diff --git a/target/linux/generic/pending-5.15/731-net-ethernet-mediatek-ppe-add-support-for-flow-accou.patch b/target/linux/generic/pending-5.15/731-net-ethernet-mediatek-ppe-add-support-for-flow-accou.patch index d290e1360..8bf036d00 100644 --- a/target/linux/generic/pending-5.15/731-net-ethernet-mediatek-ppe-add-support-for-flow-accou.patch +++ b/target/linux/generic/pending-5.15/731-net-ethernet-mediatek-ppe-add-support-for-flow-accou.patch @@ -363,7 +363,7 @@ v2: fix wrong variable name in return value check spotted by Denis Kirjanov return 0; --- a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c +++ b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c -@@ -494,6 +494,7 @@ static int +@@ -491,6 +491,7 @@ static int mtk_flow_offload_stats(struct mtk_eth *eth, struct flow_cls_offload *f) { struct mtk_flow_entry *entry; @@ -371,7 +371,7 @@ v2: fix wrong variable name in return value check spotted by Denis Kirjanov u32 idle; entry = rhashtable_lookup(ð->flow_table, &f->cookie, -@@ -504,6 +505,12 @@ mtk_flow_offload_stats(struct mtk_eth *e +@@ -501,6 +502,12 @@ mtk_flow_offload_stats(struct mtk_eth *e idle = mtk_foe_entry_idle_time(eth->ppe[entry->ppe_index], entry); f->stats.lastused = jiffies - idle * HZ; diff --git a/target/linux/generic/pending-5.15/732-01-net-ethernet-mtk_eth_soc-account-for-vlan-in-rx-head.patch b/target/linux/generic/pending-5.15/732-01-net-ethernet-mtk_eth_soc-account-for-vlan-in-rx-head.patch new file mode 100644 index 000000000..45af898cf --- /dev/null +++ b/target/linux/generic/pending-5.15/732-01-net-ethernet-mtk_eth_soc-account-for-vlan-in-rx-head.patch @@ -0,0 +1,22 @@ +From: Felix Fietkau +Date: Thu, 27 Oct 2022 19:50:31 +0200 +Subject: [PATCH] net: ethernet: mtk_eth_soc: account for vlan in rx + header length + +The network stack assumes that devices can handle an extra VLAN tag without +increasing the MTU + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -29,7 +29,7 @@ + #define MTK_TX_DMA_BUF_LEN_V2 0xffff + #define MTK_DMA_SIZE 512 + #define MTK_MAC_COUNT 2 +-#define MTK_RX_ETH_HLEN (ETH_HLEN + ETH_FCS_LEN) ++#define MTK_RX_ETH_HLEN (VLAN_ETH_HLEN + ETH_FCS_LEN) + #define MTK_RX_HLEN (NET_SKB_PAD + MTK_RX_ETH_HLEN + NET_IP_ALIGN) + #define MTK_DMA_DUMMY_DESC 0xffffffff + #define MTK_DEFAULT_MSG_ENABLE (NETIF_MSG_DRV | \ diff --git a/target/linux/generic/pending-5.15/732-02-net-ethernet-mtk_eth_soc-increase-tx-ring-side-for-Q.patch b/target/linux/generic/pending-5.15/732-02-net-ethernet-mtk_eth_soc-increase-tx-ring-side-for-Q.patch new file mode 100644 index 000000000..c6526a39a --- /dev/null +++ b/target/linux/generic/pending-5.15/732-02-net-ethernet-mtk_eth_soc-increase-tx-ring-side-for-Q.patch @@ -0,0 +1,143 @@ +From: Felix Fietkau +Date: Thu, 27 Oct 2022 19:53:57 +0200 +Subject: [PATCH] net: ethernet: mtk_eth_soc: increase tx ring side for + QDMA devices + +In order to use the hardware traffic shaper feature, a larger tx ring is +needed, especially for the scratch ring, which the hardware shaper uses to +reorder packets. + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -894,7 +894,7 @@ static int mtk_init_fq_dma(struct mtk_et + { + const struct mtk_soc_data *soc = eth->soc; + dma_addr_t phy_ring_tail; +- int cnt = MTK_DMA_SIZE; ++ int cnt = MTK_QDMA_RING_SIZE; + dma_addr_t dma_addr; + int i; + +@@ -2148,19 +2148,25 @@ static int mtk_tx_alloc(struct mtk_eth * + struct mtk_tx_ring *ring = ð->tx_ring; + int i, sz = soc->txrx.txd_size; + struct mtk_tx_dma_v2 *txd; ++ int ring_size; + +- ring->buf = kcalloc(MTK_DMA_SIZE, sizeof(*ring->buf), ++ if (MTK_HAS_CAPS(soc->caps, MTK_QDMA)) ++ ring_size = MTK_QDMA_RING_SIZE; ++ else ++ ring_size = MTK_DMA_SIZE; ++ ++ ring->buf = kcalloc(ring_size, sizeof(*ring->buf), + GFP_KERNEL); + if (!ring->buf) + goto no_tx_mem; + +- ring->dma = dma_alloc_coherent(eth->dma_dev, MTK_DMA_SIZE * sz, ++ ring->dma = dma_alloc_coherent(eth->dma_dev, ring_size * sz, + &ring->phys, GFP_KERNEL); + if (!ring->dma) + goto no_tx_mem; + +- for (i = 0; i < MTK_DMA_SIZE; i++) { +- int next = (i + 1) % MTK_DMA_SIZE; ++ for (i = 0; i < ring_size; i++) { ++ int next = (i + 1) % ring_size; + u32 next_ptr = ring->phys + next * sz; + + txd = ring->dma + i * sz; +@@ -2180,22 +2186,22 @@ static int mtk_tx_alloc(struct mtk_eth * + * descriptors in ring->dma_pdma. + */ + if (!MTK_HAS_CAPS(soc->caps, MTK_QDMA)) { +- ring->dma_pdma = dma_alloc_coherent(eth->dma_dev, MTK_DMA_SIZE * sz, ++ ring->dma_pdma = dma_alloc_coherent(eth->dma_dev, ring_size * sz, + &ring->phys_pdma, GFP_KERNEL); + if (!ring->dma_pdma) + goto no_tx_mem; + +- for (i = 0; i < MTK_DMA_SIZE; i++) { ++ for (i = 0; i < ring_size; i++) { + ring->dma_pdma[i].txd2 = TX_DMA_DESP2_DEF; + ring->dma_pdma[i].txd4 = 0; + } + } + +- ring->dma_size = MTK_DMA_SIZE; +- atomic_set(&ring->free_count, MTK_DMA_SIZE - 2); ++ ring->dma_size = ring_size; ++ atomic_set(&ring->free_count, ring_size - 2); + ring->next_free = ring->dma; + ring->last_free = (void *)txd; +- ring->last_free_ptr = (u32)(ring->phys + ((MTK_DMA_SIZE - 1) * sz)); ++ ring->last_free_ptr = (u32)(ring->phys + ((ring_size - 1) * sz)); + ring->thresh = MAX_SKB_FRAGS; + + /* make sure that all changes to the dma ring are flushed before we +@@ -2207,14 +2213,14 @@ static int mtk_tx_alloc(struct mtk_eth * + mtk_w32(eth, ring->phys, soc->reg_map->qdma.ctx_ptr); + mtk_w32(eth, ring->phys, soc->reg_map->qdma.dtx_ptr); + mtk_w32(eth, +- ring->phys + ((MTK_DMA_SIZE - 1) * sz), ++ ring->phys + ((ring_size - 1) * sz), + soc->reg_map->qdma.crx_ptr); + mtk_w32(eth, ring->last_free_ptr, soc->reg_map->qdma.drx_ptr); + mtk_w32(eth, (QDMA_RES_THRES << 8) | QDMA_RES_THRES, + soc->reg_map->qdma.qtx_cfg); + } else { + mtk_w32(eth, ring->phys_pdma, MT7628_TX_BASE_PTR0); +- mtk_w32(eth, MTK_DMA_SIZE, MT7628_TX_MAX_CNT0); ++ mtk_w32(eth, ring_size, MT7628_TX_MAX_CNT0); + mtk_w32(eth, 0, MT7628_TX_CTX_IDX0); + mtk_w32(eth, MT7628_PST_DTX_IDX0, soc->reg_map->pdma.rst_idx); + } +@@ -2232,7 +2238,7 @@ static void mtk_tx_clean(struct mtk_eth + int i; + + if (ring->buf) { +- for (i = 0; i < MTK_DMA_SIZE; i++) ++ for (i = 0; i < ring->dma_size; i++) + mtk_tx_unmap(eth, &ring->buf[i], false); + kfree(ring->buf); + ring->buf = NULL; +@@ -2240,14 +2246,14 @@ static void mtk_tx_clean(struct mtk_eth + + if (ring->dma) { + dma_free_coherent(eth->dma_dev, +- MTK_DMA_SIZE * soc->txrx.txd_size, ++ ring->dma_size * soc->txrx.txd_size, + ring->dma, ring->phys); + ring->dma = NULL; + } + + if (ring->dma_pdma) { + dma_free_coherent(eth->dma_dev, +- MTK_DMA_SIZE * soc->txrx.txd_size, ++ ring->dma_size * soc->txrx.txd_size, + ring->dma_pdma, ring->phys_pdma); + ring->dma_pdma = NULL; + } +@@ -2767,7 +2773,7 @@ static void mtk_dma_free(struct mtk_eth + netdev_reset_queue(eth->netdev[i]); + if (eth->scratch_ring) { + dma_free_coherent(eth->dma_dev, +- MTK_DMA_SIZE * soc->txrx.txd_size, ++ MTK_QDMA_RING_SIZE * soc->txrx.txd_size, + eth->scratch_ring, eth->phy_scratch_ring); + eth->scratch_ring = NULL; + eth->phy_scratch_ring = 0; +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -27,6 +27,7 @@ + #define MTK_MAX_RX_LENGTH_2K 2048 + #define MTK_TX_DMA_BUF_LEN 0x3fff + #define MTK_TX_DMA_BUF_LEN_V2 0xffff ++#define MTK_QDMA_RING_SIZE 2048 + #define MTK_DMA_SIZE 512 + #define MTK_MAC_COUNT 2 + #define MTK_RX_ETH_HLEN (VLAN_ETH_HLEN + ETH_FCS_LEN) diff --git a/target/linux/generic/pending-5.15/732-03-net-ethernet-mtk_eth_soc-avoid-port_mg-assignment-on.patch b/target/linux/generic/pending-5.15/732-03-net-ethernet-mtk_eth_soc-avoid-port_mg-assignment-on.patch new file mode 100644 index 000000000..116ae011f --- /dev/null +++ b/target/linux/generic/pending-5.15/732-03-net-ethernet-mtk_eth_soc-avoid-port_mg-assignment-on.patch @@ -0,0 +1,52 @@ +From: Felix Fietkau +Date: Fri, 4 Nov 2022 19:49:08 +0100 +Subject: [PATCH] net: ethernet: mtk_eth_soc: avoid port_mg assignment on + MT7622 and newer + +On newer chips, this field is unused and contains some bits related to queue +assignment. Initialize it to 0 in those cases. +Fix offload_version on MT7621 and MT7623, which still need the previous value. + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -4197,7 +4197,7 @@ static const struct mtk_soc_data mt7621_ + .hw_features = MTK_HW_FEATURES, + .required_clks = MT7621_CLKS_BITMAP, + .required_pctl = false, +- .offload_version = 2, ++ .offload_version = 1, + .hash_offset = 2, + .foe_entry_size = sizeof(struct mtk_foe_entry) - 16, + .txrx = { +@@ -4237,7 +4237,7 @@ static const struct mtk_soc_data mt7623_ + .hw_features = MTK_HW_FEATURES, + .required_clks = MT7623_CLKS_BITMAP, + .required_pctl = true, +- .offload_version = 2, ++ .offload_version = 1, + .hash_offset = 2, + .foe_entry_size = sizeof(struct mtk_foe_entry) - 16, + .txrx = { +--- a/drivers/net/ethernet/mediatek/mtk_ppe.c ++++ b/drivers/net/ethernet/mediatek/mtk_ppe.c +@@ -215,6 +215,8 @@ int mtk_foe_entry_prepare(struct mtk_eth + val = FIELD_PREP(MTK_FOE_IB2_DEST_PORT_V2, pse_port) | + FIELD_PREP(MTK_FOE_IB2_PORT_AG_V2, 0xf); + } else { ++ int port_mg = eth->soc->offload_version > 1 ? 0 : 0x3f; ++ + val = FIELD_PREP(MTK_FOE_IB1_STATE, MTK_FOE_STATE_BIND) | + FIELD_PREP(MTK_FOE_IB1_PACKET_TYPE, type) | + FIELD_PREP(MTK_FOE_IB1_UDP, l4proto == IPPROTO_UDP) | +@@ -222,7 +224,7 @@ int mtk_foe_entry_prepare(struct mtk_eth + entry->ib1 = val; + + val = FIELD_PREP(MTK_FOE_IB2_DEST_PORT, pse_port) | +- FIELD_PREP(MTK_FOE_IB2_PORT_MG, 0x3f) | ++ FIELD_PREP(MTK_FOE_IB2_PORT_MG, port_mg) | + FIELD_PREP(MTK_FOE_IB2_PORT_AG, 0x1f); + } + diff --git a/target/linux/generic/pending-5.15/732-04-net-ethernet-mtk_eth_soc-implement-multi-queue-suppo.patch b/target/linux/generic/pending-5.15/732-04-net-ethernet-mtk_eth_soc-implement-multi-queue-suppo.patch new file mode 100644 index 000000000..fd1898586 --- /dev/null +++ b/target/linux/generic/pending-5.15/732-04-net-ethernet-mtk_eth_soc-implement-multi-queue-suppo.patch @@ -0,0 +1,654 @@ +From: Felix Fietkau +Date: Thu, 27 Oct 2022 20:17:27 +0200 +Subject: [PATCH] net: ethernet: mtk_eth_soc: implement multi-queue + support for per-port queues + +When sending traffic to multiple ports with different link speeds, queued +packets to one port can drown out tx to other ports. +In order to better handle transmission to multiple ports, use the hardware +shaper feature to implement weighted fair queueing between ports. +Weight and maximum rate are automatically adjusted based on the link speed +of the port. +The first 3 queues are unrestricted and reserved for non-DSA direct tx on +GMAC ports. The following queues are automatically assigned by the MTK DSA +tag driver based on the target port number. +The PPE offload code configures the queues for offloaded traffic in the same +way. +This feature is only supported on devices supporting QDMA. All queues still +share the same DMA ring and descriptor pool. + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -54,6 +54,7 @@ static const struct mtk_reg_map mtk_reg_ + }, + .qdma = { + .qtx_cfg = 0x1800, ++ .qtx_sch = 0x1804, + .rx_ptr = 0x1900, + .rx_cnt_cfg = 0x1904, + .qcrx_ptr = 0x1908, +@@ -61,6 +62,7 @@ static const struct mtk_reg_map mtk_reg_ + .rst_idx = 0x1a08, + .delay_irq = 0x1a0c, + .fc_th = 0x1a10, ++ .tx_sch_rate = 0x1a14, + .int_grp = 0x1a20, + .hred = 0x1a44, + .ctx_ptr = 0x1b00, +@@ -113,6 +115,7 @@ static const struct mtk_reg_map mt7986_r + }, + .qdma = { + .qtx_cfg = 0x4400, ++ .qtx_sch = 0x4404, + .rx_ptr = 0x4500, + .rx_cnt_cfg = 0x4504, + .qcrx_ptr = 0x4508, +@@ -130,6 +133,7 @@ static const struct mtk_reg_map mt7986_r + .fq_tail = 0x4724, + .fq_count = 0x4728, + .fq_blen = 0x472c, ++ .tx_sch_rate = 0x4798, + }, + .gdm1_cnt = 0x1c00, + .gdma_to_ppe0 = 0x3333, +@@ -570,6 +574,75 @@ static void mtk_mac_link_down(struct phy + mtk_w32(mac->hw, mcr, MTK_MAC_MCR(mac->id)); + } + ++static void mtk_set_queue_speed(struct mtk_eth *eth, unsigned int idx, ++ int speed) ++{ ++ const struct mtk_soc_data *soc = eth->soc; ++ u32 ofs, val; ++ ++ if (!MTK_HAS_CAPS(soc->caps, MTK_QDMA)) ++ return; ++ ++ val = MTK_QTX_SCH_MIN_RATE_EN | ++ /* minimum: 10 Mbps */ ++ FIELD_PREP(MTK_QTX_SCH_MIN_RATE_MAN, 1) | ++ FIELD_PREP(MTK_QTX_SCH_MIN_RATE_EXP, 4) | ++ MTK_QTX_SCH_LEAKY_BUCKET_SIZE; ++ if (!MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) ++ val |= MTK_QTX_SCH_LEAKY_BUCKET_EN; ++ ++ if (IS_ENABLED(CONFIG_SOC_MT7621)) { ++ switch (speed) { ++ case SPEED_10: ++ val |= MTK_QTX_SCH_MAX_RATE_EN | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_MAN, 103) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_EXP, 2) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_WEIGHT, 1); ++ break; ++ case SPEED_100: ++ val |= MTK_QTX_SCH_MAX_RATE_EN | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_MAN, 103) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_EXP, 3); ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_WEIGHT, 1); ++ break; ++ case SPEED_1000: ++ val |= MTK_QTX_SCH_MAX_RATE_EN | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_MAN, 105) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_EXP, 4) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_WEIGHT, 10); ++ break; ++ default: ++ break; ++ } ++ } else { ++ switch (speed) { ++ case SPEED_10: ++ val |= MTK_QTX_SCH_MAX_RATE_EN | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_MAN, 1) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_EXP, 4) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_WEIGHT, 1); ++ break; ++ case SPEED_100: ++ val |= MTK_QTX_SCH_MAX_RATE_EN | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_MAN, 1) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_EXP, 5); ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_WEIGHT, 1); ++ break; ++ case SPEED_1000: ++ val |= MTK_QTX_SCH_MAX_RATE_EN | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_MAN, 10) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_EXP, 5) | ++ FIELD_PREP(MTK_QTX_SCH_MAX_RATE_WEIGHT, 10); ++ break; ++ default: ++ break; ++ } ++ } ++ ++ ofs = MTK_QTX_OFFSET * idx; ++ mtk_w32(eth, val, soc->reg_map->qdma.qtx_sch + ofs); ++} ++ + static void mtk_mac_link_up(struct phylink_config *config, + struct phy_device *phy, + unsigned int mode, phy_interface_t interface, +@@ -595,6 +668,8 @@ static void mtk_mac_link_up(struct phyli + break; + } + ++ mtk_set_queue_speed(mac->hw, mac->id, speed); ++ + /* Configure duplex */ + if (duplex == DUPLEX_FULL) + mcr |= MAC_MCR_FORCE_DPX; +@@ -1053,7 +1128,8 @@ static void mtk_tx_set_dma_desc_v1(struc + + WRITE_ONCE(desc->txd1, info->addr); + +- data = TX_DMA_SWC | TX_DMA_PLEN0(info->size); ++ data = TX_DMA_SWC | TX_DMA_PLEN0(info->size) | ++ FIELD_PREP(TX_DMA_PQID, info->qid); + if (info->last) + data |= TX_DMA_LS0; + WRITE_ONCE(desc->txd3, data); +@@ -1087,9 +1163,6 @@ static void mtk_tx_set_dma_desc_v2(struc + data |= TX_DMA_LS0; + WRITE_ONCE(desc->txd3, data); + +- if (!info->qid && mac->id) +- info->qid = MTK_QDMA_GMAC2_QID; +- + data = (mac->id + 1) << TX_DMA_FPORT_SHIFT_V2; /* forward port */ + data |= TX_DMA_SWC_V2 | QID_BITS_V2(info->qid); + WRITE_ONCE(desc->txd4, data); +@@ -1133,11 +1206,12 @@ static int mtk_tx_map(struct sk_buff *sk + .gso = gso, + .csum = skb->ip_summed == CHECKSUM_PARTIAL, + .vlan = skb_vlan_tag_present(skb), +- .qid = skb->mark & MTK_QDMA_TX_MASK, ++ .qid = skb_get_queue_mapping(skb), + .vlan_tci = skb_vlan_tag_get(skb), + .first = true, + .last = !skb_is_nonlinear(skb), + }; ++ struct netdev_queue *txq; + struct mtk_mac *mac = netdev_priv(dev); + struct mtk_eth *eth = mac->hw; + const struct mtk_soc_data *soc = eth->soc; +@@ -1145,8 +1219,10 @@ static int mtk_tx_map(struct sk_buff *sk + struct mtk_tx_dma *itxd_pdma, *txd_pdma; + struct mtk_tx_buf *itx_buf, *tx_buf; + int i, n_desc = 1; ++ int queue = skb_get_queue_mapping(skb); + int k = 0; + ++ txq = netdev_get_tx_queue(dev, queue); + itxd = ring->next_free; + itxd_pdma = qdma_to_pdma(ring, itxd); + if (itxd == ring->last_free) +@@ -1195,7 +1271,7 @@ static int mtk_tx_map(struct sk_buff *sk + memset(&txd_info, 0, sizeof(struct mtk_tx_dma_desc_info)); + txd_info.size = min_t(unsigned int, frag_size, + soc->txrx.dma_max_len); +- txd_info.qid = skb->mark & MTK_QDMA_TX_MASK; ++ txd_info.qid = queue; + txd_info.last = i == skb_shinfo(skb)->nr_frags - 1 && + !