(addr + chunks[i].len - 1, dma_size), + "unaligned DMA address: %pad %lu %zd\n", + &addr, chunks[i].len, dma_size)) { + ret = -EFAULT; + i++; + goto exit_unmap_prev; + } + } + + desc_sz = struct_size(desc, descs, nseg); + desc = kmalloc(desc_sz, GFP_KERNEL); + if (!desc) { + ret = -ENOMEM; + goto exit_unmap_all; + } + + desc->chunk_cnt = cpu_to_le32(nseg); + for (i = 0; i < nseg; i++) { + desc->descs[i].size = cpu_to_le32(chunks[i].len); + desc->descs[i].addr = cpu_to_le64(chunks[i].dma_addr); + } + + dma_desc = dma_map_single(dev, desc, desc_sz, DMA_TO_DEVICE); + ret = dma_mapping_error(dev, dma_desc); + if (ret) + goto exit_free_desc; + + arg->arg.dma = true; + arg->arg.buf = dma_desc; + ret = __nfp_nsp_command(nsp, &arg->arg); + if (ret < 0) + goto exit_unmap_desc; + + i = 0; + off = 0; + while (off < arg->out_size) { + unsigned int len; + + len = min_t(u64, chunks[i].len, arg->out_size - off); + memcpy(arg->out_buf + off, chunks[i].chunk, len); + off += len; + i++; + } + +exit_unmap_desc: + dma_unmap_single(dev, dma_desc, desc_sz, DMA_TO_DEVICE); +exit_free_desc: + kfree(desc); +exit_unmap_all: + i = nseg; +exit_unmap_prev: + while (--i >= 0) + dma_unmap_single(dev, chunks[i].dma_addr, chunks[i].len, + DMA_BIDIRECTIONAL); + i = nseg; +exit_free_prev: + while (--i >= 0) + kfree(chunks[i].chunk); + kfree(chunks); + if (ret < 0) + nfp_err(cpp, "NSP: SG DMA failed for command 0x%04x: %d (sz:%d cord:%d)\n", + arg->arg.code, ret, max_size, chunk_order); + return ret; +} + +static int +nfp_nsp_command_buf_dma(struct nfp_nsp *nsp, + struct nfp_nsp_command_buf_arg *arg, + unsigned int max_size, unsigned int dma_order) +{ + unsigned int chunk_order, buf_order; + struct nfp_cpp *cpp = nsp->cpp; + bool sg_ok; + u64 reg; + int err; + + buf_order = order_base_2(roundup_pow_of_two(max_size)); + + err = nfp_cpp_readq(cpp, nfp_resource_cpp_id(nsp->res), + nfp_resource_address(nsp->res) + NFP_CAP_CMD_DMA_SG, + ®); + if (err < 0) + return err; + sg_ok = reg & BIT_ULL(arg->arg.code - 1); + + if (!sg_ok) { + if (buf_order > dma_order) { + nfp_err(cpp, "NSP: can't service non-SG DMA for command 0x%04x\n", + arg->arg.code); + return -ENOMEM; + } + chunk_order = buf_order; + } else { + chunk_order = min_t(unsigned int, dma_order, PAGE_SHIFT); + } + + return nfp_nsp_command_buf_dma_sg(nsp, arg, max_size, chunk_order, + dma_order); +} + +static int +nfp_nsp_command_buf(struct nfp_nsp *nsp, struct nfp_nsp_command_buf_arg *arg) +{ + unsigned int dma_order, def_size, max_size; + struct nfp_cpp *cpp = nsp->cpp; + u64 reg; + int err; + + if (nsp->ver.minor < 13) { + nfp_err(cpp, "NSP: Code 0x%04x with buffer not supported (ABI %hu.%hu)\n", + arg->arg.code, nsp->ver.major, nsp->ver.minor); + return -EOPNOTSUPP; + } + + err = nfp_cpp_readq(cpp, nfp_resource_cpp_id(nsp->res), + nfp_resource_address(nsp->res) + + NSP_DFLT_BUFFER_CONFIG, + ®); + if (err < 0) + return err; + + /* Zero out undefined part of the out buffer */ + if (arg->out_buf && arg->out_size && arg->out_size > arg->in_size) + memset(arg->out_buf, 0, arg->out_size - arg->in_size); + + max_size = max(arg->in_size, arg->out_size); + def_size = FIELD_GET(NSP_DFLT_BUFFER_SIZE_MB, reg) * SZ_1M + + FIELD_GET(NSP_DFLT_BUFFER_SIZE_4KB, reg) * SZ_4K; + dma_order = FIELD_GET(NSP_DFLT_BUFFER_DMA_CHUNK_ORDER, reg); + if (def_size >= max_size) { + return nfp_nsp_command_buf_def(nsp, arg); + } else if (!dma_order) { + nfp_err(cpp, "NSP: default buffer too small for command 0x%04x (%u < %u)\n", + arg->arg.code, def_size, max_size); + return -EINVAL; + } + + return nfp_nsp_command_buf_dma(nsp, arg, max_size, dma_order); +} + int nfp_nsp_wait(struct nfp_nsp *state) { const unsigned long wait_until = jiffies + NFP_NSP_TIMEOUT_BOOT * HZ; @@ -591,10 +799,7 @@ int nfp_nsp_write_flash(struct nfp_nsp *state, const struct firmware *fw) { .code = SPCODE_NSP_WRITE_FLASH, .option = fw->size, - /* The flash time is specified