(frag_size - txd_info.size); + txd_info.addr = skb_frag_dma_map(eth->dma_dev, frag, +@@ -1234,7 +1310,7 @@ static int mtk_tx_map(struct sk_buff *sk + txd_pdma->txd2 |= TX_DMA_LS1; + } + +- netdev_sent_queue(dev, skb->len); ++ netdev_tx_sent_queue(txq, skb->len); + skb_tx_timestamp(skb); + + ring->next_free = mtk_qdma_phys_to_virt(ring, txd->txd2); +@@ -1246,8 +1322,7 @@ static int mtk_tx_map(struct sk_buff *sk + wmb(); + + if (MTK_HAS_CAPS(soc->caps, MTK_QDMA)) { +- if (netif_xmit_stopped(netdev_get_tx_queue(dev, 0)) || +- !netdev_xmit_more()) ++ if (netif_xmit_stopped(txq) || !netdev_xmit_more()) + mtk_w32(eth, txd->txd2, soc->reg_map->qdma.ctx_ptr); + } else { + int next_idx; +@@ -1316,7 +1391,7 @@ static void mtk_wake_queue(struct mtk_et + for (i = 0; i < MTK_MAC_COUNT; i++) { + if (!eth->netdev[i]) + continue; +- netif_wake_queue(eth->netdev[i]); ++ netif_tx_wake_all_queues(eth->netdev[i]); + } + } + +@@ -1340,7 +1415,7 @@ static netdev_tx_t mtk_start_xmit(struct + + tx_num = mtk_cal_txd_req(eth, skb); + if (unlikely(atomic_read(&ring->free_count) <= tx_num)) { +- netif_stop_queue(dev); ++ netif_tx_stop_all_queues(dev); + netif_err(eth, tx_queued, dev, + "Tx Ring full when queue awake!\n"); + spin_unlock(ð->page_lock); +@@ -1366,7 +1441,7 @@ static netdev_tx_t mtk_start_xmit(struct + goto drop; + + if (unlikely(atomic_read(&ring->free_count) <= ring->thresh)) +- netif_stop_queue(dev); ++ netif_tx_stop_all_queues(dev); + + spin_unlock(ð->page_lock); + +@@ -1533,10 +1608,12 @@ static int mtk_xdp_submit_frame(struct m + struct skb_shared_info *sinfo = xdp_get_shared_info_from_frame(xdpf); + const struct mtk_soc_data *soc = eth->soc; + struct mtk_tx_ring *ring = ð->tx_ring; ++ struct mtk_mac *mac = netdev_priv(dev); + struct mtk_tx_dma_desc_info txd_info = { + .size = xdpf->len, + .first = true, + .last = !xdp_frame_has_frags(xdpf), ++ .qid = mac->id, + }; + int err, index = 0, n_desc = 1, nr_frags; + struct mtk_tx_dma *htxd, *txd, *txd_pdma; +@@ -1587,6 +1664,7 @@ static int mtk_xdp_submit_frame(struct m + memset(&txd_info, 0, sizeof(struct mtk_tx_dma_desc_info)); + txd_info.size = skb_frag_size(&sinfo->frags[index]); + txd_info.last = index + 1 == nr_frags; ++ txd_info.qid = mac->id; + data = skb_frag_address(&sinfo->frags[index]); + + index++; +@@ -1938,8 +2016,46 @@ rx_done: + return done; + } + ++struct mtk_poll_state { ++ struct netdev_queue *txq; ++ unsigned int total; ++ unsigned int done; ++ unsigned int bytes; ++}; ++ ++static void ++mtk_poll_tx_done(struct mtk_eth *eth, struct mtk_poll_state *state, u8 mac, ++ struct sk_buff *skb) ++{ ++ struct netdev_queue *txq; ++ struct net_device *dev; ++ unsigned int bytes = skb->len; ++ ++ state->total++; ++ eth->tx_packets++; ++ eth->tx_bytes += bytes; ++ ++ dev = eth->netdev[mac]; ++ if (!dev) ++ return; ++ ++ txq = netdev_get_tx_queue(dev, skb_get_queue_mapping(skb)); ++ if (state->txq == txq) { ++ state->done++; ++ state->bytes += bytes; ++ return; ++ } ++ ++ if (state->txq) ++ netdev_tx_completed_queue(state->txq, state->done, state->bytes); ++ ++ state->txq = txq; ++ state->done = 1; ++ state->bytes = bytes; ++} ++ + static int mtk_poll_tx_qdma(struct mtk_eth *eth, int budget, +- unsigned int *done, unsigned int *bytes) ++ struct mtk_poll_state *state) + { + const struct mtk_reg_map *reg_map = eth->soc->reg_map; + struct mtk_tx_ring *ring = ð->tx_ring; +@@ -1969,12 +2085,9 @@ static int mtk_poll_tx_qdma(struct mtk_e + break; + + if (tx_buf->data != (void *)MTK_DMA_DUMMY_DESC) { +- if (tx_buf->type == MTK_TYPE_SKB) { +- struct sk_buff *skb = tx_buf->data; ++ if (tx_buf->type == MTK_TYPE_SKB) ++ mtk_poll_tx_done(eth, state, mac, tx_buf->data); + +- bytes[mac] += skb->len; +- done[mac]++; +- } + budget--; + } + mtk_tx_unmap(eth, tx_buf, true); +@@ -1992,7 +2105,7 @@ static int mtk_poll_tx_qdma(struct mtk_e + } + + static int mtk_poll_tx_pdma(struct mtk_eth *eth, int budget, +- unsigned int *done, unsigned int *bytes) ++ struct mtk_poll_state *state) + { + struct mtk_tx_ring *ring = ð->tx_ring; + struct mtk_tx_buf *tx_buf; +@@ -2008,12 +2121,8 @@ static int mtk_poll_tx_pdma(struct mtk_e + break; + + if (tx_buf->data != (void *)MTK_DMA_DUMMY_DESC) { +- if (tx_buf->type == MTK_TYPE_SKB) { +- struct sk_buff *skb = tx_buf->data; +- +- bytes[0] += skb->len; +- done[0]++; +- } ++ if (tx_buf->type == MTK_TYPE_SKB) ++ mtk_poll_tx_done(eth, state, 0, tx_buf->data); + budget--; + } + mtk_tx_unmap(eth, tx_buf, true); +@@ -2034,26 +2143,15 @@ static int mtk_poll_tx(struct mtk_eth *e + { + struct mtk_tx_ring *ring = ð->tx_ring; + struct dim_sample dim_sample = {}; +- unsigned int done[MTK_MAX_DEVS]; +- unsigned int bytes[MTK_MAX_DEVS]; +- int total = 0, i; +- +- memset(done, 0, sizeof(done)); +- memset(bytes, 0, sizeof(bytes)); ++ struct mtk_poll_state state = {}; + + if (MTK_HAS_CAPS(eth->soc->caps, MTK_QDMA)) +- budget = mtk_poll_tx_qdma(eth, budget, done, bytes); ++ budget = mtk_poll_tx_qdma(eth, budget, &state); + else +- budget = mtk_poll_tx_pdma(eth, budget, done, bytes); ++ budget = mtk_poll_tx_pdma(eth, budget, &state); + +- for (i = 0; i < MTK_MAC_COUNT; i++) { +- if (!eth->netdev[i] || !done[i]) +- continue; +- netdev_completed_queue(eth->netdev[i], done[i], bytes[i]); +- total += done[i]; +- eth->tx_packets += done[i]; +- eth->tx_bytes += bytes[i]; +- } ++ if (state.txq) ++ netdev_tx_completed_queue(state.txq, state.done, state.bytes); + + dim_update_sample(eth->tx_events, eth->tx_packets, eth->tx_bytes, + &dim_sample); +@@ -2063,7 +2161,7 @@ static int mtk_poll_tx(struct mtk_eth *e + (atomic_read(&ring->free_count) > ring->thresh)) + mtk_wake_queue(eth); + +- return total; ++ return state.total; + } + + static void mtk_handle_status_irq(struct mtk_eth *eth) +@@ -2149,6 +2247,7 @@ static int mtk_tx_alloc(struct mtk_eth * + int i, sz = soc->txrx.txd_size; + struct mtk_tx_dma_v2 *txd; + int ring_size; ++ u32 ofs, val; + + if (MTK_HAS_CAPS(soc->caps, MTK_QDMA)) + ring_size = MTK_QDMA_RING_SIZE; +@@ -2216,8 +2315,25 @@ static int mtk_tx_alloc(struct mtk_eth * + ring->phys + ((ring_size - 1) * sz), + soc->reg_map->qdma.crx_ptr); + mtk_w32(eth, ring->last_free_ptr, soc->reg_map->qdma.drx_ptr); +- mtk_w32(eth, (QDMA_RES_THRES << 8) | QDMA_RES_THRES, +- soc->reg_map->qdma.qtx_cfg); ++ ++ for (i = 0, ofs = 0; i < MTK_QDMA_NUM_QUEUES; i++) { ++ val = (QDMA_RES_THRES << 8) | QDMA_RES_THRES; ++ mtk_w32(eth, val, soc->reg_map->qdma.qtx_cfg + ofs); ++ ++ val = MTK_QTX_SCH_MIN_RATE_EN | ++ /* minimum: 10 Mbps */ ++ FIELD_PREP(MTK_QTX_SCH_MIN_RATE_MAN, 1) | ++ FIELD_PREP(MTK_QTX_SCH_MIN_RATE_EXP, 4) | ++ MTK_QTX_SCH_LEAKY_BUCKET_SIZE; ++ if (!MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) ++ val |= MTK_QTX_SCH_LEAKY_BUCKET_EN; ++ mtk_w32(eth, val, soc->reg_map->qdma.qtx_sch + ofs); ++ ofs += MTK_QTX_OFFSET; ++ } ++ val = MTK_QDMA_TX_SCH_MAX_WFQ | (MTK_QDMA_TX_SCH_MAX_WFQ << 16); ++ mtk_w32(eth, val, soc->reg_map->qdma.tx_sch_rate); ++ if (MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) ++ mtk_w32(eth, val, soc->reg_map->qdma.tx_sch_rate + 4); + } else { + mtk_w32(eth, ring->phys_pdma, MT7628_TX_BASE_PTR0); + mtk_w32(eth, ring_size, MT7628_TX_MAX_CNT0); +@@ -2882,7 +2998,7 @@ static int mtk_start_dma(struct mtk_eth + if (MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) + val |= MTK_MUTLI_CNT | MTK_RESV_BUF | + MTK_WCOMP_EN | MTK_DMAD_WR_WDONE | +- MTK_CHK_DDONE_EN; ++ MTK_CHK_DDONE_EN | MTK_LEAKY_BUCKET_EN; + else + val |= MTK_RX_BT_32DWORDS; + mtk_w32(eth, val, reg_map->qdma.glo_cfg); +@@ -2928,6 +3044,45 @@ static void mtk_gdm_config(struct mtk_et + mtk_w32(eth, 0, MTK_RST_GL); + } + ++static int mtk_device_event(struct notifier_block *n, unsigned long event, void *ptr) ++{ ++ struct mtk_mac *mac = container_of(n, struct mtk_mac, device_notifier); ++ struct mtk_eth *eth = mac->hw; ++ struct net_device *dev = netdev_notifier_info_to_dev(ptr); ++ struct ethtool_link_ksettings s; ++ struct net_device *ldev; ++ struct list_head *iter; ++ struct dsa_port *dp; ++ ++ if (event != NETDEV_CHANGE) ++ return NOTIFY_DONE; ++ ++ netdev_for_each_lower_dev(dev, ldev, iter) { ++ if (netdev_priv(ldev) == mac) ++ goto found; ++ } ++ ++ return NOTIFY_DONE; ++ ++found: ++ if (!dsa_slave_dev_check(dev)) ++ return NOTIFY_DONE; ++ ++ if (__ethtool_get_link_ksettings(dev, &s)) ++ return NOTIFY_DONE; ++ ++ if (s.base.speed == 0 || s.base.speed == ((__u32)-1)) ++ return NOTIFY_DONE; ++ ++ dp = dsa_port_from_netdev(dev); ++ if (dp->index >= MTK_QDMA_NUM_QUEUES) ++ return NOTIFY_DONE; ++ ++ mtk_set_queue_speed(eth, dp->index + 3, s.base.speed); ++ ++ return NOTIFY_DONE; ++} ++ + static int mtk_open(struct net_device *dev) + { + struct mtk_mac *mac = netdev_priv(dev); +@@ -2970,7 +3125,8 @@ static int mtk_open(struct net_device *d + refcount_inc(ð->dma_refcnt); + + phylink_start(mac->phylink); +- netif_start_queue(dev); ++ netif_tx_start_all_queues(dev); ++ + return 0; + } + +@@ -3486,8 +3642,12 @@ static int mtk_unreg_dev(struct mtk_eth + int i; + + for (i = 0; i < MTK_MAC_COUNT; i++) { ++ struct mtk_mac *mac; + if (!eth->netdev[i]) + continue; ++ mac = netdev_priv(eth->netdev[i]); ++ if (MTK_HAS_CAPS(eth->soc->caps, MTK_QDMA)) ++ unregister_netdevice_notifier(&mac->device_notifier); + unregister_netdev(eth->netdev[i]); + } + +@@ -3703,6 +3863,23 @@ static int mtk_set_rxnfc(struct net_devi + return ret; + } + ++static u16 mtk_select_queue(struct net_device *dev, struct sk_buff *skb, ++ struct net_device *sb_dev) ++{ ++ struct mtk_mac *mac = netdev_priv(dev); ++ unsigned int queue = 0; ++ ++ if (netdev_uses_dsa(dev)) ++ queue = skb_get_queue_mapping(skb) + 3; ++ else ++ queue = mac->id; ++ ++ if (queue >= dev->num_tx_queues) ++ queue = 0; ++ ++ return queue; ++} ++ + static const struct ethtool_ops mtk_ethtool_ops = { + .get_link_ksettings = mtk_get_link_ksettings, + .set_link_ksettings = mtk_set_link_ksettings, +@@ -3738,6 +3915,7 @@ static const struct net_device_ops mtk_n + .ndo_setup_tc = mtk_eth_setup_tc, + .ndo_bpf = mtk_xdp, + .ndo_xdp_xmit = mtk_xdp_xmit, ++ .ndo_select_queue = mtk_select_queue, + }; + + static int mtk_add_mac(struct mtk_eth *eth, struct device_node *np) +@@ -3747,6 +3925,7 @@ static int mtk_add_mac(struct mtk_eth *e + struct phylink *phylink; + struct mtk_mac *mac; + int id, err; ++ int txqs = 1; + + if (!_id) { + dev_err(eth->dev, "missing mac id\n"); +@@ -3764,7 +3943,10 @@ static int mtk_add_mac(struct mtk_eth *e + return -EINVAL; + } + +- eth->netdev[id] = alloc_etherdev(sizeof(*mac)); ++ if (MTK_HAS_CAPS(eth->soc->caps, MTK_QDMA)) ++ txqs = MTK_QDMA_NUM_QUEUES; ++ ++ eth->netdev[id] = alloc_etherdev_mqs(sizeof(*mac), txqs, 1); + if (!eth->netdev[id]) { + dev_err(eth->dev, "alloc_etherdev failed\n"); + return -ENOMEM; +@@ -3861,6 +4043,11 @@ static int mtk_add_mac(struct mtk_eth *e + else + eth->netdev[id]->max_mtu = MTK_MAX_RX_LENGTH_2K - MTK_RX_ETH_HLEN; + ++ if (MTK_HAS_CAPS(eth->soc->caps, MTK_QDMA)) { ++ mac->device_notifier.notifier_call = mtk_device_event; ++ register_netdevice_notifier(&mac->device_notifier); ++ } ++ + return 0; + + free_netdev: +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -22,6 +22,7 @@ + #include + #include "mtk_ppe.h" + ++#define MTK_QDMA_NUM_QUEUES 16 + #define MTK_QDMA_PAGE_SIZE 2048 + #define MTK_MAX_RX_LENGTH 1536 + #define MTK_MAX_RX_LENGTH_2K 2048 +@@ -203,8 +204,26 @@ + #define MTK_RING_MAX_AGG_CNT_H ((MTK_HW_LRO_MAX_AGG_CNT >> 6) & 0x3) + + /* QDMA TX Queue Configuration Registers */ ++#define MTK_QTX_OFFSET 0x10 + #define QDMA_RES_THRES 4 + ++/* QDMA Tx Queue Scheduler Configuration Registers */ ++#define MTK_QTX_SCH_TX_SEL BIT(31) ++#define MTK_QTX_SCH_TX_SEL_V2 GENMASK(31, 30) ++ ++#define MTK_QTX_SCH_LEAKY_BUCKET_EN BIT(30) ++#define MTK_QTX_SCH_LEAKY_BUCKET_SIZE GENMASK(29, 28) ++#define MTK_QTX_SCH_MIN_RATE_EN BIT(27) ++#define MTK_QTX_SCH_MIN_RATE_MAN GENMASK(26, 20) ++#define MTK_QTX_SCH_MIN_RATE_EXP GENMASK(19, 16) ++#define MTK_QTX_SCH_MAX_RATE_WEIGHT GENMASK(15, 12) ++#define MTK_QTX_SCH_MAX_RATE_EN BIT(11) ++#define MTK_QTX_SCH_MAX_RATE_MAN GENMASK(10, 4) ++#define MTK_QTX_SCH_MAX_RATE_EXP GENMASK(3, 0) ++ ++/* QDMA TX Scheduler Rate Control Register */ ++#define MTK_QDMA_TX_SCH_MAX_WFQ BIT(15) ++ + /* QDMA Global Configuration Register */ + #define MTK_RX_2B_OFFSET BIT(31) + #define MTK_RX_BT_32DWORDS (3 << 11) +@@ -223,6 +242,7 @@ + #define MTK_WCOMP_EN BIT(24) + #define MTK_RESV_BUF (0x40 << 16) + #define MTK_MUTLI_CNT (0x4 << 12) ++#define MTK_LEAKY_BUCKET_EN BIT(11) + + /* QDMA Flow Control Register */ + #define FC_THRES_DROP_MODE BIT(20) +@@ -251,8 +271,6 @@ + #define MTK_STAT_OFFSET 0x40 + + /* QDMA TX NUM */ +-#define MTK_QDMA_TX_NUM 16 +-#define MTK_QDMA_TX_MASK (MTK_QDMA_TX_NUM - 1) + #define QID_BITS_V2(x) (((x) & 0x3f) << 16) + #define MTK_QDMA_GMAC2_QID 8 + +@@ -282,6 +300,7 @@ + #define TX_DMA_PLEN0(x) (((x) & eth->soc->txrx.dma_max_len) << eth->soc->txrx.dma_len_offset) + #define TX_DMA_PLEN1(x) ((x) & eth->soc->txrx.dma_max_len) + #define TX_DMA_SWC BIT(14) ++#define TX_DMA_PQID GENMASK(3, 0) + + /* PDMA on MT7628 */ + #define TX_DMA_DONE BIT(31) +@@ -930,6 +949,7 @@ struct mtk_reg_map { + } pdma; + struct { + u32 qtx_cfg; /* tx queue configuration */ ++ u32 qtx_sch; /* tx queue scheduler configuration */ + u32 rx_ptr; /* rx base pointer */ + u32 rx_cnt_cfg; /* rx max count configuration */ + u32 qcrx_ptr; /* rx cpu pointer */ +@@ -947,6 +967,7 @@ struct mtk_reg_map { + u32 fq_tail; /* fq tail pointer */ + u32 fq_count; /* fq free page count */ + u32 fq_blen; /* fq free page buffer length */ ++ u32 tx_sch_rate; /* tx scheduler rate control registers */ + } qdma; + u32 gdm1_cnt; + u32 gdma_to_ppe0; +@@ -1139,6 +1160,7 @@ struct mtk_mac { + __be32 hwlro_ip[MTK_MAX_LRO_IP_CNT]; + int hwlro_ip_cnt; + unsigned int syscfg0; ++ struct notifier_block device_notifier; + }; + + /* the struct describing the SoC. these are declared in the soc_xyz.c files */ diff --git a/target/linux/generic/pending-5.15/732-05-net-dsa-tag_mtk-assign-per-port-queues.patch b/target/linux/generic/pending-5.15/732-05-net-dsa-tag_mtk-assign-per-port-queues.patch new file mode 100644 index 000000000..7739366ad --- /dev/null +++ b/target/linux/generic/pending-5.15/732-05-net-dsa-tag_mtk-assign-per-port-queues.patch @@ -0,0 +1,20 @@ +From: Felix Fietkau +Date: Fri, 28 Oct 2022 18:16:03 +0200 +Subject: [PATCH] net: dsa: tag_mtk: assign per-port queues + +Keeps traffic sent to the switch within link speed limits + +Signed-off-by: Felix Fietkau +--- + +--- a/net/dsa/tag_mtk.c ++++ b/net/dsa/tag_mtk.c +@@ -33,6 +33,8 @@ static struct sk_buff *mtk_tag_xmit(stru + if (__skb_put_padto(skb, ETH_ZLEN + MTK_HDR_LEN, false)) + return NULL; + ++ skb_set_queue_mapping(skb, dp->index); ++ + /* Build the special tag after the MAC Source Address. If VLAN header + * is present, it's required that VLAN header and special tag is + * being combined. Only in this way we can allow the switch can parse diff --git a/target/linux/generic/pending-5.15/732-06-net-ethernet-mediatek-ppe-assign-per-port-queues-for.patch b/target/linux/generic/pending-5.15/732-06-net-ethernet-mediatek-ppe-assign-per-port-queues-for.patch new file mode 100644 index 000000000..05161c347 --- /dev/null +++ b/target/linux/generic/pending-5.15/732-06-net-ethernet-mediatek-ppe-assign-per-port-queues-for.patch @@ -0,0 +1,93 @@ +From: Felix Fietkau +Date: Thu, 3 Nov 2022 17:49:44 +0100 +Subject: [PATCH] net: ethernet: mediatek: ppe: assign per-port queues + for offloaded traffic + +Keeps traffic sent to the switch within link speed limits + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_ppe.c ++++ b/drivers/net/ethernet/mediatek/mtk_ppe.c +@@ -445,6 +445,24 @@ static inline bool mtk_foe_entry_usable( + FIELD_GET(MTK_FOE_IB1_STATE, entry->ib1) != MTK_FOE_STATE_BIND; + } + ++int mtk_foe_entry_set_queue(struct mtk_eth *eth, struct mtk_foe_entry *entry, ++ unsigned int queue) ++{ ++ u32 *ib2 = mtk_foe_entry_ib2(eth, entry); ++ ++ if (MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) { ++ *ib2 &= ~MTK_FOE_IB2_QID_V2; ++ *ib2 |= FIELD_PREP(MTK_FOE_IB2_QID_V2, queue); ++ *ib2 |= MTK_FOE_IB2_PSE_QOS_V2; ++ } else { ++ *ib2 &= ~MTK_FOE_IB2_QID; ++ *ib2 |= FIELD_PREP(MTK_FOE_IB2_QID, queue); ++ *ib2 |= MTK_FOE_IB2_PSE_QOS; ++ } ++ ++ return 0; ++} ++ + static bool + mtk_flow_entry_match(struct mtk_eth *eth, struct mtk_flow_entry *entry, + struct mtk_foe_entry *data) +--- a/drivers/net/ethernet/mediatek/mtk_ppe.h ++++ b/drivers/net/ethernet/mediatek/mtk_ppe.h +@@ -69,7 +69,9 @@ enum { + #define MTK_FOE_IB2_DSCP GENMASK(31, 24) + + /* CONFIG_MEDIATEK_NETSYS_V2 */ ++#define MTK_FOE_IB2_QID_V2 GENMASK(6, 0) + #define MTK_FOE_IB2_PORT_MG_V2 BIT(7) ++#define MTK_FOE_IB2_PSE_QOS_V2 BIT(8) + #define MTK_FOE_IB2_DEST_PORT_V2 GENMASK(12, 9) + #define MTK_FOE_IB2_MULTICAST_V2 BIT(13) + #define MTK_FOE_IB2_WDMA_WINFO_V2 BIT(19) +@@ -369,6 +371,8 @@ int mtk_foe_entry_set_pppoe(struct mtk_e + int sid); + int mtk_foe_entry_set_wdma(struct mtk_eth *eth, struct mtk_foe_entry *entry, + int wdma_idx, int txq, int bss, int wcid); ++int mtk_foe_entry_set_queue(struct mtk_eth *eth, struct mtk_foe_entry *entry, ++ unsigned int queue); + int mtk_foe_entry_commit(struct mtk_ppe *ppe, struct mtk_flow_entry *entry); + void mtk_foe_entry_clear(struct mtk_ppe *ppe, struct mtk_flow_entry *entry); + int mtk_foe_entry_idle_time(struct mtk_ppe *ppe, struct mtk_flow_entry *entry); +--- a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c ++++ b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c +@@ -188,7 +188,7 @@ mtk_flow_set_output_device(struct mtk_et + int *wed_index) + { + struct mtk_wdma_info info = {}; +- int pse_port, dsa_port; ++ int pse_port, dsa_port, queue; + + if (mtk_flow_get_wdma_info(dev, dest_mac, &info) == 0) { + mtk_foe_entry_set_wdma(eth, foe, info.wdma_idx, info.queue, +@@ -212,8 +212,6 @@ mtk_flow_set_output_device(struct mtk_et + } + + dsa_port = mtk_flow_get_dsa_port(&dev); +- if (dsa_port >= 0) +- mtk_foe_entry_set_dsa(eth, foe, dsa_port); + + if (dev == eth->netdev[0]) + pse_port = 1; +@@ -222,6 +220,14 @@ mtk_flow_set_output_device(struct mtk_et + else + return -EOPNOTSUPP; + ++ if (dsa_port >= 0) { ++ mtk_foe_entry_set_dsa(eth, foe, dsa_port); ++ queue = 3 + dsa_port; ++ } else { ++ queue = pse_port - 1; ++ } ++ mtk_foe_entry_set_queue(eth, foe, queue); ++ + out: + mtk_foe_entry_set_pse_port(eth, foe, pse_port); + diff --git a/target/linux/generic/pending-5.15/732-07-net-ethernet-mtk_eth_soc-compile-out-netsys-v2-code-.patch b/target/linux/generic/pending-5.15/732-07-net-ethernet-mtk_eth_soc-compile-out-netsys-v2-code-.patch new file mode 100644 index 000000000..8c711ba80 --- /dev/null +++ b/target/linux/generic/pending-5.15/732-07-net-ethernet-mtk_eth_soc-compile-out-netsys-v2-code-.patch @@ -0,0 +1,28 @@ +From: Felix Fietkau +Date: Thu, 27 Oct 2022 23:39:52 +0200 +Subject: [PATCH] net: ethernet: mtk_eth_soc: compile out netsys v2 code + on mt7621 + +Avoid some branches in the hot path on low-end devices with limited CPU power, +and reduce code size + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -895,7 +895,13 @@ enum mkt_eth_capabilities { + #define MTK_MUX_GMAC12_TO_GEPHY_SGMII \ + (MTK_ETH_MUX_GMAC12_TO_GEPHY_SGMII | MTK_MUX) + +-#define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x)) ++#ifdef CONFIG_SOC_MT7621 ++#define MTK_CAP_MASK MTK_NETSYS_V2 ++#else ++#define MTK_CAP_MASK 0 ++#endif ++ ++#define MTK_HAS_CAPS(caps, _x) (((caps) & (_x) & ~(MTK_CAP_MASK)) == (_x)) + + #define MT7621_CAPS (MTK_GMAC1_RGMII | MTK_GMAC1_TRGMII | \ + MTK_GMAC2_RGMII | MTK_SHARED_INT | \ diff --git a/target/linux/generic/pending-5.15/732-08-net-dsa-add-support-for-DSA-rx-offloading-via-metada.patch b/target/linux/generic/pending-5.15/732-08-net-dsa-add-support-for-DSA-rx-offloading-via-metada.patch new file mode 100644 index 000000000..0478cb528 --- /dev/null +++ b/target/linux/generic/pending-5.15/732-08-net-dsa-add-support-for-DSA-rx-offloading-via-metada.patch @@ -0,0 +1,72 @@ +From: Felix Fietkau +Date: Tue, 8 Nov 2022 15:03:15 +0100 +Subject: [PATCH] net: dsa: add support for DSA rx offloading via + metadata dst + +If a metadata dst is present with the type METADATA_HW_PORT_MUX on a dsa cpu +port netdev, assume that it carries the port number and that there is no DSA +tag present in the skb data. + +Signed-off-by: Felix Fietkau +--- + +--- a/net/core/flow_dissector.c ++++ b/net/core/flow_dissector.c +@@ -940,12 +940,14 @@ bool __skb_flow_dissect(const struct net + #if IS_ENABLED(CONFIG_NET_DSA) + if (unlikely(skb->dev && netdev_uses_dsa(skb->dev) && + proto == htons(ETH_P_XDSA))) { ++ struct metadata_dst *md_dst = skb_metadata_dst(skb); + const struct dsa_device_ops *ops; + int offset = 0; + + ops = skb->dev->dsa_ptr->tag_ops; + /* Only DSA header taggers break flow dissection */ +- if (ops->needed_headroom) { ++ if (ops->needed_headroom && ++ (!md_dst || md_dst->type != METADATA_HW_PORT_MUX)) { + if (ops->flow_dissect) + ops->flow_dissect(skb, &proto, &offset); + else +--- a/net/dsa/dsa.c ++++ b/net/dsa/dsa.c +@@ -20,6 +20,7 @@ + #include + #include + #include ++#include + + #include "dsa_priv.h" + +@@ -225,6 +226,7 @@ static bool dsa_skb_defer_rx_timestamp(s + static int dsa_switch_rcv(struct sk_buff *skb, struct net_device *dev, + struct packet_type *pt, struct net_device *unused) + { ++ struct metadata_dst *md_dst = skb_metadata_dst(skb); + struct dsa_port *cpu_dp = dev->dsa_ptr; + struct sk_buff *nskb = NULL; + struct dsa_slave_priv *p; +@@ -238,7 +240,22 @@ static int dsa_switch_rcv(struct sk_buff + if (!skb) + return 0; + +- nskb = cpu_dp->rcv(skb, dev); ++ if (md_dst && md_dst->type == METADATA_HW_PORT_MUX) { ++ unsigned int port = md_dst->u.port_info.port_id; ++ ++ skb_dst_drop(skb); ++ if (!skb_has_extensions(skb)) ++ skb->slow_gro = 0; ++ ++ skb->dev = dsa_master_find_slave(dev, 0, port); ++ if (likely(skb->dev)) { ++ dsa_default_offload_fwd_mark(skb); ++ nskb = skb; ++ } ++ } else { ++ nskb = cpu_dp->rcv(skb, dev); ++ } ++ + if (!nskb) { + kfree_skb(skb); + return 0; diff --git a/target/linux/generic/pending-5.15/732-09-net-ethernet-mtk_eth_soc-fix-VLAN-rx-hardware-accele.patch b/target/linux/generic/pending-5.15/732-09-net-ethernet-mtk_eth_soc-fix-VLAN-rx-hardware-accele.patch new file mode 100644 index 000000000..c10906780 --- /dev/null +++ b/target/linux/generic/pending-5.15/732-09-net-ethernet-mtk_eth_soc-fix-VLAN-rx-hardware-accele.patch @@ -0,0 +1,192 @@ +From: Felix Fietkau +Date: Fri, 28 Oct 2022 11:01:12 +0200 +Subject: [PATCH] net: ethernet: mtk_eth_soc: fix VLAN rx hardware + acceleration + +- enable VLAN untagging for PDMA rx +- make it possible to disable the feature via ethtool +- pass VLAN tag to the DSA driver +- untag special tag on PDMA only if no non-DSA devices are in use +- disable special tag untagging on 7986 for now, since it's not working yet + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -23,6 +23,7 @@ + #include + #include + #include ++#include + + #include "mtk_eth_soc.h" + #include "mtk_wed.h" +@@ -1967,16 +1968,22 @@ static int mtk_poll_rx(struct napi_struc + htons(RX_DMA_VPID(trxd.rxd4)), + RX_DMA_VID(trxd.rxd4)); + } else if (trxd.rxd2 & RX_DMA_VTAG) { +- __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), ++ __vlan_hwaccel_put_tag(skb, htons(RX_DMA_VPID(trxd.rxd3)), + RX_DMA_VID(trxd.rxd3)); + } ++ } ++ ++ /* When using VLAN untagging in combination with DSA, the ++ * hardware treats the MTK special tag as a VLAN and untags it. ++ */ ++ if (skb_vlan_tag_present(skb) && netdev_uses_dsa(netdev)) { ++ unsigned int port = ntohs(skb->vlan_proto) & GENMASK(2, 0); + +- /* If the device is attached to a dsa switch, the special +- * tag inserted in VLAN field by hw switch can * be offloaded +- * by RX HW VLAN offload. Clear vlan info. +- */ +- if (netdev_uses_dsa(netdev)) +- __vlan_hwaccel_clear_tag(skb); ++ if (port < ARRAY_SIZE(eth->dsa_meta) && ++ eth->dsa_meta[port]) ++ skb_dst_set_noref(skb, ð->dsa_meta[port]->dst); ++ ++ __vlan_hwaccel_clear_tag(skb); + } + + skb_record_rx_queue(skb, 0); +@@ -2793,15 +2800,30 @@ static netdev_features_t mtk_fix_feature + + static int mtk_set_features(struct net_device *dev, netdev_features_t features) + { +- int err = 0; ++ struct mtk_mac *mac = netdev_priv(dev); ++ struct mtk_eth *eth = mac->hw; ++ netdev_features_t diff = dev->features ^ features; ++ int i; ++ ++ if ((diff & NETIF_F_LRO) && !(features & NETIF_F_LRO)) ++ mtk_hwlro_netdev_disable(dev); + +- if (!((dev->features ^ features) & NETIF_F_LRO)) ++ /* Set RX VLAN offloading */ ++ if (!(diff & NETIF_F_HW_VLAN_CTAG_RX)) + return 0; + +- if (!(features & NETIF_F_LRO)) +- mtk_hwlro_netdev_disable(dev); ++ mtk_w32(eth, !!(features & NETIF_F_HW_VLAN_CTAG_RX), ++ MTK_CDMP_EG_CTRL); + +- return err; ++ /* sync features with other MAC */ ++ for (i = 0; i < MTK_MAC_COUNT; i++) { ++ if (!eth->netdev[i] || eth->netdev[i] == dev) ++ continue; ++ eth->netdev[i]->features &= ~NETIF_F_HW_VLAN_CTAG_RX; ++ eth->netdev[i]->features |= features & NETIF_F_HW_VLAN_CTAG_RX; ++ } ++ ++ return 0; + } + + /* wait for DMA to finish whatever it is doing before we start using it again */ +@@ -3083,11 +3105,45 @@ found: + return NOTIFY_DONE; + } + ++static bool mtk_uses_dsa(struct net_device *dev) ++{ ++#if IS_ENABLED(CONFIG_NET_DSA) ++ return netdev_uses_dsa(dev) && ++ dev->dsa_ptr->tag_ops->proto == DSA_TAG_PROTO_MTK; ++#else ++ return false; ++#endif ++} ++ + static int mtk_open(struct net_device *dev) + { + struct mtk_mac *mac = netdev_priv(dev); + struct mtk_eth *eth = mac->hw; +- int err; ++ int i, err; ++ ++ if (mtk_uses_dsa(dev) && !eth->prog) { ++ for (i = 0; i < ARRAY_SIZE(eth->dsa_meta); i++) { ++ struct metadata_dst *md_dst = eth->dsa_meta[i]; ++ ++ if (md_dst) ++ continue; ++ ++ md_dst = metadata_dst_alloc(0, METADATA_HW_PORT_MUX, ++ GFP_KERNEL); ++ if (!md_dst) ++ return -ENOMEM; ++ ++ md_dst->u.port_info.port_id = i; ++ eth->dsa_meta[i] = md_dst; ++ } ++ } else { ++ /* Hardware special tag parsing needs to be disabled if at least ++ * one MAC does not use DSA. ++ */ ++ u32 val = mtk_r32(eth, MTK_CDMP_IG_CTRL); ++ val &= ~MTK_CDMP_STAG_EN; ++ mtk_w32(eth, val, MTK_CDMP_IG_CTRL); ++ } + + err = phylink_of_phy_connect(mac->phylink, mac->of_node, 0); + if (err) { +@@ -3417,6 +3473,10 @@ static int mtk_hw_init(struct mtk_eth *e + */ + val = mtk_r32(eth, MTK_CDMQ_IG_CTRL); + mtk_w32(eth, val | MTK_CDMQ_STAG_EN, MTK_CDMQ_IG_CTRL); ++ if (!MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) { ++ val = mtk_r32(eth, MTK_CDMP_IG_CTRL); ++ mtk_w32(eth, val | MTK_CDMP_STAG_EN, MTK_CDMP_IG_CTRL); ++ } + + /* Enable RX VLan Offloading */ + mtk_w32(eth, 1, MTK_CDMP_EG_CTRL); +@@ -3634,6 +3694,12 @@ static int mtk_free_dev(struct mtk_eth * + free_netdev(eth->netdev[i]); + } + ++ for (i = 0; i < ARRAY_SIZE(eth->dsa_meta); i++) { ++ if (!eth->dsa_meta[i]) ++ break; ++ metadata_dst_free(eth->dsa_meta[i]); ++ } ++ + return 0; + } + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -22,6 +22,9 @@ + #include + #include "mtk_ppe.h" + ++#define MTK_MAX_DSA_PORTS 7 ++#define MTK_DSA_PORT_MASK GENMASK(2, 0) ++ + #define MTK_QDMA_NUM_QUEUES 16 + #define MTK_QDMA_PAGE_SIZE 2048 + #define MTK_MAX_RX_LENGTH 1536 +@@ -93,6 +96,9 @@ + #define MTK_CDMQ_IG_CTRL 0x1400 + #define MTK_CDMQ_STAG_EN BIT(0) + ++/* CDMQ Exgress Control Register */ ++#define MTK_CDMQ_EG_CTRL 0x1404 ++ + /* CDMP Ingress Control Register */ + #define MTK_CDMP_IG_CTRL 0x400 + #define MTK_CDMP_STAG_EN BIT(0) +@@ -1140,6 +1146,8 @@ struct mtk_eth { + + int ip_align; + ++ struct metadata_dst *dsa_meta[MTK_MAX_DSA_PORTS]; ++ + struct mtk_ppe *ppe[2]; + struct rhashtable flow_table; + diff --git a/target/linux/generic/pending-5.15/732-10-net-ethernet-mtk_eth_soc-work-around-issue-with-send.patch b/target/linux/generic/pending-5.15/732-10-net-ethernet-mtk_eth_soc-work-around-issue-with-send.patch new file mode 100644 index 000000000..ba86686ee --- /dev/null +++ b/target/linux/generic/pending-5.15/732-10-net-ethernet-mtk_eth_soc-work-around-issue-with-send.patch @@ -0,0 +1,78 @@ +From: Felix Fietkau +Date: Thu, 3 Nov 2022 12:38:49 +0100 +Subject: [PATCH] net: ethernet: mtk_eth_soc: work around issue with + sending small fragments + +When frames are sent with very small fragments, the DMA engine appears to +lock up and transmit attempts time out. Fix this by detecting the presence +of small fragments and use skb_gso_segment + skb_linearize to deal with +them + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -1396,12 +1396,28 @@ static void mtk_wake_queue(struct mtk_et + } + } + ++static bool mtk_skb_has_small_frag(struct sk_buff *skb) ++{ ++ int min_size = 16; ++ int i; ++ ++ if (skb_headlen(skb) < min_size) ++ return true; ++ ++ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) ++ if (skb_frag_size(&skb_shinfo(skb)->frags[i]) < min_size) ++ return true; ++ ++ return false; ++} ++ + static netdev_tx_t mtk_start_xmit(struct sk_buff *skb, struct net_device *dev) + { + struct mtk_mac *mac = netdev_priv(dev); + struct mtk_eth *eth = mac->hw; + struct mtk_tx_ring *ring = ð->tx_ring; + struct net_device_stats *stats = &dev->stats; ++ struct sk_buff *segs, *next; + bool gso = false; + int tx_num; + +@@ -1423,6 +1439,17 @@ static netdev_tx_t mtk_start_xmit(struct + return NETDEV_TX_BUSY; + } + ++ if (skb_is_gso(skb) && mtk_skb_has_small_frag(skb)) { ++ segs = skb_gso_segment(skb, dev->features & ~NETIF_F_ALL_TSO); ++ if (IS_ERR(segs)) ++ goto drop; ++ ++ if (segs) { ++ consume_skb(skb); ++ skb = segs; ++ } ++ } ++ + /* TSO: fill MSS info in tcp checksum field */ + if (skb_is_gso(skb)) { + if (skb_cow_head(skb, 0)) { +@@ -1438,8 +1465,13 @@ static netdev_tx_t mtk_start_xmit(struct + } + } + +- if (mtk_tx_map(skb, dev, tx_num, ring, gso) < 0) +- goto drop; ++ skb_list_walk_safe(skb, skb, next) { ++ if ((mtk_skb_has_small_frag(skb) && skb_linearize(skb)) || ++ mtk_tx_map(skb, dev, tx_num, ring, gso) < 0) { ++ stats->tx_dropped++; ++ dev_kfree_skb_any(skb); ++ } ++ } + + if (unlikely(atomic_read(&ring->free_count) <= ring->thresh)) + netif_tx_stop_all_queues(dev); diff --git a/target/linux/generic/pending-5.15/732-11-net-ethernet-mtk_eth_soc-set-NETIF_F_ALL_TSO.patch b/target/linux/generic/pending-5.15/732-11-net-ethernet-mtk_eth_soc-set-NETIF_F_ALL_TSO.patch new file mode 100644 index 000000000..f89ad6adb --- /dev/null +++ b/target/linux/generic/pending-5.15/732-11-net-ethernet-mtk_eth_soc-set-NETIF_F_ALL_TSO.patch @@ -0,0 +1,21 @@ +From: Felix Fietkau +Date: Fri, 28 Oct 2022 12:54:48 +0200 +Subject: [PATCH] net: ethernet: mtk_eth_soc: set NETIF_F_ALL_TSO + +Significantly improves performance by avoiding unnecessary segmentation + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -49,8 +49,7 @@ + NETIF_F_RXCSUM | \ + NETIF_F_HW_VLAN_CTAG_TX | \ + NETIF_F_HW_VLAN_CTAG_RX | \ +- NETIF_F_SG | NETIF_F_TSO | \ +- NETIF_F_TSO6 | \ ++ NETIF_F_SG | NETIF_F_ALL_TSO | \ + NETIF_F_IPV6_CSUM |\ + NETIF_F_HW_TC) + #define MTK_HW_FEATURES_MT7628 (NETIF_F_SG | NETIF_F_RXCSUM) diff --git a/target/linux/generic/pending-5.15/732-12-net-ethernet-mtk_eth_soc-drop-packets-to-WDMA-if-the.patch b/target/linux/generic/pending-5.15/732-12-net-ethernet-mtk_eth_soc-drop-packets-to-WDMA-if-the.patch new file mode 100644 index 000000000..9fa384e6f --- /dev/null +++ b/target/linux/generic/pending-5.15/732-12-net-ethernet-mtk_eth_soc-drop-packets-to-WDMA-if-the.patch @@ -0,0 +1,37 @@ +From: Felix Fietkau +Date: Thu, 3 Nov 2022 17:46:25 +0100 +Subject: [PATCH] net: ethernet: mtk_eth_soc: drop packets to WDMA if the + ring is full + +Improves handling of DMA ring overflow. +Clarify other WDMA drop related comment. + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -3529,9 +3529,12 @@ static int mtk_hw_init(struct mtk_eth *e + mtk_w32(eth, 0x21021000, MTK_FE_INT_GRP); + + if (MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) { +- /* PSE should not drop port8 and port9 packets */ ++ /* PSE should not drop port8 and port9 packets from WDMA Tx */ + mtk_w32(eth, 0x00000300, PSE_DROP_CFG); + ++ /* PSE should drop packets to port 8/9 on WDMA Rx ring full */ ++ mtk_w32(eth, 0x00000300, PSE_PPE0_DROP); ++ + /* PSE Free Queue Flow Control */ + mtk_w32(eth, 0x01fa01f4, PSE_FQFC_CFG2); + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -127,6 +127,7 @@ + #define PSE_FQFC_CFG1 0x100 + #define PSE_FQFC_CFG2 0x104 + #define PSE_DROP_CFG 0x108 ++#define PSE_PPE0_DROP 0x110 + + /* PSE Input Queue Reservation Register*/ + #define PSE_IQ_REV(x) (0x140 + (((x) - 1) << 2)) diff --git a/target/linux/generic/pending-5.15/732-13-net-ethernet-mtk_eth_soc-fix-flow_offload-related-re.patch b/target/linux/generic/pending-5.15/732-13-net-ethernet-mtk_eth_soc-fix-flow_offload-related-re.patch new file mode 100644 index 000000000..0664e0106 --- /dev/null +++ b/target/linux/generic/pending-5.15/732-13-net-ethernet-mtk_eth_soc-fix-flow_offload-related-re.patch @@ -0,0 +1,52 @@ +From: Felix Fietkau +Date: Thu, 17 Nov 2022 11:58:21 +0100 +Subject: [PATCH] net: ethernet: mtk_eth_soc: fix flow_offload related refcount + bug + +Since we call flow_block_cb_decref on FLOW_BLOCK_UNBIND, we need to call +flow_block_cb_incref unconditionally, even for a newly allocated cb. +Fixes a use-after-free bug + +Fixes: 502e84e2382d ("net: ethernet: mtk_eth_soc: add flow offloading support") +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c ++++ b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c +@@ -561,6 +561,7 @@ mtk_eth_setup_tc_block(struct net_device + struct mtk_eth *eth = mac->hw; + static LIST_HEAD(block_cb_list); + struct flow_block_cb *block_cb; ++ bool register_block = false; + flow_setup_cb_t *cb; + + if (!eth->soc->offload_version) +@@ -575,16 +576,20 @@ mtk_eth_setup_tc_block(struct net_device + switch (f->command) { + case FLOW_BLOCK_BIND: + block_cb = flow_block_cb_lookup(f->block, cb, dev); +- if (block_cb) { +- flow_block_cb_incref(block_cb); +- return 0; ++ if (!block_cb) { ++ block_cb = flow_block_cb_alloc(cb, dev, dev, NULL); ++ if (IS_ERR(block_cb)) ++ return PTR_ERR(block_cb); ++ ++ register_block = true; + } +- block_cb = flow_block_cb_alloc(cb, dev, dev, NULL); +- if (IS_ERR(block_cb)) +- return PTR_ERR(block_cb); + +- flow_block_cb_add(block_cb, f); +- list_add_tail(&block_cb->driver_list, &block_cb_list); ++ flow_block_cb_incref(block_cb); ++ ++ if (register_block) { ++ flow_block_cb_add(block_cb, f); ++ list_add_tail(&block_cb->driver_list, &block_cb_list); ++ } + return 0; + case FLOW_BLOCK_UNBIND: + block_cb = flow_block_cb_lookup(f->block, cb, dev); diff --git a/target/linux/generic/pending-5.15/732-14-net-ethernet-mtk_eth_soc-drop-generic-vlan-rx-offloa.patch b/target/linux/generic/pending-5.15/732-14-net-ethernet-mtk_eth_soc-drop-generic-vlan-rx-offloa.patch new file mode 100644 index 000000000..42f98e6af --- /dev/null +++ b/target/linux/generic/pending-5.15/732-14-net-ethernet-mtk_eth_soc-drop-generic-vlan-rx-offloa.patch @@ -0,0 +1,181 @@ +From: Felix Fietkau +Date: Sun, 20 Nov 2022 23:01:00 +0100 +Subject: [PATCH] net: ethernet: mtk_eth_soc: drop generic vlan rx offload, + only use DSA untagging + +Through testing I found out that hardware vlan rx offload support seems to +have some hardware issues. At least when using multiple MACs and when receiving +tagged packets on the secondary MAC, the hardware can sometimes start to emit +wrong tags on the first MAC as well. + +In order to avoid such issues, drop the feature configuration and use the +offload feature only for DSA hardware untagging on MT7621/MT7622 devices which +only use one MAC. + +Signed-off-by: Felix Fietkau +--- + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c +@@ -1993,29 +1993,16 @@ static int mtk_poll_rx(struct napi_struc + if (reason == MTK_PPE_CPU_REASON_HIT_UNBIND_RATE_REACHED) + mtk_ppe_check_skb(eth->ppe[0], skb, hash); + +- if (netdev->features & NETIF_F_HW_VLAN_CTAG_RX) { +- if (MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) { +- if (trxd.rxd3 & RX_DMA_VTAG_V2) +- __vlan_hwaccel_put_tag(skb, +- htons(RX_DMA_VPID(trxd.rxd4)), +- RX_DMA_VID(trxd.rxd4)); +- } else if (trxd.rxd2 & RX_DMA_VTAG) { +- __vlan_hwaccel_put_tag(skb, htons(RX_DMA_VPID(trxd.rxd3)), +- RX_DMA_VID(trxd.rxd3)); +- } +- } +- + /* When using VLAN untagging in combination with DSA, the + * hardware treats the MTK special tag as a VLAN and untags it. + */ +- if (skb_vlan_tag_present(skb) && netdev_uses_dsa(netdev)) { +- unsigned int port = ntohs(skb->vlan_proto) & GENMASK(2, 0); ++ if (!MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2) && ++ (trxd.rxd2 & RX_DMA_VTAG) && netdev_uses_dsa(netdev)) { ++ unsigned int port = RX_DMA_VPID(trxd.rxd3) & GENMASK(2, 0); + + if (port < ARRAY_SIZE(eth->dsa_meta) && + eth->dsa_meta[port]) + skb_dst_set_noref(skb, ð->dsa_meta[port]->dst); +- +- __vlan_hwaccel_clear_tag(skb); + } + + skb_record_rx_queue(skb, 0); +@@ -2832,29 +2819,11 @@ static netdev_features_t mtk_fix_feature + + static int mtk_set_features(struct net_device *dev, netdev_features_t features) + { +- struct mtk_mac *mac = netdev_priv(dev); +- struct mtk_eth *eth = mac->hw; + netdev_features_t diff = dev->features ^ features; +- int i; + + if ((diff & NETIF_F_LRO) && !(features & NETIF_F_LRO)) + mtk_hwlro_netdev_disable(dev); + +- /* Set RX VLAN offloading */ +- if (!(diff & NETIF_F_HW_VLAN_CTAG_RX)) +- return 0; +- +- mtk_w32(eth, !!(features & NETIF_F_HW_VLAN_CTAG_RX), +- MTK_CDMP_EG_CTRL); +- +- /* sync features with other MAC */ +- for (i = 0; i < MTK_MAC_COUNT; i++) { +- if (!eth->netdev[i] || eth->netdev[i] == dev) +- continue; +- eth->netdev[i]->features &= ~NETIF_F_HW_VLAN_CTAG_RX; +- eth->netdev[i]->features |= features & NETIF_F_HW_VLAN_CTAG_RX; +- } +- + return 0; + } + +@@ -3153,30 +3122,6 @@ static int mtk_open(struct net_device *d + struct mtk_eth *eth = mac->hw; + int i, err; + +- if (mtk_uses_dsa(dev) && !eth->prog) { +- for (i = 0; i < ARRAY_SIZE(eth->dsa_meta); i++) { +- struct metadata_dst *md_dst = eth->dsa_meta[i]; +- +- if (md_dst) +- continue; +- +- md_dst = metadata_dst_alloc(0, METADATA_HW_PORT_MUX, +- GFP_KERNEL); +- if (!md_dst) +- return -ENOMEM; +- +- md_dst->u.port_info.port_id = i; +- eth->dsa_meta[i] = md_dst; +- } +- } else { +- /* Hardware special tag parsing needs to be disabled if at least +- * one MAC does not use DSA. +- */ +- u32 val = mtk_r32(eth, MTK_CDMP_IG_CTRL); +- val &= ~MTK_CDMP_STAG_EN; +- mtk_w32(eth, val, MTK_CDMP_IG_CTRL); +- } +- + err = phylink_of_phy_connect(mac->phylink, mac->of_node, 0); + if (err) { + netdev_err(dev, "%s: could not attach PHY: %d\n", __func__, +@@ -3215,6 +3160,35 @@ static int mtk_open(struct net_device *d + phylink_start(mac->phylink); + netif_tx_start_all_queues(dev); + ++ if (MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) ++ return 0; ++ ++ if (mtk_uses_dsa(dev) && !eth->prog) { ++ for (i = 0; i < ARRAY_SIZE(eth->dsa_meta); i++) { ++ struct metadata_dst *md_dst = eth->dsa_meta[i]; ++ ++ if (md_dst) ++ continue; ++ ++ md_dst = metadata_dst_alloc(0, METADATA_HW_PORT_MUX, ++ GFP_KERNEL); ++ if (!md_dst) ++ return -ENOMEM; ++ ++ md_dst->u.port_info.port_id = i; ++ eth->dsa_meta[i] = md_dst; ++ } ++ } else { ++ /* Hardware special tag parsing needs to be disabled if at least ++ * one MAC does not use DSA. ++ */ ++ u32 val = mtk_r32(eth, MTK_CDMP_IG_CTRL); ++ val &= ~MTK_CDMP_STAG_EN; ++ mtk_w32(eth, val, MTK_CDMP_IG_CTRL); ++ ++ mtk_w32(eth, 0, MTK_CDMP_EG_CTRL); ++ } ++ + return 0; + } + +@@ -3508,10 +3482,9 @@ static int mtk_hw_init(struct mtk_eth *e + if (!MTK_HAS_CAPS(eth->soc->caps, MTK_NETSYS_V2)) { + val = mtk_r32(eth, MTK_CDMP_IG_CTRL); + mtk_w32(eth, val | MTK_CDMP_STAG_EN, MTK_CDMP_IG_CTRL); +- } + +- /* Enable RX VLan Offloading */ +- mtk_w32(eth, 1, MTK_CDMP_EG_CTRL); ++ mtk_w32(eth, 1, MTK_CDMP_EG_CTRL); ++ } + + /* set interrupt delays based on current Net DIM sample */ + mtk_dim_rx(ð->rx_dim.work); +@@ -4132,7 +4105,7 @@ static int mtk_add_mac(struct mtk_eth *e + eth->netdev[id]->hw_features |= NETIF_F_LRO; + + eth->netdev[id]->vlan_features = eth->soc->hw_features & +- ~(NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX); ++ ~NETIF_F_HW_VLAN_CTAG_TX; + eth->netdev[id]->features |= eth->soc->hw_features; + eth->netdev[id]->ethtool_ops = &mtk_ethtool_ops; + +--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h ++++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h +@@ -48,7 +48,6 @@ + #define MTK_HW_FEATURES (NETIF_F_IP_CSUM | \ + NETIF_F_RXCSUM | \ + NETIF_F_HW_VLAN_CTAG_TX | \ +- NETIF_F_HW_VLAN_CTAG_RX | \ + NETIF_F_SG | NETIF_F_ALL_TSO | \ + NETIF_F_IPV6_CSUM |\ + NETIF_F_HW_TC) diff --git a/target/linux/generic/pending-5.15/733-01-net-ethernet-mtk_wed-introduce-wed-mcu-support.patch b/target/linux/generic/pending-5.15/733-01-net-ethernet-mtk_wed-introduce-wed-mcu-support.patch new file mode 100644 index 000000000..c48613929 --- /dev/null +++ b/target/linux/generic/pending-5.15/733-01-net-ethernet-mtk_wed-introduce-wed-mcu-support.patch @@ -0,0 +1,591 @@ +From: Sujuan Chen +Date: Sat, 5 Nov 2022 23:36:18 +0100 +Subject: [PATCH] net: ethernet: mtk_wed: introduce wed mcu support + +Introduce WED mcu support used to configure WED WO chip. +This is a preliminary patch in order to add RX Wireless +Ethernet Dispatch available on MT7986 SoC. + +Tested-by: Daniel Golle +Co-developed-by: Lorenzo Bianconi +Signed-off-by: Lorenzo Bianconi +Signed-off-by: Sujuan Chen +Signed-off-by: David S. Miller +--- + create mode 100644 drivers/net/ethernet/mediatek/mtk_wed_mcu.c + create mode 100644 drivers/net/ethernet/mediatek/mtk_wed_wo.h + +--- a/drivers/net/ethernet/mediatek/Makefile ++++ b/drivers/net/ethernet/mediatek/Makefile +@@ -5,7 +5,7 @@ + + obj-$(CONFIG_NET_MEDIATEK_SOC) += mtk_eth.o + mtk_eth-y := mtk_eth_soc.o mtk_sgmii.o mtk_eth_path.o mtk_ppe.o mtk_ppe_debugfs.o mtk_ppe_offload.o +-mtk_eth-$(CONFIG_NET_MEDIATEK_SOC_WED) += mtk_wed.o ++mtk_eth-$(CONFIG_NET_MEDIATEK_SOC_WED) += mtk_wed.o mtk_wed_mcu.o + ifdef CONFIG_DEBUG_FS + mtk_eth-$(CONFIG_NET_MEDIATEK_SOC_WED) += mtk_wed_debugfs.o + endif +--- /dev/null ++++ b/drivers/net/ethernet/mediatek/mtk_wed_mcu.c +@@ -0,0 +1,359 @@ ++// SPDX-License-Identifier: GPL-2.0-only ++/* Copyright (C) 2022 MediaTek Inc. ++ * ++ * Author: Lorenzo Bianconi ++ * Sujuan Chen ++ */ ++ ++#include ++#include ++#include ++#include ++#include ++ ++#include "mtk_wed_regs.h" ++#include "mtk_wed_wo.h" ++#include "mtk_wed.h" ++ ++static u32 wo_r32(struct mtk_wed_wo *wo, u32 reg) ++{ ++ return readl(wo->boot.addr + reg); ++} ++ ++static void wo_w32(struct mtk_wed_wo *wo, u32 reg, u32 val) ++{ ++ writel(val, wo->boot.addr + reg); ++} ++ ++static struct sk_buff * ++mtk_wed_mcu_msg_alloc(const void *data, int data_len) ++{ ++ int length = sizeof(struct mtk_wed_mcu_hdr) + data_len; ++ struct sk_buff *skb; ++ ++ skb = alloc_skb(length, GFP_KERNEL); ++ if (!skb) ++ return NULL; ++ ++ memset(skb->head, 0, length); ++ skb_reserve(skb, sizeof(struct mtk_wed_mcu_hdr)); ++ if (data && data_len) ++ skb_put_data(skb, data, data_len); ++ ++ return skb; ++} ++ ++static struct sk_buff * ++mtk_wed_mcu_get_response(struct mtk_wed_wo *wo, unsigned long expires) ++{ ++ if (!time_is_after_jiffies(expires)) ++ return NULL; ++ ++ wait_event_timeout(wo->mcu.wait, !skb_queue_empty(&wo->mcu.res_q), ++ expires - jiffies); ++ return skb_dequeue(&wo->mcu.res_q); ++} ++ ++void mtk_wed_mcu_rx_event(struct mtk_wed_wo *wo, struct sk_buff *skb) ++{ ++ skb_queue_tail(&wo->mcu.res_q, skb); ++ wake_up(&wo->mcu.wait); ++} ++ ++void mtk_wed_mcu_rx_unsolicited_event(struct mtk_wed_wo *wo, ++ struct sk_buff *skb) ++{ ++ struct mtk_wed_mcu_hdr *hdr = (struct mtk_wed_mcu_hdr *)skb->data; ++ ++ switch (hdr->cmd) { ++ case MTK_WED_WO_EVT_LOG_DUMP: { ++ const char *msg = (const char *)(skb->data + sizeof(*hdr)); ++ ++ dev_notice(wo->hw->dev, "%s\n", msg); ++ break; ++ } ++ case MTK_WED_WO_EVT_PROFILING: { ++ struct mtk_wed_wo_log_info *info; ++ u32 count = (skb->len - sizeof(*hdr)) / sizeof(*info); ++ int i; ++ ++ info = (struct mtk_wed_wo_log_info *)(skb->data + sizeof(*hdr)); ++ for (i = 0 ; i < count ; i++) ++ dev_notice(wo->hw->dev, ++ "SN:%u latency: total=%u, rro:%u, mod:%u\n", ++ le32_to_cpu(info[i].sn), ++ le32_to_cpu(info[i].total), ++ le32_to_cpu(info[i].rro), ++ le32_to_cpu(info[i].mod)); ++ break; ++ } ++ case MTK_WED_WO_EVT_RXCNT_INFO: ++ break; ++ default: ++ break; ++ } ++ ++ dev_kfree_skb(skb); ++} ++ ++static int ++mtk_wed_mcu_skb_send_msg(struct mtk_wed_wo *wo, struct sk_buff *skb, ++ int id, int cmd, u16 *wait_seq, bool wait_resp) ++{ ++ struct mtk_wed_mcu_hdr *hdr; ++ ++ /* TODO: make it dynamic based on cmd */ ++ wo->mcu.timeout = 20 * HZ; ++ ++ hdr = (struct mtk_wed_mcu_hdr *)skb_push(skb, sizeof(*hdr)); ++ hdr->cmd = cmd; ++ hdr->length = cpu_to_le16(skb->len); ++ ++ if (wait_resp && wait_seq) { ++ u16 seq = ++wo->mcu.seq; ++ ++ if (!seq) ++ seq = ++wo->mcu.seq; ++ *wait_seq = seq; ++ ++ hdr->flag |= cpu_to_le16(MTK_WED_WARP_CMD_FLAG_NEED_RSP); ++ hdr->seq = cpu_to_le16(seq); ++ } ++ if (id == MTK_WED_MODULE_ID_WO) ++ hdr->flag |= cpu_to_le16(MTK_WED_WARP_CMD_FLAG_FROM_TO_WO); ++ ++ dev_kfree_skb(skb); ++ return 0; ++} ++ ++static int ++mtk_wed_mcu_parse_response(struct mtk_wed_wo *wo, struct sk_buff *skb, ++ int cmd, int seq) ++{ ++ struct mtk_wed_mcu_hdr *hdr; ++ ++ if (!skb) { ++ dev_err(wo->hw->dev, "Message %08x (seq %d) timeout\n", ++ cmd, seq); ++ return -ETIMEDOUT; ++ } ++ ++ hdr = (struct mtk_wed_mcu_hdr *)skb->data; ++ if (le16_to_cpu(hdr->seq) != seq) ++ return -EAGAIN; ++ ++ skb_pull(skb, sizeof(*hdr)); ++ switch (cmd) { ++ case MTK_WED_WO_CMD_RXCNT_INFO: ++ default: ++ break; ++ } ++ ++ return 0; ++} ++ ++int mtk_wed_mcu_send_msg(struct mtk_wed_wo *wo, int id, int cmd, ++ const void *data, int len, bool wait_resp) ++{ ++ unsigned long expires; ++ struct sk_buff *skb; ++ u16 seq; ++ int ret; ++ ++ skb = mtk_wed_mcu_msg_alloc(data, len); ++ if (!skb) ++ return -ENOMEM; ++ ++ mutex_lock(&wo->mcu.mutex); ++ ++ ret = mtk_wed_mcu_skb_send_msg(wo, skb, id, cmd, &seq, wait_resp); ++ if (ret || !wait_resp) ++ goto unlock; ++ ++ expires = jiffies + wo->mcu.timeout; ++ do { ++ skb = mtk_wed_mcu_get_response(wo, expires); ++ ret = mtk_wed_mcu_parse_response(wo, skb, cmd, seq); ++ dev_kfree_skb(skb); ++ } while (ret == -EAGAIN); ++ ++unlock: ++ mutex_unlock(&wo->mcu.mutex); ++ ++ return ret; ++} ++ ++static int ++mtk_wed_get_memory_region(struct mtk_wed_wo *wo, ++ struct mtk_wed_wo_memory_region *region) ++{ ++ struct reserved_mem *rmem; ++ struct device_node *np; ++ int index; ++ ++ index = of_property_match_string(wo->hw->node, "memory-region-names", ++ region->name); ++ if (index < 0) ++ return index; ++ ++ np = of_parse_phandle(wo->hw->node, "memory-region", index); ++ if (!np) ++ return -ENODEV; ++ ++ rmem = of_reserved_mem_lookup(np); ++ of_node_put(np); ++ ++ if (!rmem) ++ return -ENODEV; ++ ++ region->phy_addr = rmem->base; ++ region->size = rmem->size; ++ region->addr = devm_ioremap(wo->hw->dev, region->phy_addr, region->size); ++ ++ return !region->addr ? -EINVAL : 0; ++} ++ ++static int ++mtk_wed_mcu_run_firmware(struct mtk_wed_wo *wo, const struct firmware *fw, ++ struct mtk_wed_wo_memory_region *region) ++{ ++ const u8 *first_region_ptr, *region_ptr, *trailer_ptr, *ptr = fw->data; ++ const struct mtk_wed_fw_trailer *trailer; ++ const struct mtk_wed_fw_region *fw_region; ++ ++ trailer_ptr = fw->data + fw->size - sizeof(*trailer); ++ trailer = (const struct mtk_wed_fw_trailer *)trailer_ptr; ++ region_ptr = trailer_ptr - trailer->num_region * sizeof(*fw_region); ++ first_region_ptr = region_ptr; ++ ++ while (region_ptr < trailer_ptr) { ++ u32 length; ++ ++ fw_region = (const struct mtk_wed_fw_region *)region_ptr; ++ length = le32_to_cpu(fw_region->len); ++ ++ if (region->phy_addr != le32_to_cpu(fw_region->addr)) ++ goto next; ++ ++ if (region->size < length) ++ goto next; ++ ++ if (first_region_ptr < ptr + length) ++ goto next; ++ ++ if (region->shared && region->consumed) ++ return 0; ++ ++ if (!region->shared || !region->consumed) { ++ memcpy_toio(region->addr, ptr, length); ++ region->consumed = true; ++ return 0; ++ } ++next: ++ region_ptr += sizeof(*fw_region); ++ ptr += length; ++ } ++ ++ return -EINVAL; ++} ++ ++static int ++mtk_wed_mcu_load_firmware(struct mtk_wed_wo *wo) ++{ ++ static struct mtk_wed_wo_memory_region mem_region[] = { ++ [MTK_WED_WO_REGION_EMI] = { ++ .name = "wo-emi", ++ }, ++ [MTK_WED_WO_REGION_ILM] = { ++ .name = "wo-ilm", ++ }, ++ [MTK_WED_WO_REGION_DATA] = { ++ .name = "wo-data", ++ .shared = true, ++ }, ++ }; ++ const struct mtk_wed_fw_trailer *trailer; ++ const struct firmware *fw; ++ const char *fw_name; ++ u32 val, boot_cr; ++ int ret, i; ++ ++ /* load firmware region metadata */ ++ for (i = 0; i < ARRAY_SIZE(mem_region); i++) { ++ ret = mtk_wed_get_memory_region(wo, &mem_region[i]); ++ if (ret) ++ return ret; ++ } ++ ++ wo->boot.name = "wo-boot"; ++ ret = mtk_wed_get_memory_region(wo, &wo->boot); ++ if (ret) ++ return ret; ++ ++ /* set dummy cr */ ++ wed_w32(wo->hw->wed_dev, MTK_WED_SCR0 + 4 * MTK_WED_DUMMY_CR_FWDL, ++ wo->hw->index + 1); ++ ++ /* load firmware */ ++ fw_name = wo->hw->index ? MT7986_FIRMWARE_WO1 : MT7986_FIRMWARE_WO0; ++ ret = request_firmware(&fw, fw_name, wo->hw->dev); ++ if (ret) ++ return ret; ++ ++ trailer = (void *)(fw->data + fw->size - ++ sizeof(struct mtk_wed_fw_trailer)); ++ dev_info(wo->hw->dev, ++ "MTK WED WO Firmware Version: %.10s, Build Time: %.15s\n", ++ trailer->fw_ver, trailer->build_date); ++ dev_info(wo->hw->dev, "MTK WED WO Chip ID %02x Region %d\n", ++ trailer->chip_id, trailer->num_region); ++ ++ for (i = 0; i < ARRAY_SIZE(mem_region); i++) { ++ ret = mtk_wed_mcu_run_firmware(wo, fw, &mem_region[i]); ++ if (ret) ++ goto out; ++ } ++ ++ /* set the start address */ ++ boot_cr = wo->hw->index ? MTK_WO_MCU_CFG_LS_WA_BOOT_ADDR_ADDR ++ : MTK_WO_MCU_CFG_LS_WM_BOOT_ADDR_ADDR; ++ wo_w32(wo, boot_cr, mem_region[MTK_WED_WO_REGION_EMI].phy_addr >> 16); ++ /* wo firmware reset */ ++ wo_w32(wo, MTK_WO_MCU_CFG_LS_WF_MCCR_CLR_ADDR, 0xc00); ++ ++ val = wo_r32(wo, MTK_WO_MCU_CFG_LS_WF_MCU_CFG_WM_WA_ADDR); ++ val |= wo->hw->index ? MTK_WO_MCU_CFG_LS_WF_WM_WA_WA_CPU_RSTB_MASK ++ : MTK_WO_MCU_CFG_LS_WF_WM_WA_WM_CPU_RSTB_MASK; ++ wo_w32(wo, MTK_WO_MCU_CFG_LS_WF_MCU_CFG_WM_WA_ADDR, val); ++out: ++ release_firmware(fw); ++ ++ return ret; ++} ++ ++static u32 ++mtk_wed_mcu_read_fw_dl(struct mtk_wed_wo *wo) ++{ ++ return wed_r32(wo->hw->wed_dev, ++ MTK_WED_SCR0 + 4 * MTK_WED_DUMMY_CR_FWDL); ++} ++ ++int mtk_wed_mcu_init(struct mtk_wed_wo *wo) ++{ ++ u32 val; ++ int ret; ++ ++ skb_queue_head_init(&wo->mcu.res_q); ++ init_waitqueue_head(&wo->mcu.wait); ++ mutex_init(&wo->mcu.mutex); ++ ++ ret = mtk_wed_mcu_load_firmware(wo); ++ if (ret) ++ return ret; ++ ++ return readx_poll_timeout(mtk_wed_mcu_read_fw_dl, wo, val, !val, ++ 100, MTK_FW_DL_TIMEOUT); ++} ++ ++MODULE_FIRMWARE(MT7986_FIRMWARE_WO0); ++MODULE_FIRMWARE(MT7986_FIRMWARE_WO1); +--- a/drivers/net/ethernet/mediatek/mtk_wed_regs.h ++++ b/drivers/net/ethernet/mediatek/mtk_wed_regs.h +@@ -152,6 +152,7 @@ struct mtk_wdma_desc { + + #define MTK_WED_RING_RX(_n) (0x400 + (_n) * 0x10) + ++#define MTK_WED_SCR0 0x3c0 + #define MTK_WED_WPDMA_INT_TRIGGER 0x504 + #define MTK_WED_WPDMA_INT_TRIGGER_RX_DONE BIT(1) + #define MTK_WED_WPDMA_INT_TRIGGER_TX_DONE GENMASK(5, 4) +--- /dev/null ++++ b/drivers/net/ethernet/mediatek/mtk_wed_wo.h +@@ -0,0 +1,150 @@ ++/* SPDX-License-Identifier: GPL-2.0-only */ ++/* Copyright (C) 2022 Lorenzo Bianconi */ ++ ++#ifndef __MTK_WED_WO_H ++#define __MTK_WED_WO_H ++ ++#include ++#include ++ ++struct mtk_wed_hw; ++ ++struct mtk_wed_mcu_hdr { ++ /* DW0 */ ++ u8 version; ++ u8 cmd; ++ __le16 length; ++ ++ /* DW1 */ ++ __le16 seq; ++ __le16 flag; ++ ++ /* DW2 */ ++ __le32 status; ++ ++ /* DW3 */ ++ u8 rsv[20]; ++}; ++ ++struct mtk_wed_wo_log_info { ++ __le32 sn; ++ __le32 total; ++ __le32 rro; ++ __le32 mod; ++}; ++ ++enum mtk_wed_wo_event { ++ MTK_WED_WO_EVT_LOG_DUMP = 0x1, ++ MTK_WED_WO_EVT_PROFILING = 0x2, ++ MTK_WED_WO_EVT_RXCNT_INFO = 0x3, ++}; ++ ++#define MTK_WED_MODULE_ID_WO 1 ++#define MTK_FW_DL_TIMEOUT 4000000 /* us */ ++#define MTK_WOCPU_TIMEOUT 2000000 /* us */ ++ ++enum { ++ MTK_WED_WARP_CMD_FLAG_RSP = BIT(0), ++ MTK_WED_WARP_CMD_FLAG_NEED_RSP = BIT(1), ++ MTK_WED_WARP_CMD_FLAG_FROM_TO_WO = BIT(2), ++}; ++ ++enum { ++ MTK_WED_WO_REGION_EMI, ++ MTK_WED_WO_REGION_ILM, ++ MTK_WED_WO_REGION_DATA, ++ MTK_WED_WO_REGION_BOOT, ++ __MTK_WED_WO_REGION_MAX, ++}; ++ ++enum mtk_wed_dummy_cr_idx { ++ MTK_WED_DUMMY_CR_FWDL, ++ MTK_WED_DUMMY_CR_WO_STATUS, ++}; ++ ++#define MT7986_FIRMWARE_WO0 "mediatek/mt7986_wo_0.bin" ++#define MT7986_FIRMWARE_WO1 "mediatek/mt7986_wo_1.bin" ++ ++#define MTK_WO_MCU_CFG_LS_BASE 0 ++#define MTK_WO_MCU_CFG_LS_HW_VER_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x000) ++#define MTK_WO_MCU_CFG_LS_FW_VER_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x004) ++#define MTK_WO_MCU_CFG_LS_CFG_DBG1_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x00c) ++#define MTK_WO_MCU_CFG_LS_CFG_DBG2_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x010) ++#define MTK_WO_MCU_CFG_LS_WF_MCCR_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x014) ++#define MTK_WO_MCU_CFG_LS_WF_MCCR_SET_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x018) ++#define MTK_WO_MCU_CFG_LS_WF_MCCR_CLR_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x01c) ++#define MTK_WO_MCU_CFG_LS_WF_MCU_CFG_WM_WA_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x050) ++#define MTK_WO_MCU_CFG_LS_WM_BOOT_ADDR_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x060) ++#define MTK_WO_MCU_CFG_LS_WA_BOOT_ADDR_ADDR (MTK_WO_MCU_CFG_LS_BASE + 0x064) ++ ++#define MTK_WO_MCU_CFG_LS_WF_WM_WA_WM_CPU_RSTB_MASK BIT(5) ++#define MTK_WO_MCU_CFG_LS_WF_WM_WA_WA_CPU_RSTB_MASK BIT(0) ++ ++struct mtk_wed_wo_memory_region { ++ const char *name; ++ void __iomem *addr; ++ phys_addr_t phy_addr; ++ u32 size; ++ bool shared:1; ++ bool consumed:1; ++}; ++ ++struct mtk_wed_fw_region { ++ __le32 decomp_crc; ++ __le32 decomp_len; ++ __le32 decomp_blk_sz; ++ u8 rsv0[4]; ++ __le32 addr; ++ __le32 len; ++ u8 feature_set; ++ u8 rsv1[15]; ++} __packed; ++ ++struct mtk_wed_fw_trailer { ++ u8 chip_id; ++ u8 eco_code; ++ u8 num_region; ++ u8 format_ver; ++ u8 format_flag; ++ u8 rsv[2]; ++ char fw_ver[10]; ++ char build_date[15]; ++ u32 crc; ++}; ++ ++struct mtk_wed_wo { ++ struct mtk_wed_hw *hw; ++ struct mtk_wed_wo_memory_region boot; ++ ++ struct { ++ struct mutex mutex; ++ int timeout; ++ u16 seq; ++ ++ struct sk_buff_head res_q; ++ wait_queue_head_t wait; ++ } mcu; ++}; ++ ++static inline int ++mtk_wed_mcu_check_msg(struct mtk_wed_wo *wo, struct sk_buff *skb) ++{ ++ struct mtk_wed_mcu_hdr *hdr = (struct mtk_wed_mcu_hdr *)skb->data; ++ ++ if (hdr->version) ++ return -EINVAL; ++ ++ if (skb->len < sizeof(*hdr) || skb->len != le16_to_cpu(hdr->length)) ++ return -EINVAL; ++ ++ return 0; ++} ++ ++void mtk_wed_mcu_rx_event(struct mtk_wed_wo *wo, struct sk_buff *skb); ++void mtk_wed_mcu_rx_unsolicited_event(struct mtk_wed_wo *wo, ++ struct sk_buff *skb); ++int mtk_wed_mcu_send_msg(struct mtk_wed_wo *wo, int id, int cmd, ++ const void *data, int len, bool wait_resp); ++int mtk_wed_mcu_init(struct mtk_wed_wo *wo); ++ ++#endif /* __MTK_WED_WO_H */ +--- a/include/linux/soc/mediatek/mtk_wed.h ++++ b/include/linux/soc/mediatek/mtk_wed.h +@@ -11,6 +11,35 @@ + struct mtk_wed_hw; + struct mtk_wdma_desc; + ++enum mtk_wed_wo_cmd { ++ MTK_WED_WO_CMD_WED_CFG, ++ MTK_WED_WO_CMD_WED_RX_STAT, ++ MTK_WED_WO_CMD_RRO_SER, ++ MTK_WED_WO_CMD_DBG_INFO, ++ MTK_WED_WO_CMD_DEV_INFO, ++ MTK_WED_WO_CMD_BSS_INFO, ++ MTK_WED_WO_CMD_STA_REC, ++ MTK_WED_WO_CMD_DEV_INFO_DUMP, ++ MTK_WED_WO_CMD_BSS_INFO_DUMP, ++ MTK_WED_WO_CMD_STA_REC_DUMP, ++ MTK_WED_WO_CMD_BA_INFO_DUMP, ++ MTK_WED_WO_CMD_FBCMD_Q_DUMP, ++ MTK_WED_WO_CMD_FW_LOG_CTRL, ++ MTK_WED_WO_CMD_LOG_FLUSH, ++ MTK_WED_WO_CMD_CHANGE_STATE, ++ MTK_WED_WO_CMD_CPU_STATS_ENABLE, ++ MTK_WED_WO_CMD_CPU_STATS_DUMP, ++ MTK_WED_WO_CMD_EXCEPTION_INIT, ++ MTK_WED_WO_CMD_PROF_CTRL, ++ MTK_WED_WO_CMD_STA_BA_DUMP, ++ MTK_WED_WO_CMD_BA_CTRL_DUMP, ++ MTK_WED_WO_CMD_RXCNT_CTRL, ++ MTK_WED_WO_CMD_RXCNT_INFO, ++ MTK_WED_WO_CMD_SET_CAP, ++ MTK_WED_WO_CMD_CCIF_RING_DUMP, ++ MTK_WED_WO_CMD_WED_END ++}; ++ + enum mtk_wed_bus_tye { + MTK_WED_BUS_PCIE, + MTK_WED_BUS_AXI, diff --git a/target/linux/generic/pending-5.15/733-02-net-ethernet-mtk_wed-introduce-wed-wo-support.patch b/target/linux/generic/pending-5.15/733-02-net-ethernet-mtk_wed-introduce-wed-wo-support.patch new file mode 100644 index 000000000..dbd7e30fb --- /dev/null +++ b/target/linux/generic/pending-5.15/733-02-net-ethernet-mtk_wed-introduce-wed-wo-support.patch @@ -0,0 +1,737 @@ +From: Lorenzo Bianconi +Date: Sat, 5 Nov 2022 23:36:19 +0100 +Subject: [PATCH] net: ethernet: mtk_wed: introduce wed wo support + +Introduce WO chip support to mtk wed driver. MTK WED WO is used to +implement RX Wireless Ethernet Dispatch and offload traffic received by +wlan nic to the wired interface. + +Tested-by: Daniel Golle +Co-developed-by: Sujuan Chen +Signed-off-by: Sujuan Chen +Signed-off-by: Lorenzo Bianconi +Signed-off-by: David S. Miller +--- + create mode 100644 drivers/net/ethernet/mediatek/mtk_wed_wo.c + +--- a/drivers/net/ethernet/mediatek/Makefile ++++ b/drivers/net/ethernet/mediatek/Makefile +@@ -5,7 +5,7 @@ + + obj-$(CONFIG_NET_MEDIATEK_SOC) += mtk_eth.o + mtk_eth-y := mtk_eth_soc.o mtk_sgmii.o mtk_eth_path.o mtk_ppe.o mtk_ppe_debugfs.o mtk_ppe_offload.o +-mtk_eth-$(CONFIG_NET_MEDIATEK_SOC_WED) += mtk_wed.o mtk_wed_mcu.o ++mtk_eth-$(CONFIG_NET_MEDIATEK_SOC_WED) += mtk_wed.o mtk_wed_mcu.o mtk_wed_wo.o + ifdef CONFIG_DEBUG_FS + mtk_eth-$(CONFIG_NET_MEDIATEK_SOC_WED) += mtk_wed_debugfs.o + endif +--- a/drivers/net/ethernet/mediatek/mtk_wed.c ++++ b/drivers/net/ethernet/mediatek/mtk_wed.c +@@ -16,6 +16,7 @@ + #include "mtk_wed_regs.h" + #include "mtk_wed.h" + #include "mtk_ppe.h" ++#include "mtk_wed_wo.h" + + #define MTK_PCIE_BASE(n) (0x1a143000 + (n) * 0x2000) + +@@ -355,6 +356,8 @@ mtk_wed_detach(struct mtk_wed_device *de + + mtk_wed_free_buffer(dev); + mtk_wed_free_tx_rings(dev); ++ if (hw->version != 1) ++ mtk_wed_wo_deinit(hw); + + if (dev->wlan.bus_type == MTK_WED_BUS_PCIE) { + struct device_node *wlan_node; +@@ -885,9 +888,11 @@ mtk_wed_attach(struct mtk_wed_device *de + } + + mtk_wed_hw_init_early(dev); +- if (hw->hifsys) ++ if (hw->version == 1) + regmap_update_bits(hw->hifsys, HIFSYS_DMA_AG_MAP, + BIT(hw->index), 0); ++ else ++ ret = mtk_wed_wo_init(hw); + + out: + mutex_unlock(&hw_lock); +--- a/drivers/net/ethernet/mediatek/mtk_wed.h ++++ b/drivers/net/ethernet/mediatek/mtk_wed.h +@@ -10,6 +10,7 @@ + #include + + struct mtk_eth; ++struct mtk_wed_wo; + + struct mtk_wed_hw { + struct device_node *node; +@@ -22,6 +23,7 @@ struct mtk_wed_hw { + struct regmap *mirror; + struct dentry *debugfs_dir; + struct mtk_wed_device *wed_dev; ++ struct mtk_wed_wo *wed_wo; + u32 debugfs_reg; + u32 num_flows; + u8 version; +--- a/drivers/net/ethernet/mediatek/mtk_wed_mcu.c ++++ b/drivers/net/ethernet/mediatek/mtk_wed_mcu.c +@@ -122,8 +122,7 @@ mtk_wed_mcu_skb_send_msg(struct mtk_wed_ + if (id == MTK_WED_MODULE_ID_WO) + hdr->flag |= cpu_to_le16(MTK_WED_WARP_CMD_FLAG_FROM_TO_WO); + +- dev_kfree_skb(skb); +- return 0; ++ return mtk_wed_wo_queue_tx_skb(wo, &wo->q_tx, skb); + } + + static int +--- /dev/null ++++ b/drivers/net/ethernet/mediatek/mtk_wed_wo.c +@@ -0,0 +1,508 @@ ++// SPDX-License-Identifier: GPL-2.0-only ++/* Copyright (C) 2022 MediaTek Inc. ++ * ++ * Author: Lorenzo Bianconi ++ * Sujuan Chen ++ */ ++ ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++ ++#include "mtk_wed.h" ++#include "mtk_wed_regs.h" ++#include "mtk_wed_wo.h" ++ ++static u32 ++mtk_wed_mmio_r32(struct mtk_wed_wo *wo, u32 reg) ++{ ++ u32 val; ++ ++ if (regmap_read(wo->mmio.regs, reg, &val)) ++ val = ~0; ++ ++ return val; ++} ++ ++static void ++mtk_wed_mmio_w32(struct mtk_wed_wo *wo, u32 reg, u32 val) ++{ ++ regmap_write(wo->mmio.regs, reg, val); ++} ++ ++static u32 ++mtk_wed_wo_get_isr(struct mtk_wed_wo *wo) ++{ ++ u32 val = mtk_wed_mmio_r32(wo, MTK_WED_WO_CCIF_RCHNUM); ++ ++ return val & MTK_WED_WO_CCIF_RCHNUM_MASK; ++} ++ ++static void ++mtk_wed_wo_set_isr(struct mtk_wed_wo *wo, u32 mask) ++{ ++ mtk_wed_mmio_w32(wo, MTK_WED_WO_CCIF_IRQ0_MASK, mask); ++} ++ ++static void ++mtk_wed_wo_set_ack(struct mtk_wed_wo *wo, u32 mask) ++{ ++ mtk_wed_mmio_w32(wo, MTK_WED_WO_CCIF_ACK, mask); ++} ++ ++static void ++mtk_wed_wo_set_isr_mask(struct mtk_wed_wo *wo, u32 mask, u32 val, bool set) ++{ ++ unsigned long flags; ++ ++ spin_lock_irqsave(&wo->mmio.lock, flags); ++ wo->mmio.irq_mask &= ~mask; ++ wo->mmio.irq_mask |= val; ++ if (set) ++ mtk_wed_wo_set_isr(wo, wo->mmio.irq_mask); ++ spin_unlock_irqrestore(&wo->mmio.lock, flags); ++} ++ ++static void ++mtk_wed_wo_irq_enable(struct mtk_wed_wo *wo, u32 mask) ++{ ++ mtk_wed_wo_set_isr_mask(wo, 0, mask, false); ++ tasklet_schedule(&wo->mmio.irq_tasklet); ++} ++ ++static void ++mtk_wed_wo_irq_disable(struct mtk_wed_wo *wo, u32 mask) ++{ ++ mtk_wed_wo_set_isr_mask(wo, mask, 0, true); ++} ++ ++static void ++mtk_wed_wo_kickout(struct mtk_wed_wo *wo) ++{ ++ mtk_wed_mmio_w32(wo, MTK_WED_WO_CCIF_BUSY, 1 << MTK_WED_WO_TXCH_NUM); ++ mtk_wed_mmio_w32(wo, MTK_WED_WO_CCIF_TCHNUM, MTK_WED_WO_TXCH_NUM); ++} ++ ++static void ++mtk_wed_wo_queue_kick(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q, ++ u32 val) ++{ ++ wmb(); ++ mtk_wed_mmio_w32(wo, q->regs.cpu_idx, val); ++} ++ ++static void * ++mtk_wed_wo_dequeue(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q, u32 *len, ++ bool flush) ++{ ++ int buf_len = SKB_WITH_OVERHEAD(q->buf_size); ++ int index = (q->tail + 1) % q->n_desc; ++ struct mtk_wed_wo_queue_entry *entry; ++ struct mtk_wed_wo_queue_desc *desc; ++ void *buf; ++ ++ if (!q->queued) ++ return NULL; ++ ++ if (flush) ++ q->desc[index].ctrl |= cpu_to_le32(MTK_WED_WO_CTL_DMA_DONE); ++ else if (!(q->desc[index].ctrl & cpu_to_le32(MTK_WED_WO_CTL_DMA_DONE))) ++ return NULL; ++ ++ q->tail = index; ++ q->queued--; ++ ++ desc = &q->desc[index]; ++ entry = &q->entry[index]; ++ buf = entry->buf; ++ if (len) ++ *len = FIELD_GET(MTK_WED_WO_CTL_SD_LEN0, ++ le32_to_cpu(READ_ONCE(desc->ctrl))); ++ if (buf) ++ dma_unmap_single(wo->hw->dev, entry->addr, buf_len, ++ DMA_FROM_DEVICE); ++ entry->buf = NULL; ++ ++ return buf; ++} ++ ++static int ++mtk_wed_wo_queue_refill(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q, ++ gfp_t gfp, bool rx) ++{ ++ enum dma_data_direction dir = rx ? DMA_FROM_DEVICE : DMA_TO_DEVICE; ++ int n_buf = 0; ++ ++ spin_lock_bh(&q->lock); ++ while (q->queued < q->n_desc) { ++ void *buf = page_frag_alloc(&q->cache, q->buf_size, gfp); ++ struct mtk_wed_wo_queue_entry *entry; ++ dma_addr_t addr; ++ ++ if (!buf) ++ break; ++ ++ addr = dma_map_single(wo->hw->dev, buf, q->buf_size, dir); ++ if (unlikely(dma_mapping_error(wo->hw->dev, addr))) { ++ skb_free_frag(buf); ++ break; ++ } ++ ++ q->head = (q->head + 1) % q->n_desc; ++ entry = &q->entry[q->head]; ++ entry->addr = addr; ++ entry->len = q->buf_size; ++ q->entry[q->head].buf = buf; ++ ++ if (rx) { ++ struct mtk_wed_wo_queue_desc *desc = &q->desc[q->head]; ++ u32 ctrl = MTK_WED_WO_CTL_LAST_SEC0 | ++ FIELD_PREP(MTK_WED_WO_CTL_SD_LEN0, ++ entry->len); ++ ++ WRITE_ONCE(desc->buf0, cpu_to_le32(addr)); ++ WRITE_ONCE(desc->ctrl, cpu_to_le32(ctrl)); ++ } ++ q->queued++; ++ n_buf++; ++ } ++ spin_unlock_bh(&q->lock); ++ ++ return n_buf; ++} ++ ++static void ++mtk_wed_wo_rx_complete(struct mtk_wed_wo *wo) ++{ ++ mtk_wed_wo_set_ack(wo, MTK_WED_WO_RXCH_INT_MASK); ++ mtk_wed_wo_irq_enable(wo, MTK_WED_WO_RXCH_INT_MASK); ++} ++ ++static void ++mtk_wed_wo_rx_run_queue(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q) ++{ ++ for (;;) { ++ struct mtk_wed_mcu_hdr *hdr; ++ struct sk_buff *skb; ++ void *data; ++ u32 len; ++ ++ data = mtk_wed_wo_dequeue(wo, q, &len, false); ++ if (!data) ++ break; ++ ++ skb = build_skb(data, q->buf_size); ++ if (!skb) { ++ skb_free_frag(data); ++ continue; ++ } ++ ++ __skb_put(skb, len); ++ if (mtk_wed_mcu_check_msg(wo, skb)) { ++ dev_kfree_skb(skb); ++ continue; ++ } ++ ++ hdr = (struct mtk_wed_mcu_hdr *)skb->data; ++ if (hdr->flag & cpu_to_le16(MTK_WED_WARP_CMD_FLAG_RSP)) ++ mtk_wed_mcu_rx_event(wo, skb); ++ else ++ mtk_wed_mcu_rx_unsolicited_event(wo, skb); ++ } ++ ++ if (mtk_wed_wo_queue_refill(wo, q, GFP_ATOMIC, true)) { ++ u32 index = (q->head - 1) % q->n_desc; ++ ++ mtk_wed_wo_queue_kick(wo, q, index); ++ } ++} ++ ++static irqreturn_t ++mtk_wed_wo_irq_handler(int irq, void *data) ++{ ++ struct mtk_wed_wo *wo = data; ++ ++ mtk_wed_wo_set_isr(wo, 0); ++ tasklet_schedule(&wo->mmio.irq_tasklet); ++ ++ return IRQ_HANDLED; ++} ++ ++static void mtk_wed_wo_irq_tasklet(struct tasklet_struct *t) ++{ ++ struct mtk_wed_wo *wo = from_tasklet(wo, t, mmio.irq_tasklet); ++ u32 intr, mask; ++ ++ /* disable interrupts */ ++ mtk_wed_wo_set_isr(wo, 0); ++ ++ intr = mtk_wed_wo_get_isr(wo); ++ intr &= wo->mmio.irq_mask; ++ mask = intr & (MTK_WED_WO_RXCH_INT_MASK | MTK_WED_WO_EXCEPTION_INT_MASK); ++ mtk_wed_wo_irq_disable(wo, mask); ++ ++ if (intr & MTK_WED_WO_RXCH_INT_MASK) { ++ mtk_wed_wo_rx_run_queue(wo, &wo->q_rx); ++ mtk_wed_wo_rx_complete(wo); ++ } ++} ++ ++/* mtk wed wo hw queues */ ++ ++static int ++mtk_wed_wo_queue_alloc(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q, ++ int n_desc, int buf_size, int index, ++ struct mtk_wed_wo_queue_regs *regs) ++{ ++ spin_lock_init(&q->lock); ++ q->regs = *regs; ++ q->n_desc = n_desc; ++ q->buf_size = buf_size; ++ ++ q->desc = dmam_alloc_coherent(wo->hw->dev, n_desc * sizeof(*q->desc), ++ &q->desc_dma, GFP_KERNEL); ++ if (!q->desc) ++ return -ENOMEM; ++ ++ q->entry = devm_kzalloc(wo->hw->dev, n_desc * sizeof(*q->entry), ++ GFP_KERNEL); ++ if (!q->entry) ++ return -ENOMEM; ++ ++ return 0; ++} ++ ++static void ++mtk_wed_wo_queue_free(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q) ++{ ++ mtk_wed_mmio_w32(wo, q->regs.cpu_idx, 0); ++ dma_free_coherent(wo->hw->dev, q->n_desc * sizeof(*q->desc), q->desc, ++ q->desc_dma); ++} ++ ++static void ++mtk_wed_wo_queue_tx_clean(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q) ++{ ++ struct page *page; ++ int i; ++ ++ spin_lock_bh(&q->lock); ++ for (i = 0; i < q->n_desc; i++) { ++ struct mtk_wed_wo_queue_entry *entry = &q->entry[i]; ++ ++ dma_unmap_single(wo->hw->dev, entry->addr, entry->len, ++ DMA_TO_DEVICE); ++ skb_free_frag(entry->buf); ++ entry->buf = NULL; ++ } ++ spin_unlock_bh(&q->lock); ++ ++ if (!q->cache.va) ++ return; ++ ++ page = virt_to_page(q->cache.va); ++ __page_frag_cache_drain(page, q->cache.pagecnt_bias); ++ memset(&q->cache, 0, sizeof(q->cache)); ++} ++ ++static void ++mtk_wed_wo_queue_rx_clean(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q) ++{ ++ struct page *page; ++ ++ spin_lock_bh(&q->lock); ++ for (;;) { ++ void *buf = mtk_wed_wo_dequeue(wo, q, NULL, true); ++ ++ if (!buf) ++ break; ++ ++ skb_free_frag(buf); ++ } ++ spin_unlock_bh(&q->lock); ++ ++ if (!q->cache.va) ++ return; ++ ++ page = virt_to_page(q->cache.va); ++ __page_frag_cache_drain(page, q->cache.pagecnt_bias); ++ memset(&q->cache, 0, sizeof(q->cache)); ++} ++ ++static void ++mtk_wed_wo_queue_reset(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q) ++{ ++ mtk_wed_mmio_w32(wo, q->regs.cpu_idx, 0); ++ mtk_wed_mmio_w32(wo, q->regs.desc_base, q->desc_dma); ++ mtk_wed_mmio_w32(wo, q->regs.ring_size, q->n_desc); ++} ++ ++int mtk_wed_wo_queue_tx_skb(struct mtk_wed_wo *wo, struct mtk_wed_wo_queue *q, ++ struct sk_buff *skb) ++{ ++ struct mtk_wed_wo_queue_entry *entry; ++ struct mtk_wed_wo_queue_desc *desc; ++ int ret = 0, index; ++ u32 ctrl; ++ ++ spin_lock_bh(&q->lock); ++ ++ q->tail = mtk_wed_mmio_r32(wo, q->regs.dma_idx); ++ index = (q->head + 1) % q->n_desc; ++ if (q->tail == index) { ++ ret = -ENOMEM; ++ goto out; ++ } ++ ++ entry = &q->entry[index]; ++ if (skb->len > entry->len) { ++ ret = -ENOMEM; ++ goto out; ++ } ++ ++ desc = &q->desc[index]; ++ q->head = index; ++ ++ dma_sync_single_for_cpu(wo->hw->dev, entry->addr, skb->len, ++ DMA_TO_DEVICE); ++ memcpy(entry->buf, skb->data, skb->len); ++ dma_sync_single_for_device(wo->hw->dev, entry->addr, skb->len, ++ DMA_TO_DEVICE); ++ ++ ctrl = FIELD_PREP(MTK_WED_WO_CTL_SD_LEN0, skb->len) | ++ MTK_WED_WO_CTL_LAST_SEC0 | MTK_WED_WO_CTL_DMA_DONE; ++ WRITE_ONCE(desc->buf0, cpu_to_le32(entry->addr)); ++ WRITE_ONCE(desc->ctrl, cpu_to_le32(ctrl)); ++ ++ mtk_wed_wo_queue_kick(wo, q, q->head); ++ mtk_wed_wo_kickout(wo); ++out: ++ spin_unlock_bh(&q->lock); ++ ++ dev_kfree_skb(skb); ++ ++ return ret; ++} ++ ++static int ++mtk_wed_wo_exception_init(struct mtk_wed_wo *wo) ++{ ++ return 0; ++} ++ ++static int ++mtk_wed_wo_hardware_init(struct mtk_wed_wo *wo) ++{ ++ struct mtk_wed_wo_queue_regs regs; ++ struct device_node *np; ++ int ret; ++ ++ np = of_parse_phandle(wo->hw->node, "mediatek,wo-ccif", 0); ++ if (!np) ++ return -ENODEV; ++ ++ wo->mmio.regs = syscon_regmap_lookup_by_phandle(np, NULL); ++ if (IS_ERR_OR_NULL(wo->mmio.regs)) ++ return PTR_ERR(wo->mmio.regs); ++ ++ wo->mmio.irq = irq_of_parse_and_map(np, 0); ++ wo->mmio.irq_mask = MTK_WED_WO_ALL_INT_MASK; ++ spin_lock_init(&wo->mmio.lock); ++ tasklet_setup(&wo->mmio.irq_tasklet, mtk_wed_wo_irq_tasklet); ++ ++ ret = devm_request_irq(wo->hw->dev, wo->mmio.irq, ++ mtk_wed_wo_irq_handler, IRQF_TRIGGER_HIGH, ++ KBUILD_MODNAME, wo); ++ if (ret) ++ goto error; ++ ++ regs.desc_base = MTK_WED_WO_CCIF_DUMMY1; ++ regs.ring_size = MTK_WED_WO_CCIF_DUMMY2; ++ regs.dma_idx = MTK_WED_WO_CCIF_SHADOW4; ++ regs.cpu_idx = MTK_WED_WO_CCIF_DUMMY3; ++ ++ ret = mtk_wed_wo_queue_alloc(wo, &wo->q_tx, MTK_WED_WO_RING_SIZE, ++ MTK_WED_WO_CMD_LEN, MTK_WED_WO_TXCH_NUM, ++ ®s); ++ if (ret) ++ goto error; ++ ++ mtk_wed_wo_queue_refill(wo, &wo->q_tx, GFP_KERNEL, false); ++ mtk_wed_wo_queue_reset(wo, &wo->q_tx); ++ ++ regs.desc_base = MTK_WED_WO_CCIF_DUMMY5; ++ regs.ring_size = MTK_WED_WO_CCIF_DUMMY6; ++ regs.dma_idx = MTK_WED_WO_CCIF_SHADOW8; ++ regs.cpu_idx = MTK_WED_WO_CCIF_DUMMY7; ++ ++ ret = mtk_wed_wo_queue_alloc(wo, &wo->q_rx, MTK_WED_WO_RING_SIZE, ++ MTK_WED_WO_CMD_LEN, MTK_WED_WO_RXCH_NUM, ++ ®s); ++ if (ret) ++ goto error; ++ ++ mtk_wed_wo_queue_refill(wo, &wo->q_rx, GFP_KERNEL, true); ++ mtk_wed_wo_queue_reset(wo, &wo->q_rx); ++ ++ /* rx queue irqmask */ ++ mtk_wed_wo_set_isr(wo, wo->mmio.irq_mask); ++ ++ return 0; ++ ++error: ++ devm_free_irq(wo->hw->dev, wo->mmio.irq, wo); ++ ++ return ret; ++} ++ ++static void ++mtk_wed_wo_hw_deinit(struct mtk_wed_wo *wo) ++{ ++ /* disable interrupts */ ++ mtk_wed_wo_set_isr(wo, 0); ++ ++ tasklet_disable(&wo->mmio.irq_tasklet); ++ ++ disable_irq(wo->mmio.irq); ++ devm_free_irq(wo->hw->dev, wo->mmio.irq, wo); ++ ++ mtk_wed_wo_queue_tx_clean(wo, &wo->q_tx); ++ mtk_wed_wo_queue_rx_clean(wo, &wo->q_rx); ++ mtk_wed_wo_queue_free(wo, &wo->q_tx); ++ mtk_wed_wo_queue_free(wo, &wo->q_rx); ++} ++ ++int mtk_wed_wo_init(struct mtk_wed_hw *hw) ++{ ++ struct mtk_wed_wo *wo; ++ int ret; ++ ++ wo = devm_kzalloc(hw->dev, sizeof(*wo), GFP_KERNEL); ++ if (!wo) ++ return -ENOMEM; ++ ++ hw->wed_wo = wo; ++ wo->hw = hw; ++ ++ ret = mtk_wed_wo_hardware_init(wo); ++ if (ret) ++ return ret; ++ ++ ret = mtk_wed_mcu_init(wo); ++ if (ret) ++ return ret; ++ ++ return mtk_wed_wo_exception_init(wo); ++} ++ ++void mtk_wed_wo_deinit(struct mtk_wed_hw *hw) ++{ ++ struct mtk_wed_wo *wo = hw->wed_wo; ++ ++ mtk_wed_wo_hw_deinit(wo); ++} +--- a/drivers/net/ethernet/mediatek/mtk_wed_wo.h ++++ b/drivers/net/ethernet/mediatek/mtk_wed_wo.h +@@ -80,6 +80,54 @@ enum mtk_wed_dummy_cr_idx { + #define MTK_WO_MCU_CFG_LS_WF_WM_WA_WM_CPU_RSTB_MASK BIT(5) + #define MTK_WO_MCU_CFG_LS_WF_WM_WA_WA_CPU_RSTB_MASK BIT(0) + ++#define MTK_WED_WO_RING_SIZE 256 ++#define MTK_WED_WO_CMD_LEN 1504 ++ ++#define MTK_WED_WO_TXCH_NUM 0 ++#define MTK_WED_WO_RXCH_NUM 1 ++#define MTK_WED_WO_RXCH_WO_EXCEPTION 7 ++ ++#define MTK_WED_WO_TXCH_INT_MASK BIT(0) ++#define MTK_WED_WO_RXCH_INT_MASK BIT(1) ++#define MTK_WED_WO_EXCEPTION_INT_MASK BIT(7) ++#define MTK_WED_WO_ALL_INT_MASK (MTK_WED_WO_RXCH_INT_MASK | \ ++ MTK_WED_WO_EXCEPTION_INT_MASK) ++ ++#define MTK_WED_WO_CCIF_BUSY 0x004 ++#define MTK_WED_WO_CCIF_START 0x008 ++#define MTK_WED_WO_CCIF_TCHNUM 0x00c ++#define MTK_WED_WO_CCIF_RCHNUM 0x010 ++#define MTK_WED_WO_CCIF_RCHNUM_MASK GENMASK(7, 0) ++ ++#define MTK_WED_WO_CCIF_ACK 0x014 ++#define MTK_WED_WO_CCIF_IRQ0_MASK 0x018 ++#define MTK_WED_WO_CCIF_IRQ1_MASK 0x01c ++#define MTK_WED_WO_CCIF_DUMMY1 0x020 ++#define MTK_WED_WO_CCIF_DUMMY2 0x024 ++#define MTK_WED_WO_CCIF_DUMMY3 0x028 ++#define MTK_WED_WO_CCIF_DUMMY4 0x02c ++#define MTK_WED_WO_CCIF_SHADOW1 0x030 ++#define MTK_WED_WO_CCIF_SHADOW2 0x034 ++#define MTK_WED_WO_CCIF_SHADOW3 0x038 ++#define MTK_WED_WO_CCIF_SHADOW4 0x03c ++#define MTK_WED_WO_CCIF_DUMMY5 0x050 ++#define MTK_WED_WO_CCIF_DUMMY6 0x054 ++#define MTK_WED_WO_CCIF_DUMMY7 0x058 ++#define MTK_WED_WO_CCIF_DUMMY8 0x05c ++#define MTK_WED_WO_CCIF_SHADOW5 0x060 ++#define MTK_WED_WO_CCIF_SHADOW6 0x064 ++#define MTK_WED_WO_CCIF_SHADOW7 0x068 ++#define MTK_WED_WO_CCIF_SHADOW8 0x06c ++ ++#define MTK_WED_WO_CTL_SD_LEN1 GENMASK(13, 0) ++#define MTK_WED_WO_CTL_LAST_SEC1 BIT(14) ++#define MTK_WED_WO_CTL_BURST BIT(15) ++#define MTK_WED_WO_CTL_SD_LEN0_SHIFT 16 ++#define MTK_WED_WO_CTL_SD_LEN0 GENMASK(29, 16) ++#define MTK_WED_WO_CTL_LAST_SEC0 BIT(30) ++#define MTK_WED_WO_CTL_DMA_DONE BIT(31) ++#define MTK_WED_WO_INFO_WINFO GENMASK(15, 0) ++ + struct mtk_wed_wo_memory_region { + const char *name; + void __iomem *addr; +@@ -112,10 +160,53 @@ struct mtk_wed_fw_trailer { + u32 crc; + }; + ++struct mtk_wed_wo_queue_regs { ++ u32 desc_base; ++ u32 ring_size; ++ u32 cpu_idx; ++ u32 dma_idx; ++}; ++ ++struct mtk_wed_wo_queue_desc { ++ __le32 buf0; ++ __le32 ctrl; ++ __le32 buf1; ++ __le32 info; ++ __le32 reserved[4]; ++} __packed __aligned(32); ++ ++struct mtk_wed_wo_queue_entry { ++ dma_addr_t addr; ++ void *buf; ++ u32 len; ++}; ++ ++struct mtk_wed_wo_queue { ++ struct mtk_wed_wo_queue_regs regs; ++ ++ struct page_frag_cache cache; ++ spinlock_t lock; ++ ++ struct mtk_wed_wo_queue_desc *desc; ++ dma_addr_t desc_dma; ++ ++ struct mtk_wed_wo_queue_entry *entry; ++ ++ u16 head; ++ u16 tail; ++ int n_desc; ++ int queued; ++ int buf_size; ++ ++}; ++ + struct mtk_wed_wo { + struct mtk_wed_hw *hw; + struct mtk_wed_wo_memory_region boot; + ++ struct mtk_wed_wo_queue q_tx; ++ struct mtk_wed_wo_queue q_rx; ++ + struct { + struct mutex mutex; + int timeout; +@@ -124,6 +215,15 @@ struct mtk_wed_wo { + struct sk_buff_head res_q; + wait_queue_head_t wait; + } mcu; ++ ++ struct { ++ struct regmap *regs; ++ ++ spinlock_t lock; ++ struct tasklet_struct irq_tasklet; ++ int irq; ++ u32 irq_mask; ++ } mmio; + }; + + static inline int +@@ -146,5 +246,9 @@ void mtk_wed_mcu_rx_unsolicited_event(st + int mtk_wed_mcu_send_msg(struct mtk_wed_wo *wo, int id, int cmd, + const void *data, int len, bool wait_resp); + int mtk_wed_mcu_init(struct mtk_wed_wo *wo); ++int mtk_wed_wo_init(struct mtk_wed_hw *hw); ++void mtk_wed_wo_deinit(struct mtk_wed_hw *hw); ++int mtk_wed_wo_queue_tx_skb(struct mtk_wed_wo *dev, struct mtk_wed_wo_queue *q, ++ struct sk_buff *skb); + + #endif /* __MTK_WED_WO_H */ diff --git a/target/linux/generic/pending-5.15/733-03-net-ethernet-mtk_wed-rename-tx_wdma-array-in-rx_wdma.patch b/target/linux/generic/pending-5.15/733-03-net-ethernet-mtk_wed-rename-tx_wdma-array-in-rx_wdma.patch new file mode 100644 index 000000000..ffd6bc358 --- /dev/null +++ b/target/linux/generic/pending-5.15/733-03-net-ethernet-mtk_wed-rename-tx_wdma-array-in-rx_wdma.patch @@ -0,0 +1,79 @@ +From: Lorenzo Bianconi +Date: Sat, 5 Nov 2022 23:36:20 +0100 +Subject: [PATCH] net: ethernet: mtk_wed: rename tx_wdma array in rx_wdma + +Rename tx_wdma queue array in rx_wdma since this is rx side of wdma soc. +Moreover rename mtk_wed_wdma_ring_setup routine in +mtk_wed_wdma_rx_ring_setup() + +Signed-off-by: Lorenzo Bianconi +Signed-off-by: David S. Miller +--- + +--- a/drivers/net/ethernet/mediatek/mtk_wed.c ++++ b/drivers/net/ethernet/mediatek/mtk_wed.c +@@ -253,8 +253,8 @@ mtk_wed_free_tx_rings(struct mtk_wed_dev + + for (i = 0; i < ARRAY_SIZE(dev->tx_ring); i++) + mtk_wed_free_ring(dev, &dev->tx_ring[i]); +- for (i = 0; i < ARRAY_SIZE(dev->tx_wdma); i++) +- mtk_wed_free_ring(dev, &dev->tx_wdma[i]); ++ for (i = 0; i < ARRAY_SIZE(dev->rx_wdma); i++) ++ mtk_wed_free_ring(dev, &dev->rx_wdma[i]); + } + + static void +@@ -695,10 +695,10 @@ mtk_wed_ring_alloc(struct mtk_wed_device + } + + static int +-mtk_wed_wdma_ring_setup(struct mtk_wed_device *dev, int idx, int size) ++mtk_wed_wdma_rx_ring_setup(struct mtk_wed_device *dev, int idx, int size) + { + u32 desc_size = sizeof(struct mtk_wdma_desc) * dev->hw->version; +- struct mtk_wed_ring *wdma = &dev->tx_wdma[idx]; ++ struct mtk_wed_ring *wdma = &dev->rx_wdma[idx]; + + if (mtk_wed_ring_alloc(dev, wdma, MTK_WED_WDMA_RING_SIZE, desc_size)) + return -ENOMEM; +@@ -812,9 +812,9 @@ mtk_wed_start(struct mtk_wed_device *dev + { + int i; + +- for (i = 0; i < ARRAY_SIZE(dev->tx_wdma); i++) +- if (!dev->tx_wdma[i].desc) +- mtk_wed_wdma_ring_setup(dev, i, 16); ++ for (i = 0; i < ARRAY_SIZE(dev->rx_wdma); i++) ++ if (!dev->rx_wdma[i].desc) ++ mtk_wed_wdma_rx_ring_setup(dev, i, 16); + + mtk_wed_hw_init(dev); + mtk_wed_configure_irq(dev, irq_mask); +@@ -923,7 +923,7 @@ mtk_wed_tx_ring_setup(struct mtk_wed_dev + sizeof(*ring->desc))) + return -ENOMEM; + +- if (mtk_wed_wdma_ring_setup(dev, idx, MTK_WED_WDMA_RING_SIZE)) ++ if (mtk_wed_wdma_rx_ring_setup(dev, idx, MTK_WED_WDMA_RING_SIZE)) + return -ENOMEM; + + ring->reg_base = MTK_WED_RING_TX(idx); +--- a/include/linux/soc/mediatek/mtk_wed.h ++++ b/include/linux/soc/mediatek/mtk_wed.h +@@ -7,6 +7,7 @@ + #include + + #define MTK_WED_TX_QUEUES 2 ++#define MTK_WED_RX_QUEUES 2 + + struct mtk_wed_hw; + struct mtk_wdma_desc; +@@ -66,7 +67,7 @@ struct mtk_wed_device { + + struct mtk_wed_ring tx_ring[MTK_WED_TX_QUEUES]; + struct mtk_wed_ring txfree_ring; +- struct mtk_wed_ring tx_wdma[MTK_WED_TX_QUEUES]; ++ struct mtk_wed_ring rx_wdma[MTK_WED_RX_QUEUES]; + + struct { + int size; diff --git a/target/linux/generic/pending-5.15/733-04-net-ethernet-mtk_wed-add-configure-wed-wo-support.patch b/target/linux/generic/pending-5.15/733-04-net-ethernet-mtk_wed-add-configure-wed-wo-support.patch new file mode 100644 index 000000000..4c34d0cb3 --- /dev/null +++ b/target/linux/generic/pending-5.15/733-04-net-ethernet-mtk_wed-add-configure-wed-wo-support.patch @@ -0,0 +1,1521 @@ +From: Lorenzo Bianconi +Date: Sat, 5 Nov 2022 23:36:21 +0100 +Subject: [PATCH] net: ethernet: mtk_wed: add configure wed wo support + +Enable RX Wireless Ethernet Dispatch available on MT7986 Soc. + +Tested-by: Daniel Golle +Co-developed-by: Sujuan Chen +Signed-off-by: Sujuan Chen +Signed-off-by: Lorenzo Bianconi +Signed-off-by: David S. Miller +--- + +--- a/drivers/net/ethernet/mediatek/mtk_wed.c ++++ b/drivers/net/ethernet/mediatek/mtk_wed.c +@@ -9,6 +9,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -23,6 +24,7 @@ + #define MTK_WED_PKT_SIZE 1900 + #define MTK_WED_BUF_SIZE 2048 + #define MTK_WED_BUF_PER_PAGE (PAGE_SIZE / 2048) ++#define MTK_WED_RX_RING_SIZE 1536 + + #define MTK_WED_TX_RING_SIZE 2048 + #define MTK_WED_WDMA_RING_SIZE 1024 +@@ -31,6 +33,10 @@ + #define MTK_WED_PER_GROUP_PKT 128 + + #define MTK_WED_FBUF_SIZE 128 ++#define MTK_WED_MIOD_CNT 16 ++#define MTK_WED_FB_CMD_CNT 1024 ++#define MTK_WED_RRO_QUE_CNT 8192 ++#define MTK_WED_MIOD_ENTRY_CNT 128 + + static struct mtk_wed_hw *hw_list[2]; + static DEFINE_MUTEX(hw_lock); +@@ -65,12 +71,76 @@ wdma_set(struct mtk_wed_device *dev, u32 + wdma_m32(dev, reg, 0, mask); + } + ++static void ++wdma_clr(struct mtk_wed_device *dev, u32 reg, u32 mask) ++{ ++ wdma_m32(dev, reg, mask, 0); ++} ++ ++static u32 ++wifi_r32(struct mtk_wed_device *dev, u32 reg) ++{ ++ return readl(dev->wlan.base + reg); ++} ++ ++static void ++wifi_w32(struct mtk_wed_device *dev, u32 reg, u32 val) ++{ ++ writel(val, dev->wlan.base + reg); ++} ++ + static u32 + mtk_wed_read_reset(struct mtk_wed_device *dev) + { + return wed_r32(dev, MTK_WED_RESET); + } + ++static u32 ++mtk_wdma_read_reset(struct mtk_wed_device *dev) ++{ ++ return wdma_r32(dev, MTK_WDMA_GLO_CFG); ++} ++ ++static void ++mtk_wdma_rx_reset(struct mtk_wed_device *dev) ++{ ++ u32 status, mask = MTK_WDMA_GLO_CFG_RX_DMA_BUSY; ++ int i; ++ ++ wdma_clr(dev, MTK_WDMA_GLO_CFG, MTK_WDMA_GLO_CFG_RX_DMA_EN); ++ if (readx_poll_timeout(mtk_wdma_read_reset, dev, status, ++ !(status & mask), 0, 1000)) ++ dev_err(dev->hw->dev, "rx reset failed\n"); ++ ++ for (i = 0; i < ARRAY_SIZE(dev->rx_wdma); i++) { ++ if (dev->rx_wdma[i].desc) ++ continue; ++ ++ wdma_w32(dev, ++ MTK_WDMA_RING_RX(i) + MTK_WED_RING_OFS_CPU_IDX, 0); ++ } ++} ++ ++static void ++mtk_wdma_tx_reset(struct mtk_wed_device *dev) ++{ ++ u32 status, mask = MTK_WDMA_GLO_CFG_TX_DMA_BUSY; ++ int i; ++ ++ wdma_clr(dev, MTK_WDMA_GLO_CFG, MTK_WDMA_GLO_CFG_TX_DMA_EN); ++ if (readx_poll_timeout(mtk_wdma_read_reset, dev, status, ++ !