to take a maximum of 70s - * so we add an additional factor to this spec time. - */ - .timeout_sec = 2.5 * 70, + .timeout_sec = 900, }, .in_buf = fw->data, .in_size = fw->size, @@ -711,3 +916,52 @@ int nfp_nsp_hwinfo_lookup(struct nfp_nsp *state, void *buf, unsigned int size) return 0; } + +int nfp_nsp_versions(struct nfp_nsp *state, void *buf, unsigned int size) +{ + struct nfp_nsp_command_buf_arg versions = { + { + .code = SPCODE_VERSIONS, + .option = min_t(u32, size, NFP_VERSIONS_SIZE), + }, + .out_buf = buf, + .out_size = min_t(u32, size, NFP_VERSIONS_SIZE), + }; + + return nfp_nsp_command_buf(state, &versions); +} + +const char *nfp_nsp_versions_get(enum nfp_nsp_versions id, bool flash, + const u8 *buf, unsigned int size) +{ + static const u32 id2off[] = { + [NFP_VERSIONS_BSP] = NFP_VERSIONS_BSP_OFF, + [NFP_VERSIONS_CPLD] = NFP_VERSIONS_CPLD_OFF, + [NFP_VERSIONS_APP] = NFP_VERSIONS_APP_OFF, + [NFP_VERSIONS_BUNDLE] = NFP_VERSIONS_BUNDLE_OFF, + [NFP_VERSIONS_UNDI] = NFP_VERSIONS_UNDI_OFF, + [NFP_VERSIONS_NCSI] = NFP_VERSIONS_NCSI_OFF, + [NFP_VERSIONS_CFGR] = NFP_VERSIONS_CFGR_OFF, + }; + unsigned int field, buf_field_cnt, buf_off; + + if (id >= ARRAY_SIZE(id2off) || !id2off[id]) + return ERR_PTR(-EINVAL); + + field = id * 2 + flash; + + buf_field_cnt = get_unaligned_le16(buf); + if (buf_field_cnt <= field) + return ERR_PTR(-ENOENT); + + buf_off = get_unaligned_le16(buf + id2off[id] + flash * 2); + if (!buf_off) + return ERR_PTR(-ENOENT); + + if (buf_off >= size) + return ERR_PTR(-EINVAL); + if (strnlen(&buf[buf_off], size - buf_off) == size - buf_off) + return ERR_PTR(-EINVAL); + + return (const char *)&buf[buf_off]; +} diff --git a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.h b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.h index ff33ac54097a..bd9c358c646f 100644 --- a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.h +++ b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp.h @@ -38,12 +38,18 @@ static inline bool nfp_nsp_has_hwinfo_lookup(struct nfp_nsp *state) return nfp_nsp_get_abi_ver_minor(state) > 24; } +static inline bool nfp_nsp_has_versions(struct nfp_nsp *state) +{ + return nfp_nsp_get_abi_ver_minor(state) > 27; +} + enum nfp_eth_interface { NFP_INTERFACE_NONE = 0, NFP_INTERFACE_SFP = 1, NFP_INTERFACE_SFPP = 10, NFP_INTERFACE_SFP28 = 28, NFP_INTERFACE_QSFP = 40, + NFP_INTERFACE_RJ45 = 45, NFP_INTERFACE_CXP = 100, NFP_INTERFACE_QSFP28 = 112, }; @@ -208,4 +214,19 @@ enum nfp_nsp_sensor_id { int nfp_hwmon_read_sensor(struct nfp_cpp *cpp, enum nfp_nsp_sensor_id id, long *val); +#define NFP_NSP_VERSION_BUFSZ 1024 /* reasonable size, not in the ABI */ + +enum nfp_nsp_versions { + NFP_VERSIONS_BSP, + NFP_VERSIONS_CPLD, + NFP_VERSIONS_APP, + NFP_VERSIONS_BUNDLE, + NFP_VERSIONS_UNDI, + NFP_VERSIONS_NCSI, + NFP_VERSIONS_CFGR, +}; + +int nfp_nsp_versions(struct nfp_nsp *state, void *buf, unsigned int size); +const char *nfp_nsp_versions_get(enum nfp_nsp_versions id, bool flash, + const u8 *buf, unsigned int size); #endif diff --git a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp_eth.c b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp_eth.c index 802c9224bb32..311a5be25acb 100644 --- a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp_eth.c +++ b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_nsp_eth.c @@ -206,6 +206,9 @@ nfp_eth_calc_port_type(struct nfp_cpp *cpp, struct nfp_eth_table_port *entry) if (entry->interface == NFP_INTERFACE_NONE) { entry->port_type = PORT_NONE; return; + } else if (entry->interface == NFP_INTERFACE_RJ45) { + entry->port_type = PORT_TP; + return; } if (entry->media == NFP_MEDIA_FIBRE) @@ -269,8 +272,7 @@ __nfp_eth_read_ports(struct nfp_cpp *cpp, struct nfp_nsp *nsp) goto err; } - table = kzalloc(sizeof(*table) + - sizeof(struct nfp_eth_table_port) * cnt, GFP_KERNEL); + table = kzalloc(struct_size(table, ports, cnt), GFP_KERNEL); if (!table) goto err; |