(status & mask), 0, 1000)) ++ dev_err(dev->hw->dev, "tx reset failed\n"); ++ ++ for (i = 0; i < ARRAY_SIZE(dev->tx_wdma); i++) { ++ if (dev->tx_wdma[i].desc) ++ continue; ++ ++ wdma_w32(dev, ++ MTK_WDMA_RING_TX(i) + MTK_WED_RING_OFS_CPU_IDX, 0); ++ } ++} ++ + static void + mtk_wed_reset(struct mtk_wed_device *dev, u32 mask) + { +@@ -82,6 +152,54 @@ mtk_wed_reset(struct mtk_wed_device *dev + WARN_ON_ONCE(1); + } + ++static u32 ++mtk_wed_wo_read_status(struct mtk_wed_device *dev) ++{ ++ return wed_r32(dev, MTK_WED_SCR0 + 4 * MTK_WED_DUMMY_CR_WO_STATUS); ++} ++ ++static void ++mtk_wed_wo_reset(struct mtk_wed_device *dev) ++{ ++ struct mtk_wed_wo *wo = dev->hw->wed_wo; ++ u8 state = MTK_WED_WO_STATE_DISABLE; ++ void __iomem *reg; ++ u32 val; ++ ++ mtk_wdma_tx_reset(dev); ++ mtk_wed_reset(dev, MTK_WED_RESET_WED); ++ ++ mtk_wed_mcu_send_msg(wo, MTK_WED_MODULE_ID_WO, ++ MTK_WED_WO_CMD_CHANGE_STATE, &state, ++ sizeof(state), false); ++ ++ if (readx_poll_timeout(mtk_wed_wo_read_status, dev, val, ++ val == MTK_WED_WOIF_DISABLE_DONE, ++ 100, MTK_WOCPU_TIMEOUT)) ++ dev_err(dev->hw->dev, "failed to disable wed-wo\n"); ++ ++ reg = ioremap(MTK_WED_WO_CPU_MCUSYS_RESET_ADDR, 4); ++ ++ val = readl(reg); ++ switch (dev->hw->index) { ++ case 0: ++ val |= MTK_WED_WO_CPU_WO0_MCUSYS_RESET_MASK; ++ writel(val, reg); ++ val &= ~MTK_WED_WO_CPU_WO0_MCUSYS_RESET_MASK; ++ writel(val, reg); ++ break; ++ case 1: ++ val |= MTK_WED_WO_CPU_WO1_MCUSYS_RESET_MASK; ++ writel(val, reg); ++ val &= ~MTK_WED_WO_CPU_WO1_MCUSYS_RESET_MASK; ++ writel(val, reg); ++ break; ++ default: ++ break; ++ } ++ iounmap(reg); ++} ++ + static struct mtk_wed_hw * + mtk_wed_assign(struct mtk_wed_device *dev) + { +@@ -116,7 +234,7 @@ out: + } + + static int +-mtk_wed_buffer_alloc(struct mtk_wed_device *dev) ++mtk_wed_tx_buffer_alloc(struct mtk_wed_device *dev) + { + struct mtk_wdma_desc *desc; + dma_addr_t desc_phys; +@@ -133,16 +251,16 @@ mtk_wed_buffer_alloc(struct mtk_wed_devi + if (!page_list) + return -ENOMEM; + +- dev->buf_ring.size = ring_size; +- dev->buf_ring.pages = page_list; ++ dev->tx_buf_ring.size = ring_size; ++ dev->tx_buf_ring.pages = page_list; + + desc = dma_alloc_coherent(dev->hw->dev, ring_size * sizeof(*desc), + &desc_phys, GFP_KERNEL); + if (!desc) + return -ENOMEM; + +- dev->buf_ring.desc = desc; +- dev->buf_ring.desc_phys = desc_phys; ++ dev->tx_buf_ring.desc = desc; ++ dev->tx_buf_ring.desc_phys = desc_phys; + + for (i = 0, page_idx = 0; i < ring_size; i += MTK_WED_BUF_PER_PAGE) { + dma_addr_t page_phys, buf_phys; +@@ -203,10 +321,10 @@ mtk_wed_buffer_alloc(struct mtk_wed_devi + } + + static void +-mtk_wed_free_buffer(struct mtk_wed_device *dev) ++mtk_wed_free_tx_buffer(struct mtk_wed_device *dev) + { +- struct mtk_wdma_desc *desc = dev->buf_ring.desc; +- void **page_list = dev->buf_ring.pages; ++ struct mtk_wdma_desc *desc = dev->tx_buf_ring.desc; ++ void **page_list = dev->tx_buf_ring.pages; + int page_idx; + int i; + +@@ -216,7 +334,8 @@ mtk_wed_free_buffer(struct mtk_wed_devic + if (!desc) + goto free_pagelist; + +- for (i = 0, page_idx = 0; i < dev->buf_ring.size; i += MTK_WED_BUF_PER_PAGE) { ++ for (i = 0, page_idx = 0; i < dev->tx_buf_ring.size; ++ i += MTK_WED_BUF_PER_PAGE) { + void *page = page_list[page_idx++]; + dma_addr_t buf_addr; + +@@ -229,13 +348,59 @@ mtk_wed_free_buffer(struct mtk_wed_devic + __free_page(page); + } + +- dma_free_coherent(dev->hw->dev, dev->buf_ring.size * sizeof(*desc), +- desc, dev->buf_ring.desc_phys); ++ dma_free_coherent(dev->hw->dev, dev->tx_buf_ring.size * sizeof(*desc), ++ desc, dev->tx_buf_ring.desc_phys); + + free_pagelist: + kfree(page_list); + } + ++static int ++mtk_wed_rx_buffer_alloc(struct mtk_wed_device *dev) ++{ ++ struct mtk_rxbm_desc *desc; ++ dma_addr_t desc_phys; ++ ++ dev->rx_buf_ring.size = dev->wlan.rx_nbuf; ++ desc = dma_alloc_coherent(dev->hw->dev, ++ dev->wlan.rx_nbuf * sizeof(*desc), ++ &desc_phys, GFP_KERNEL); ++ if (!desc) ++ return -ENOMEM; ++ ++ dev->rx_buf_ring.desc = desc; ++ dev->rx_buf_ring.desc_phys = desc_phys; ++ dev->wlan.init_rx_buf(dev, dev->wlan.rx_npkt); ++ ++ return 0; ++} ++ ++static void ++mtk_wed_free_rx_buffer(struct mtk_wed_device *dev) ++{ ++ struct mtk_rxbm_desc *desc = dev->rx_buf_ring.desc; ++ ++ if (!desc) ++ return; ++ ++ dev->wlan.release_rx_buf(dev); ++ dma_free_coherent(dev->hw->dev, dev->rx_buf_ring.size * sizeof(*desc), ++ desc, dev->rx_buf_ring.desc_phys); ++} ++ ++static void ++mtk_wed_rx_buffer_hw_init(struct mtk_wed_device *dev) ++{ ++ wed_w32(dev, MTK_WED_RX_BM_RX_DMAD, ++ FIELD_PREP(MTK_WED_RX_BM_RX_DMAD_SDL0, dev->wlan.rx_size)); ++ wed_w32(dev, MTK_WED_RX_BM_BASE, dev->rx_buf_ring.desc_phys); ++ wed_w32(dev, MTK_WED_RX_BM_INIT_PTR, MTK_WED_RX_BM_INIT_SW_TAIL | ++ FIELD_PREP(MTK_WED_RX_BM_SW_TAIL, dev->wlan.rx_npkt)); ++ wed_w32(dev, MTK_WED_RX_BM_DYN_ALLOC_TH, ++ FIELD_PREP(MTK_WED_RX_BM_DYN_ALLOC_TH_H, 0xffff)); ++ wed_set(dev, MTK_WED_CTRL, MTK_WED_CTRL_WED_RX_BM_EN); ++} ++ + static void + mtk_wed_free_ring(struct mtk_wed_device *dev, struct mtk_wed_ring *ring) + { +@@ -247,6 +412,13 @@ mtk_wed_free_ring(struct mtk_wed_device + } + + static void ++mtk_wed_free_rx_rings(struct mtk_wed_device *dev) ++{ ++ mtk_wed_free_rx_buffer(dev); ++ mtk_wed_free_ring(dev, &dev->rro.ring); ++} ++ ++static void + mtk_wed_free_tx_rings(struct mtk_wed_device *dev) + { + int i; +@@ -291,6 +463,38 @@ mtk_wed_set_512_support(struct mtk_wed_d + } + } + ++#define MTK_WFMDA_RX_DMA_EN BIT(2) ++static void ++mtk_wed_check_wfdma_rx_fill(struct mtk_wed_device *dev, int idx) ++{ ++ u32 val; ++ int i; ++ ++ if (!(dev->rx_ring[idx].flags & MTK_WED_RING_CONFIGURED)) ++ return; /* queue is not configured by mt76 */ ++ ++ for (i = 0; i < 3; i++) { ++ u32 cur_idx; ++ ++ cur_idx = wed_r32(dev, ++ MTK_WED_WPDMA_RING_RX_DATA(idx) + ++ MTK_WED_RING_OFS_CPU_IDX); ++ if (cur_idx == MTK_WED_RX_RING_SIZE - 1) ++ break; ++ ++ usleep_range(100000, 200000); ++ } ++ ++ if (i == 3) { ++ dev_err(dev->hw->dev, "rx dma enable failed\n"); ++ return; ++ } ++ ++ val = wifi_r32(dev, dev->wlan.wpdma_rx_glo - dev->wlan.phy_base) | ++ MTK_WFMDA_RX_DMA_EN; ++ wifi_w32(dev, dev->wlan.wpdma_rx_glo - dev->wlan.phy_base, val); ++} ++ + static void + mtk_wed_dma_disable(struct mtk_wed_device *dev) + { +@@ -304,20 +508,25 @@ mtk_wed_dma_disable(struct mtk_wed_devic + MTK_WED_GLO_CFG_TX_DMA_EN | + MTK_WED_GLO_CFG_RX_DMA_EN); + +- wdma_m32(dev, MTK_WDMA_GLO_CFG, ++ wdma_clr(dev, MTK_WDMA_GLO_CFG, + MTK_WDMA_GLO_CFG_TX_DMA_EN | + MTK_WDMA_GLO_CFG_RX_INFO1_PRERES | +- MTK_WDMA_GLO_CFG_RX_INFO2_PRERES, 0); ++ MTK_WDMA_GLO_CFG_RX_INFO2_PRERES); + + if (dev->hw->version == 1) { + regmap_write(dev->hw->mirror, dev->hw->index * 4, 0); +- wdma_m32(dev, MTK_WDMA_GLO_CFG, +- MTK_WDMA_GLO_CFG_RX_INFO3_PRERES, 0); ++ wdma_clr(dev, MTK_WDMA_GLO_CFG, ++ MTK_WDMA_GLO_CFG_RX_INFO3_PRERES); + } else { + wed_clr(dev, MTK_WED_WPDMA_GLO_CFG, + MTK_WED_WPDMA_GLO_CFG_RX_DRV_R0_PKT_PROC | + MTK_WED_WPDMA_GLO_CFG_RX_DRV_R0_CRX_SYNC); + ++ wed_clr(dev, MTK_WED_WPDMA_RX_D_GLO_CFG, ++ MTK_WED_WPDMA_RX_D_RX_DRV_EN); ++ wed_clr(dev, MTK_WED_WDMA_GLO_CFG, ++ MTK_WED_WDMA_GLO_CFG_TX_DDONE_CHK); ++ + mtk_wed_set_512_support(dev, false); + } + } +@@ -338,6 +547,13 @@ mtk_wed_stop(struct mtk_wed_device *dev) + wdma_w32(dev, MTK_WDMA_INT_MASK, 0); + wdma_w32(dev, MTK_WDMA_INT_GRP2, 0); + wed_w32(dev, MTK_WED_WPDMA_INT_MASK, 0); ++ ++ if (dev->hw->version == 1) ++ return; ++ ++ wed_w32(dev, MTK_WED_EXT_INT_MASK1, 0); ++ wed_w32(dev, MTK_WED_EXT_INT_MASK2, 0); ++ wed_clr(dev, MTK_WED_CTRL, MTK_WED_CTRL_WED_RX_BM_EN); + } + + static void +@@ -353,11 +569,21 @@ mtk_wed_detach(struct mtk_wed_device *de + wdma_w32(dev, MTK_WDMA_RESET_IDX, 0); + + mtk_wed_reset(dev, MTK_WED_RESET_WED); ++ if (mtk_wed_get_rx_capa(dev)) { ++ wdma_clr(dev, MTK_WDMA_GLO_CFG, MTK_WDMA_GLO_CFG_TX_DMA_EN); ++ wdma_w32(dev, MTK_WDMA_RESET_IDX, MTK_WDMA_RESET_IDX_TX); ++ wdma_w32(dev, MTK_WDMA_RESET_IDX, 0); ++ } + +- mtk_wed_free_buffer(dev); ++ mtk_wed_free_tx_buffer(dev); + mtk_wed_free_tx_rings(dev); +- if (hw->version != 1) ++ ++ if (mtk_wed_get_rx_capa(dev)) { ++ mtk_wed_wo_reset(dev); ++ mtk_wed_free_rx_rings(dev); + mtk_wed_wo_deinit(hw); ++ mtk_wdma_rx_reset(dev); ++ } + + if (dev->wlan.bus_type == MTK_WED_BUS_PCIE) { + struct device_node *wlan_node; +@@ -441,10 +667,12 @@ mtk_wed_set_wpdma(struct mtk_wed_device + } else { + mtk_wed_bus_init(dev); + +- wed_w32(dev, MTK_WED_WPDMA_CFG_BASE, dev->wlan.wpdma_int); +- wed_w32(dev, MTK_WED_WPDMA_CFG_INT_MASK, dev->wlan.wpdma_mask); +- wed_w32(dev, MTK_WED_WPDMA_CFG_TX, dev->wlan.wpdma_tx); +- wed_w32(dev, MTK_WED_WPDMA_CFG_TX_FREE, dev->wlan.wpdma_txfree); ++ wed_w32(dev, MTK_WED_WPDMA_CFG_BASE, dev->wlan.wpdma_int); ++ wed_w32(dev, MTK_WED_WPDMA_CFG_INT_MASK, dev->wlan.wpdma_mask); ++ wed_w32(dev, MTK_WED_WPDMA_CFG_TX, dev->wlan.wpdma_tx); ++ wed_w32(dev, MTK_WED_WPDMA_CFG_TX_FREE, dev->wlan.wpdma_txfree); ++ wed_w32(dev, MTK_WED_WPDMA_RX_GLO_CFG, dev->wlan.wpdma_rx_glo); ++ wed_w32(dev, MTK_WED_WPDMA_RX_RING, dev->wlan.wpdma_rx); + } + } + +@@ -494,6 +722,132 @@ mtk_wed_hw_init_early(struct mtk_wed_dev + } + } + ++static int ++mtk_wed_rro_ring_alloc(struct mtk_wed_device *dev, struct mtk_wed_ring *ring, ++ int size) ++{ ++ ring->desc = dma_alloc_coherent(dev->hw->dev, ++ size * sizeof(*ring->desc), ++ &ring->desc_phys, GFP_KERNEL); ++ if (!ring->desc) ++ return -ENOMEM; ++ ++ ring->desc_size = sizeof(*ring->desc); ++ ring->size = size; ++ memset(ring->desc, 0, size); ++ ++ return 0; ++} ++ ++#define MTK_WED_MIOD_COUNT (MTK_WED_MIOD_ENTRY_CNT * MTK_WED_MIOD_CNT) ++static int ++mtk_wed_rro_alloc(struct mtk_wed_device *dev) ++{ ++ struct reserved_mem *rmem; ++ struct device_node *np; ++ int index; ++ ++ index = of_property_match_string(dev->hw->node, "memory-region-names", ++ "wo-dlm"); ++ if (index < 0) ++ return index; ++ ++ np = of_parse_phandle(dev->hw->node, "memory-region", index); ++ if (!np) ++ return -ENODEV; ++ ++ rmem = of_reserved_mem_lookup(np); ++ of_node_put(np); ++ ++ if (!rmem) ++ return -ENODEV; ++ ++ dev->rro.miod_phys = rmem->base; ++ dev->rro.fdbk_phys = MTK_WED_MIOD_COUNT + dev->rro.miod_phys; ++ ++ return mtk_wed_rro_ring_alloc(dev, &dev->rro.ring, ++ MTK_WED_RRO_QUE_CNT); ++} ++ ++static int ++mtk_wed_rro_cfg(struct mtk_wed_device *dev) ++{ ++ struct mtk_wed_wo *wo = dev->hw->wed_wo; ++ struct { ++ struct { ++ __le32 base; ++ __le32 cnt; ++ __le32 unit; ++ } ring[2]; ++ __le32 wed; ++ u8 version; ++ } req = { ++ .ring[0] = { ++ .base = cpu_to_le32(MTK_WED_WOCPU_VIEW_MIOD_BASE), ++ .cnt = cpu_to_le32(MTK_WED_MIOD_CNT), ++ .unit = cpu_to_le32(MTK_WED_MIOD_ENTRY_CNT), ++ }, ++ .ring[1] = { ++ .base = cpu_to_le32(MTK_WED_WOCPU_VIEW_MIOD_BASE + ++ MTK_WED_MIOD_COUNT), ++ .cnt = cpu_to_le32(MTK_WED_FB_CMD_CNT), ++ .unit = cpu_to_le32(4), ++ }, ++ }; ++ ++ return mtk_wed_mcu_send_msg(wo, MTK_WED_MODULE_ID_WO, ++ MTK_WED_WO_CMD_WED_CFG, ++ &req, sizeof(req), true); ++} ++ ++static void ++mtk_wed_rro_hw_init(struct mtk_wed_device *dev) ++{ ++ wed_w32(dev, MTK_WED_RROQM_MIOD_CFG, ++ FIELD_PREP(MTK_WED_RROQM_MIOD_MID_DW, 0x70 >> 2) | ++ FIELD_PREP(MTK_WED_RROQM_MIOD_MOD_DW, 0x10 >> 2) | ++ FIELD_PREP(MTK_WED_RROQM_MIOD_ENTRY_DW, ++ MTK_WED_MIOD_ENTRY_CNT >> 2)); ++ ++ wed_w32(dev, MTK_WED_RROQM_MIOD_CTRL0, dev->rro.miod_phys); ++ wed_w32(dev, MTK_WED_RROQM_MIOD_CTRL1, ++ FIELD_PREP(MTK_WED_RROQM_MIOD_CNT, MTK_WED_MIOD_CNT)); ++ wed_w32(dev, MTK_WED_RROQM_FDBK_CTRL0, dev->rro.fdbk_phys); ++ wed_w32(dev, MTK_WED_RROQM_FDBK_CTRL1, ++ FIELD_PREP(MTK_WED_RROQM_FDBK_CNT, MTK_WED_FB_CMD_CNT)); ++ wed_w32(dev, MTK_WED_RROQM_FDBK_CTRL2, 0); ++ wed_w32(dev, MTK_WED_RROQ_BASE_L, dev->rro.ring.desc_phys); ++ ++ wed_set(dev, MTK_WED_RROQM_RST_IDX, ++ MTK_WED_RROQM_RST_IDX_MIOD | ++ MTK_WED_RROQM_RST_IDX_FDBK); ++ ++ wed_w32(dev, MTK_WED_RROQM_RST_IDX, 0); ++ wed_w32(dev, MTK_WED_RROQM_MIOD_CTRL2, MTK_WED_MIOD_CNT - 1); ++ wed_set(dev, MTK_WED_CTRL, MTK_WED_CTRL_RX_RRO_QM_EN); ++} ++ ++static void ++mtk_wed_route_qm_hw_init(struct mtk_wed_device *dev) ++{ ++ wed_w32(dev, MTK_WED_RESET, MTK_WED_RESET_RX_ROUTE_QM); ++ ++ for (;;) { ++ usleep_range(100, 200); ++ if (!(wed_r32(dev, MTK_WED_RESET) & MTK_WED_RESET_RX_ROUTE_QM)) ++ break; ++ } ++ ++ /* configure RX_ROUTE_QM */ ++ wed_clr(dev, MTK_WED_RTQM_GLO_CFG, MTK_WED_RTQM_Q_RST); ++ wed_clr(dev, MTK_WED_RTQM_GLO_CFG, MTK_WED_RTQM_TXDMAD_FPORT); ++ wed_set(dev, MTK_WED_RTQM_GLO_CFG, ++ FIELD_PREP(MTK_WED_RTQM_TXDMAD_FPORT, 0x3 + dev->hw->index)); ++ wed_clr(dev, MTK_WED_RTQM_GLO_CFG, MTK_WED_RTQM_Q_RST); ++ /* enable RX_ROUTE_QM */ ++ wed_set(dev, MTK_WED_CTRL, MTK_WED_CTRL_RX_ROUTE_QM_EN); ++} ++ + static void + mtk_wed_hw_init(struct mtk_wed_device *dev) + { +@@ -505,11 +859,11 @@ mtk_wed_hw_init(struct mtk_wed_device *d + wed_w32(dev, MTK_WED_TX_BM_CTRL, + MTK_WED_TX_BM_CTRL_PAUSE | + FIELD_PREP(MTK_WED_TX_BM_CTRL_VLD_GRP_NUM, +- dev->buf_ring.size / 128) | ++ dev->tx_buf_ring.size / 128) | + FIELD_PREP(MTK_WED_TX_BM_CTRL_RSV_GRP_NUM, + MTK_WED_TX_RING_SIZE / 256)); + +- wed_w32(dev, MTK_WED_TX_BM_BASE, dev->buf_ring.desc_phys); ++ wed_w32(dev, MTK_WED_TX_BM_BASE, dev->tx_buf_ring.desc_phys); + + wed_w32(dev, MTK_WED_TX_BM_BUF_LEN, MTK_WED_PKT_SIZE); + +@@ -536,9 +890,9 @@ mtk_wed_hw_init(struct mtk_wed_device *d + wed_w32(dev, MTK_WED_TX_TKID_CTRL, + MTK_WED_TX_TKID_CTRL_PAUSE | + FIELD_PREP(MTK_WED_TX_TKID_CTRL_VLD_GRP_NUM, +- dev->buf_ring.size / 128) | ++ dev->tx_buf_ring.size / 128) | + FIELD_PREP(MTK_WED_TX_TKID_CTRL_RSV_GRP_NUM, +- dev->buf_ring.size / 128)); ++ dev->tx_buf_ring.size / 128)); + wed_w32(dev, MTK_WED_TX_TKID_DYN_THR, + FIELD_PREP(MTK_WED_TX_TKID_DYN_THR_LO, 0) | + MTK_WED_TX_TKID_DYN_THR_HI); +@@ -546,18 +900,28 @@ mtk_wed_hw_init(struct mtk_wed_device *d + + mtk_wed_reset(dev, MTK_WED_RESET_TX_BM); + +- if (dev->hw->version == 1) ++ if (dev->hw->version == 1) { + wed_set(dev, MTK_WED_CTRL, + MTK_WED_CTRL_WED_TX_BM_EN | + MTK_WED_CTRL_WED_TX_FREE_AGENT_EN); +- else ++ } else { + wed_clr(dev, MTK_WED_TX_TKID_CTRL, MTK_WED_TX_TKID_CTRL_PAUSE); ++ /* rx hw init */ ++ wed_w32(dev, MTK_WED_WPDMA_RX_D_RST_IDX, ++ MTK_WED_WPDMA_RX_D_RST_CRX_IDX | ++ MTK_WED_WPDMA_RX_D_RST_DRV_IDX); ++ wed_w32(dev, MTK_WED_WPDMA_RX_D_RST_IDX, 0); ++ ++ mtk_wed_rx_buffer_hw_init(dev); ++ mtk_wed_rro_hw_init(dev); ++ mtk_wed_route_qm_hw_init(dev); ++ } + + wed_clr(dev, MTK_WED_TX_BM_CTRL, MTK_WED_TX_BM_CTRL_PAUSE); + } + + static void +-mtk_wed_ring_reset(struct mtk_wed_ring *ring, int size) ++mtk_wed_ring_reset(struct mtk_wed_ring *ring, int size, bool tx) + { + void *head = (void *)ring->desc; + int i; +@@ -567,7 +931,10 @@ mtk_wed_ring_reset(struct mtk_wed_ring * + + desc = (struct mtk_wdma_desc *)(head + i * ring->desc_size); + desc->buf0 = 0; +- desc->ctrl = cpu_to_le32(MTK_WDMA_DESC_CTRL_DMA_DONE); ++ if (tx) ++ desc->ctrl = cpu_to_le32(MTK_WDMA_DESC_CTRL_DMA_DONE); ++ else ++ desc->ctrl = cpu_to_le32(MTK_WFDMA_DESC_CTRL_TO_HOST); + desc->buf1 = 0; + desc->info = 0; + } +@@ -623,7 +990,8 @@ mtk_wed_reset_dma(struct mtk_wed_device + if (!dev->tx_ring[i].desc) + continue; + +- mtk_wed_ring_reset(&dev->tx_ring[i], MTK_WED_TX_RING_SIZE); ++ mtk_wed_ring_reset(&dev->tx_ring[i], MTK_WED_TX_RING_SIZE, ++ true); + } + + if (mtk_wed_poll_busy(dev)) +@@ -641,6 +1009,9 @@ mtk_wed_reset_dma(struct mtk_wed_device + wdma_w32(dev, MTK_WDMA_RESET_IDX, MTK_WDMA_RESET_IDX_RX); + wdma_w32(dev, MTK_WDMA_RESET_IDX, 0); + ++ if (mtk_wed_get_rx_capa(dev)) ++ mtk_wdma_rx_reset(dev); ++ + if (busy) { + mtk_wed_reset(dev, MTK_WED_RESET_WDMA_INT_AGENT); + mtk_wed_reset(dev, MTK_WED_RESET_WDMA_RX_DRV); +@@ -675,12 +1046,11 @@ mtk_wed_reset_dma(struct mtk_wed_device + MTK_WED_WPDMA_RESET_IDX_RX); + wed_w32(dev, MTK_WED_WPDMA_RESET_IDX, 0); + } +- + } + + static int + mtk_wed_ring_alloc(struct mtk_wed_device *dev, struct mtk_wed_ring *ring, +- int size, u32 desc_size) ++ int size, u32 desc_size, bool tx) + { + ring->desc = dma_alloc_coherent(dev->hw->dev, size * desc_size, + &ring->desc_phys, GFP_KERNEL); +@@ -689,7 +1059,7 @@ mtk_wed_ring_alloc(struct mtk_wed_device + + ring->desc_size = desc_size; + ring->size = size; +- mtk_wed_ring_reset(ring, size); ++ mtk_wed_ring_reset(ring, size, tx); + + return 0; + } +@@ -698,9 +1068,14 @@ static int + mtk_wed_wdma_rx_ring_setup(struct mtk_wed_device *dev, int idx, int size) + { + u32 desc_size = sizeof(struct mtk_wdma_desc) * dev->hw->version; +- struct mtk_wed_ring *wdma = &dev->rx_wdma[idx]; ++ struct mtk_wed_ring *wdma; + +- if (mtk_wed_ring_alloc(dev, wdma, MTK_WED_WDMA_RING_SIZE, desc_size)) ++ if (idx >= ARRAY_SIZE(dev->rx_wdma)) ++ return -EINVAL; ++ ++ wdma = &dev->rx_wdma[idx]; ++ if (mtk_wed_ring_alloc(dev, wdma, MTK_WED_WDMA_RING_SIZE, desc_size, ++ true)) + return -ENOMEM; + + wdma_w32(dev, MTK_WDMA_RING_RX(idx) + MTK_WED_RING_OFS_BASE, +@@ -717,6 +1092,60 @@ mtk_wed_wdma_rx_ring_setup(struct mtk_we + return 0; + } + ++static int ++mtk_wed_wdma_tx_ring_setup(struct mtk_wed_device *dev, int idx, int size) ++{ ++ u32 desc_size = sizeof(struct mtk_wdma_desc) * dev->hw->version; ++ struct mtk_wed_ring *wdma; ++ ++ if (idx >= ARRAY_SIZE(dev->tx_wdma)) ++ return -EINVAL; ++ ++ wdma = &dev->tx_wdma[idx]; ++ if (mtk_wed_ring_alloc(dev, wdma, MTK_WED_WDMA_RING_SIZE, desc_size, ++ true)) ++ return -ENOMEM; ++ ++ wdma_w32(dev, MTK_WDMA_RING_TX(idx) + MTK_WED_RING_OFS_BASE, ++ wdma->desc_phys); ++ wdma_w32(dev, MTK_WDMA_RING_TX(idx) + MTK_WED_RING_OFS_COUNT, ++ size); ++ wdma_w32(dev, MTK_WDMA_RING_TX(idx) + MTK_WED_RING_OFS_CPU_IDX, 0); ++ wdma_w32(dev, MTK_WDMA_RING_TX(idx) + MTK_WED_RING_OFS_DMA_IDX, 0); ++ ++ if (!idx) { ++ wed_w32(dev, MTK_WED_WDMA_RING_TX + MTK_WED_RING_OFS_BASE, ++ wdma->desc_phys); ++ wed_w32(dev, MTK_WED_WDMA_RING_TX + MTK_WED_RING_OFS_COUNT, ++ size); ++ wed_w32(dev, MTK_WED_WDMA_RING_TX + MTK_WED_RING_OFS_CPU_IDX, ++ 0); ++ wed_w32(dev, MTK_WED_WDMA_RING_TX + MTK_WED_RING_OFS_DMA_IDX, ++ 0); ++ } ++ ++ return 0; ++} ++ ++static void ++mtk_wed_ppe_check(struct mtk_wed_device *dev, struct sk_buff *skb, ++ u32 reason, u32 hash) ++{ ++ struct mtk_eth *eth = dev->hw->eth; ++ struct ethhdr *eh; ++ ++ if (!skb) ++ return; ++ ++ if (reason != MTK_PPE_CPU_REASON_HIT_UNBIND_RATE_REACHED) ++ return; ++ ++ skb_set_mac_header(skb, 0); ++ eh = eth_hdr(skb); ++ skb->protocol = eh->h_proto; ++ mtk_ppe_check_skb(eth->ppe[dev->hw->index], skb, hash); ++} ++ + static void + mtk_wed_configure_irq(struct mtk_wed_device *dev, u32 irq_mask) + { +@@ -739,6 +1168,8 @@ mtk_wed_configure_irq(struct mtk_wed_dev + + wed_clr(dev, MTK_WED_WDMA_INT_CTRL, wdma_mask); + } else { ++ wdma_mask |= FIELD_PREP(MTK_WDMA_INT_MASK_TX_DONE, ++ GENMASK(1, 0)); + /* initail tx interrupt trigger */ + wed_w32(dev, MTK_WED_WPDMA_INT_CTRL_TX, + MTK_WED_WPDMA_INT_CTRL_TX0_DONE_EN | +@@ -757,6 +1188,16 @@ mtk_wed_configure_irq(struct mtk_wed_dev + FIELD_PREP(MTK_WED_WPDMA_INT_CTRL_TX_FREE_DONE_TRIG, + dev->wlan.txfree_tbit)); + ++ wed_w32(dev, MTK_WED_WPDMA_INT_CTRL_RX, ++ MTK_WED_WPDMA_INT_CTRL_RX0_EN | ++ MTK_WED_WPDMA_INT_CTRL_RX0_CLR | ++ MTK_WED_WPDMA_INT_CTRL_RX1_EN | ++ MTK_WED_WPDMA_INT_CTRL_RX1_CLR | ++ FIELD_PREP(MTK_WED_WPDMA_INT_CTRL_RX0_DONE_TRIG, ++ dev->wlan.rx_tbit[0]) | ++ FIELD_PREP(MTK_WED_WPDMA_INT_CTRL_RX1_DONE_TRIG, ++ dev->wlan.rx_tbit[1])); ++ + wed_w32(dev, MTK_WED_WDMA_INT_CLR, wdma_mask); + wed_set(dev, MTK_WED_WDMA_INT_CTRL, + FIELD_PREP(MTK_WED_WDMA_INT_CTRL_POLL_SRC_SEL, +@@ -794,9 +1235,15 @@ mtk_wed_dma_enable(struct mtk_wed_device + wdma_set(dev, MTK_WDMA_GLO_CFG, + MTK_WDMA_GLO_CFG_RX_INFO3_PRERES); + } else { ++ int i; ++ + wed_set(dev, MTK_WED_WPDMA_CTRL, + MTK_WED_WPDMA_CTRL_SDL1_FIXED); + ++ wed_set(dev, MTK_WED_WDMA_GLO_CFG, ++ MTK_WED_WDMA_GLO_CFG_TX_DRV_EN | ++ MTK_WED_WDMA_GLO_CFG_TX_DDONE_CHK); ++ + wed_set(dev, MTK_WED_WPDMA_GLO_CFG, + MTK_WED_WPDMA_GLO_CFG_RX_DRV_R0_PKT_PROC | + MTK_WED_WPDMA_GLO_CFG_RX_DRV_R0_CRX_SYNC); +@@ -804,6 +1251,15 @@ mtk_wed_dma_enable(struct mtk_wed_device + wed_clr(dev, MTK_WED_WPDMA_GLO_CFG, + MTK_WED_WPDMA_GLO_CFG_TX_TKID_KEEP | + MTK_WED_WPDMA_GLO_CFG_TX_DMAD_DW3_PREV); ++ ++ wed_set(dev, MTK_WED_WPDMA_RX_D_GLO_CFG, ++ MTK_WED_WPDMA_RX_D_RX_DRV_EN | ++ FIELD_PREP(MTK_WED_WPDMA_RX_D_RXD_READ_LEN, 0x18) | ++ FIELD_PREP(MTK_WED_WPDMA_RX_D_INIT_PHASE_RXEN_SEL, ++ 0x2)); ++ ++ for (i = 0; i < MTK_WED_RX_QUEUES; i++) ++ mtk_wed_check_wfdma_rx_fill(dev, i); + } + } + +@@ -829,7 +1285,19 @@ mtk_wed_start(struct mtk_wed_device *dev + val |= BIT(0) | (BIT(1) * !!dev->hw->index); + regmap_write(dev->hw->mirror, dev->hw->index * 4, val); + } else { +- mtk_wed_set_512_support(dev, true); ++ /* driver set mid ready and only once */ ++ wed_w32(dev, MTK_WED_EXT_INT_MASK1, ++ MTK_WED_EXT_INT_STATUS_WPDMA_MID_RDY); ++ wed_w32(dev, MTK_WED_EXT_INT_MASK2, ++ MTK_WED_EXT_INT_STATUS_WPDMA_MID_RDY); ++ ++ wed_r32(dev, MTK_WED_EXT_INT_MASK1); ++ wed_r32(dev, MTK_WED_EXT_INT_MASK2); ++ ++ if (mtk_wed_rro_cfg(dev)) ++ return; ++ ++ mtk_wed_set_512_support(dev, dev->wlan.wcid_512); + } + + mtk_wed_dma_enable(dev); +@@ -863,7 +1331,7 @@ mtk_wed_attach(struct mtk_wed_device *de + if (!hw) { + module_put(THIS_MODULE); + ret = -ENODEV; +- goto out; ++ goto unlock; + } + + device = dev->wlan.bus_type == MTK_WED_BUS_PCIE +@@ -876,15 +1344,24 @@ mtk_wed_attach(struct mtk_wed_device *de + dev->dev = hw->dev; + dev->irq = hw->irq; + dev->wdma_idx = hw->index; ++ dev->version = hw->version; + + if (hw->eth->dma_dev == hw->eth->dev && + of_dma_is_coherent(hw->eth->dev->of_node)) + mtk_eth_set_dma_device(hw->eth, hw->dev); + +- ret = mtk_wed_buffer_alloc(dev); +- if (ret) { +- mtk_wed_detach(dev); ++ ret = mtk_wed_tx_buffer_alloc(dev); ++ if (ret) + goto out; ++ ++ if (mtk_wed_get_rx_capa(dev)) { ++ ret = mtk_wed_rx_buffer_alloc(dev); ++ if (ret) ++ goto out; ++ ++ ret = mtk_wed_rro_alloc(dev); ++ if (ret) ++ goto out; + } + + mtk_wed_hw_init_early(dev); +@@ -893,8 +1370,10 @@ mtk_wed_attach(struct mtk_wed_device *de + BIT(hw->index), 0); + else + ret = mtk_wed_wo_init(hw); +- + out: ++ if (ret) ++ mtk_wed_detach(dev); ++unlock: + mutex_unlock(&hw_lock); + + return ret; +@@ -917,10 +1396,11 @@ mtk_wed_tx_ring_setup(struct mtk_wed_dev + * WDMA RX. + */ + +- BUG_ON(idx >= ARRAY_SIZE(dev->tx_ring)); ++ if (WARN_ON(idx >= ARRAY_SIZE(dev->tx_ring))) ++ return -EINVAL; + + if (mtk_wed_ring_alloc(dev, ring, MTK_WED_TX_RING_SIZE, +- sizeof(*ring->desc))) ++ sizeof(*ring->desc), true)) + return -ENOMEM; + + if (mtk_wed_wdma_rx_ring_setup(dev, idx, MTK_WED_WDMA_RING_SIZE)) +@@ -967,6 +1447,37 @@ mtk_wed_txfree_ring_setup(struct mtk_wed + return 0; + } + ++static int ++mtk_wed_rx_ring_setup(struct mtk_wed_device *dev, int idx, void __iomem *regs) ++{ ++ struct mtk_wed_ring *ring = &dev->rx_ring[idx]; ++ ++ if (WARN_ON(idx >= ARRAY_SIZE(dev->rx_ring))) ++ return -EINVAL; ++ ++ if (mtk_wed_ring_alloc(dev, ring, MTK_WED_RX_RING_SIZE, ++ sizeof(*ring->desc), false)) ++ return -ENOMEM; ++ ++ if (mtk_wed_wdma_tx_ring_setup(dev, idx, MTK_WED_WDMA_RING_SIZE)) ++ return -ENOMEM; ++ ++ ring->reg_base = MTK_WED_RING_RX_DATA(idx); ++ ring->wpdma = regs; ++ ring->flags |= MTK_WED_RING_CONFIGURED; ++ ++ /* WPDMA -> WED */ ++ wpdma_rx_w32(dev, idx, MTK_WED_RING_OFS_BASE, ring->desc_phys); ++ wpdma_rx_w32(dev, idx, MTK_WED_RING_OFS_COUNT, MTK_WED_RX_RING_SIZE); ++ ++ wed_w32(dev, MTK_WED_WPDMA_RING_RX_DATA(idx) + MTK_WED_RING_OFS_BASE, ++ ring->desc_phys); ++ wed_w32(dev, MTK_WED_WPDMA_RING_RX_DATA(idx) + MTK_WED_RING_OFS_COUNT, ++ MTK_WED_RX_RING_SIZE); ++ ++ return 0; ++} ++ + static u32 + mtk_wed_irq_get(struct mtk_wed_device *dev, u32 mask) + { +@@ -1063,7 +1574,9 @@ void mtk_wed_add_hw(struct device_node * + static const struct mtk_wed_ops wed_ops = { + .attach = mtk_wed_attach, + .tx_ring_setup = mtk_wed_tx_ring_setup, ++ .rx_ring_setup = mtk_wed_rx_ring_setup, + .txfree_ring_setup = mtk_wed_txfree_ring_setup, ++ .msg_update = mtk_wed_mcu_msg_update, + .start = mtk_wed_start, + .stop = mtk_wed_stop, + .reset_dma = mtk_wed_reset_dma, +@@ -1072,6 +1585,7 @@ void mtk_wed_add_hw(struct device_node * + .irq_get = mtk_wed_irq_get, + .irq_set_mask = mtk_wed_irq_set_mask, + .detach = mtk_wed_detach, ++ .ppe_check = mtk_wed_ppe_check, + }; + struct device_node *eth_np = eth->dev->of_node; + struct platform_device *pdev; +--- a/drivers/net/ethernet/mediatek/mtk_wed.h ++++ b/drivers/net/ethernet/mediatek/mtk_wed.h +@@ -87,6 +87,24 @@ wpdma_tx_w32(struct mtk_wed_device *dev, + } + + static inline u32 ++wpdma_rx_r32(struct mtk_wed_device *dev, int ring, u32 reg) ++{ ++ if (!dev->rx_ring[ring].wpdma) ++ return 0; ++ ++ return readl(dev->rx_ring[ring].wpdma + reg); ++} ++ ++static inline void ++wpdma_rx_w32(struct mtk_wed_device *dev, int ring, u32 reg, u32 val) ++{ ++ if (!dev->rx_ring[ring].wpdma) ++ return; ++ ++ writel(val, dev->rx_ring[ring].wpdma + reg); ++} ++ ++static inline u32 + wpdma_txfree_r32(struct mtk_wed_device *dev, u32 reg) + { + if (!dev->txfree_ring.wpdma) +@@ -128,6 +146,7 @@ static inline int mtk_wed_flow_add(int i + static inline void mtk_wed_flow_remove(int index) + { + } ++ + #endif + + #ifdef CONFIG_DEBUG_FS +--- a/drivers/net/ethernet/mediatek/mtk_wed_mcu.c ++++ b/drivers/net/ethernet/mediatek/mtk_wed_mcu.c +@@ -10,6 +10,7 @@ + #include + #include + #include ++#include + + #include "mtk_wed_regs.h" + #include "mtk_wed_wo.h" +@@ -60,24 +61,37 @@ void mtk_wed_mcu_rx_event(struct mtk_wed + wake_up(&wo->mcu.wait); + } + ++static void ++mtk_wed_update_rx_stats(struct mtk_wed_device *wed, struct sk_buff *skb) ++{ ++ u32 count = get_unaligned_le32(skb->data); ++ struct mtk_wed_wo_rx_stats *stats; ++ int i; ++ ++ if (count * sizeof(*stats) > skb->len - sizeof(u32)) ++ return; ++ ++ stats = (struct mtk_wed_wo_rx_stats *)(skb->data + sizeof(u32)); ++ for (i = 0 ; i < count ; i++) ++ wed->wlan.update_wo_rx_stats(wed, &stats[i]); ++} ++ + void mtk_wed_mcu_rx_unsolicited_event(struct mtk_wed_wo *wo, + struct sk_buff *skb) + { + struct mtk_wed_mcu_hdr *hdr = (struct mtk_wed_mcu_hdr *)skb->data; + +- switch (hdr->cmd) { +- case MTK_WED_WO_EVT_LOG_DUMP: { +- const char *msg = (const char *)(skb->data + sizeof(*hdr)); ++ skb_pull(skb, sizeof(*hdr)); + +- dev_notice(wo->hw->dev, "%s\n", msg); ++ switch (hdr->cmd) { ++ case MTK_WED_WO_EVT_LOG_DUMP: ++ dev_notice(wo->hw->dev, "%s\n", skb->data); + break; +- } + case MTK_WED_WO_EVT_PROFILING: { +- struct mtk_wed_wo_log_info *info; +- u32 count = (skb->len - sizeof(*hdr)) / sizeof(*info); ++ struct mtk_wed_wo_log_info *info = (void *)skb->data; ++ u32 count = skb->len / sizeof(*info); + int i; + +- info = (struct mtk_wed_wo_log_info *)(skb->data + sizeof(*hdr)); + for (i = 0 ; i < count ; i++) + dev_notice(wo->hw->dev, + "SN:%u latency: total=%u, rro:%u, mod:%u\n", +@@ -88,6 +102,7 @@ void mtk_wed_mcu_rx_unsolicited_event(st + break; + } + case MTK_WED_WO_EVT_RXCNT_INFO: ++ mtk_wed_update_rx_stats(wo->hw->wed_dev, skb); + break; + default: + break; +@@ -144,6 +159,8 @@ mtk_wed_mcu_parse_response(struct mtk_we + skb_pull(skb, sizeof(*hdr)); + switch (cmd) { + case MTK_WED_WO_CMD_RXCNT_INFO: ++ mtk_wed_update_rx_stats(wo->hw->wed_dev, skb); ++ break; + default: + break; + } +@@ -182,6 +199,18 @@ unlock: + return ret; + } + ++int mtk_wed_mcu_msg_update(struct mtk_wed_device *dev, int id, void *data, ++ int len) ++{ ++ struct mtk_wed_wo *wo = dev->hw->wed_wo; ++ ++ if (dev->hw->version == 1) ++ return 0; ++ ++ return mtk_wed_mcu_send_msg(wo, MTK_WED_MODULE_ID_WO, id, data, len, ++ true); ++} ++ + static int + mtk_wed_get_memory_region(struct mtk_wed_wo *wo, + struct mtk_wed_wo_memory_region *region) +--- a/drivers/net/ethernet/mediatek/mtk_wed_regs.h ++++ b/drivers/net/ethernet/mediatek/mtk_wed_regs.h +@@ -4,6 +4,7 @@ + #ifndef __MTK_WED_REGS_H + #define __MTK_WED_REGS_H + ++#define MTK_WFDMA_DESC_CTRL_TO_HOST BIT(8) + #define MTK_WDMA_DESC_CTRL_LEN1 GENMASK(14, 0) + #define MTK_WDMA_DESC_CTRL_LEN1_V2 GENMASK(13, 0) + #define MTK_WDMA_DESC_CTRL_LAST_SEG1 BIT(15) +@@ -28,6 +29,8 @@ struct mtk_wdma_desc { + #define MTK_WED_RESET_WED_TX_DMA BIT(12) + #define MTK_WED_RESET_WDMA_RX_DRV BIT(17) + #define MTK_WED_RESET_WDMA_INT_AGENT BIT(19) ++#define MTK_WED_RESET_RX_RRO_QM BIT(20) ++#define MTK_WED_RESET_RX_ROUTE_QM BIT(21) + #define MTK_WED_RESET_WED BIT(31) + + #define MTK_WED_CTRL 0x00c +@@ -39,8 +42,12 @@ struct mtk_wdma_desc { + #define MTK_WED_CTRL_WED_TX_BM_BUSY BIT(9) + #define MTK_WED_CTRL_WED_TX_FREE_AGENT_EN BIT(10) + #define MTK_WED_CTRL_WED_TX_FREE_AGENT_BUSY BIT(11) +-#define MTK_WED_CTRL_RESERVE_EN BIT(12) +-#define MTK_WED_CTRL_RESERVE_BUSY BIT(13) ++#define MTK_WED_CTRL_WED_RX_BM_EN BIT(12) ++#define MTK_WED_CTRL_WED_RX_BM_BUSY BIT(13) ++#define MTK_WED_CTRL_RX_RRO_QM_EN BIT(14) ++#define MTK_WED_CTRL_RX_RRO_QM_BUSY BIT(15) ++#define MTK_WED_CTRL_RX_ROUTE_QM_EN BIT(16) ++#define MTK_WED_CTRL_RX_ROUTE_QM_BUSY BIT(17) + #define MTK_WED_CTRL_FINAL_DIDX_READ BIT(24) + #define MTK_WED_CTRL_ETH_DMAD_FMT BIT(25) + #define MTK_WED_CTRL_MIB_READ_CLEAR BIT(28) +@@ -62,6 +69,9 @@ struct mtk_wdma_desc { + #define MTK_WED_EXT_INT_STATUS_TX_DMA_R_RESP_ERR BIT(22) + #define MTK_WED_EXT_INT_STATUS_TX_DMA_W_RESP_ERR BIT(23) + #define MTK_WED_EXT_INT_STATUS_RX_DRV_DMA_RECYCLE BIT(24) ++#define MTK_WED_EXT_INT_STATUS_RX_DRV_GET_BM_DMAD_SKIP BIT(25) ++#define MTK_WED_EXT_INT_STATUS_WPDMA_RX_D_DRV_ERR BIT(26) ++#define MTK_WED_EXT_INT_STATUS_WPDMA_MID_RDY BIT(27) + #define MTK_WED_EXT_INT_STATUS_ERROR_MASK (MTK_WED_EXT_INT_STATUS_TF_LEN_ERR | \ + MTK_WED_EXT_INT_STATUS_TKID_WO_PYLD | \ + MTK_WED_EXT_INT_STATUS_TKID_TITO_INVALID | \ +@@ -71,6 +81,8 @@ struct mtk_wdma_desc { + MTK_WED_EXT_INT_STATUS_TX_DMA_R_RESP_ERR) + + #define MTK_WED_EXT_INT_MASK 0x028 ++#define MTK_WED_EXT_INT_MASK1 0x02c ++#define MTK_WED_EXT_INT_MASK2 0x030 + + #define MTK_WED_STATUS 0x060 + #define MTK_WED_STATUS_TX GENMASK(15, 8) +@@ -151,6 +163,7 @@ struct mtk_wdma_desc { + #define MTK_WED_RING_TX(_n) (0x300 + (_n) * 0x10) + + #define MTK_WED_RING_RX(_n) (0x400 + (_n) * 0x10) ++#define MTK_WED_RING_RX_DATA(_n) (0x420 + (_n) * 0x10) + + #define MTK_WED_SCR0 0x3c0 + #define MTK_WED_WPDMA_INT_TRIGGER 0x504 +@@ -213,6 +226,12 @@ struct mtk_wdma_desc { + #define MTK_WED_WPDMA_INT_CTRL_TX1_DONE_TRIG GENMASK(14, 10) + + #define MTK_WED_WPDMA_INT_CTRL_RX 0x534 ++#define MTK_WED_WPDMA_INT_CTRL_RX0_EN BIT(0) ++#define MTK_WED_WPDMA_INT_CTRL_RX0_CLR BIT(1) ++#define MTK_WED_WPDMA_INT_CTRL_RX0_DONE_TRIG GENMASK(6, 2) ++#define MTK_WED_WPDMA_INT_CTRL_RX1_EN BIT(8) ++#define MTK_WED_WPDMA_INT_CTRL_RX1_CLR BIT(9) ++#define MTK_WED_WPDMA_INT_CTRL_RX1_DONE_TRIG GENMASK(14, 10) + + #define MTK_WED_WPDMA_INT_CTRL_TX_FREE 0x538 + #define MTK_WED_WPDMA_INT_CTRL_TX_FREE_DONE_EN BIT(0) +@@ -242,11 +261,34 @@ struct mtk_wdma_desc { + + #define MTK_WED_WPDMA_RING_TX(_n) (0x600 + (_n) * 0x10) + #define MTK_WED_WPDMA_RING_RX(_n) (0x700 + (_n) * 0x10) ++#define MTK_WED_WPDMA_RING_RX_DATA(_n) (0x730 + (_n) * 0x10) ++ ++#define MTK_WED_WPDMA_RX_D_GLO_CFG 0x75c ++#define MTK_WED_WPDMA_RX_D_RX_DRV_EN BIT(0) ++#define MTK_WED_WPDMA_RX_D_INIT_PHASE_RXEN_SEL GENMASK(11, 7) ++#define MTK_WED_WPDMA_RX_D_RXD_READ_LEN GENMASK(31, 24) ++ ++#define MTK_WED_WPDMA_RX_D_RST_IDX 0x760 ++#define MTK_WED_WPDMA_RX_D_RST_CRX_IDX GENMASK(17, 16) ++#define MTK_WED_WPDMA_RX_D_RST_DRV_IDX GENMASK(25, 24) ++ ++#define MTK_WED_WPDMA_RX_GLO_CFG 0x76c ++#define MTK_WED_WPDMA_RX_RING 0x770 ++ ++#define MTK_WED_WPDMA_RX_D_MIB(_n) (0x774 + (_n) * 4) ++#define MTK_WED_WPDMA_RX_D_PROCESSED_MIB(_n) (0x784 + (_n) * 4) ++#define MTK_WED_WPDMA_RX_D_COHERENT_MIB 0x78c ++ ++#define MTK_WED_WDMA_RING_TX 0x800 ++ ++#define MTK_WED_WDMA_TX_MIB 0x810 ++ + #define MTK_WED_WDMA_RING_RX(_n) (0x900 + (_n) * 0x10) + #define MTK_WED_WDMA_RX_THRES(_n) (0x940 + (_n) * 0x4) + + #define MTK_WED_WDMA_GLO_CFG 0xa04 + #define MTK_WED_WDMA_GLO_CFG_TX_DRV_EN BIT(0) ++#define MTK_WED_WDMA_GLO_CFG_TX_DDONE_CHK BIT(1) + #define MTK_WED_WDMA_GLO_CFG_RX_DRV_EN BIT(2) + #define MTK_WED_WDMA_GLO_CFG_RX_DRV_BUSY BIT(3) + #define MTK_WED_WDMA_GLO_CFG_BT_SIZE GENMASK(5, 4) +@@ -291,6 +333,20 @@ struct mtk_wdma_desc { + #define MTK_WED_WDMA_RX_RECYCLE_MIB(_n) (0xae8 + (_n) * 4) + #define MTK_WED_WDMA_RX_PROCESSED_MIB(_n) (0xaf0 + (_n) * 4) + ++#define MTK_WED_RX_BM_RX_DMAD 0xd80 ++#define MTK_WED_RX_BM_RX_DMAD_SDL0 GENMASK(13, 0) ++ ++#define MTK_WED_RX_BM_BASE 0xd84 ++#define MTK_WED_RX_BM_INIT_PTR 0xd88 ++#define MTK_WED_RX_BM_SW_TAIL GENMASK(15, 0) ++#define MTK_WED_RX_BM_INIT_SW_TAIL BIT(16) ++ ++#define MTK_WED_RX_PTR 0xd8c ++ ++#define MTK_WED_RX_BM_DYN_ALLOC_TH 0xdb4 ++#define MTK_WED_RX_BM_DYN_ALLOC_TH_H GENMASK(31, 16) ++#define MTK_WED_RX_BM_DYN_ALLOC_TH_L GENMASK(15, 0) ++ + #define MTK_WED_RING_OFS_BASE 0x00 + #define MTK_WED_RING_OFS_COUNT 0x04 + #define MTK_WED_RING_OFS_CPU_IDX 0x08 +@@ -301,7 +357,9 @@ struct mtk_wdma_desc { + + #define MTK_WDMA_GLO_CFG 0x204 + #define MTK_WDMA_GLO_CFG_TX_DMA_EN BIT(0) ++#define MTK_WDMA_GLO_CFG_TX_DMA_BUSY BIT(1) + #define MTK_WDMA_GLO_CFG_RX_DMA_EN BIT(2) ++#define MTK_WDMA_GLO_CFG_RX_DMA_BUSY BIT(3) + #define MTK_WDMA_GLO_CFG_RX_INFO3_PRERES BIT(26) + #define MTK_WDMA_GLO_CFG_RX_INFO2_PRERES BIT(27) + #define MTK_WDMA_GLO_CFG_RX_INFO1_PRERES BIT(28) +@@ -330,4 +388,70 @@ struct mtk_wdma_desc { + /* DMA channel mapping */ + #define HIFSYS_DMA_AG_MAP 0x008 + ++#define MTK_WED_RTQM_GLO_CFG 0xb00 ++#define MTK_WED_RTQM_BUSY BIT(1) ++#define MTK_WED_RTQM_Q_RST BIT(2) ++#define MTK_WED_RTQM_Q_DBG_BYPASS BIT(5) ++#define MTK_WED_RTQM_TXDMAD_FPORT GENMASK(23, 20) ++ ++#define MTK_WED_RTQM_R2H_MIB(_n) (0xb70 + (_n) * 0x4) ++#define MTK_WED_RTQM_R2Q_MIB(_n) (0xb78 + (_n) * 0x4) ++#define MTK_WED_RTQM_Q2N_MIB 0xb80 ++#define MTK_WED_RTQM_Q2H_MIB(_n) (0xb84 + (_n) * 0x4) ++ ++#define MTK_WED_RTQM_Q2B_MIB 0xb8c ++#define MTK_WED_RTQM_PFDBK_MIB 0xb90 ++ ++#define MTK_WED_RROQM_GLO_CFG 0xc04 ++#define MTK_WED_RROQM_RST_IDX 0xc08 ++#define MTK_WED_RROQM_RST_IDX_MIOD BIT(0) ++#define MTK_WED_RROQM_RST_IDX_FDBK BIT(4) ++ ++#define MTK_WED_RROQM_MIOD_CTRL0 0xc40 ++#define MTK_WED_RROQM_MIOD_CTRL1 0xc44 ++#define MTK_WED_RROQM_MIOD_CNT GENMASK(11, 0) ++ ++#define MTK_WED_RROQM_MIOD_CTRL2 0xc48 ++#define MTK_WED_RROQM_MIOD_CTRL3 0xc4c ++ ++#define MTK_WED_RROQM_FDBK_CTRL0 0xc50 ++#define MTK_WED_RROQM_FDBK_CTRL1 0xc54 ++#define MTK_WED_RROQM_FDBK_CNT GENMASK(11, 0) ++ ++#define MTK_WED_RROQM_FDBK_CTRL2 0xc58 ++ ++#define MTK_WED_RROQ_BASE_L 0xc80 ++#define MTK_WED_RROQ_BASE_H 0xc84 ++ ++#define MTK_WED_RROQM_MIOD_CFG 0xc8c ++#define MTK_WED_RROQM_MIOD_MID_DW GENMASK(5, 0) ++#define MTK_WED_RROQM_MIOD_MOD_DW GENMASK(13, 8) ++#define MTK_WED_RROQM_MIOD_ENTRY_DW GENMASK(22, 16) ++ ++#define MTK_WED_RROQM_MID_MIB 0xcc0 ++#define MTK_WED_RROQM_MOD_MIB 0xcc4 ++#define MTK_WED_RROQM_MOD_COHERENT_MIB 0xcc8 ++#define MTK_WED_RROQM_FDBK_MIB 0xcd0 ++#define MTK_WED_RROQM_FDBK_COHERENT_MIB 0xcd4 ++#define MTK_WED_RROQM_FDBK_IND_MIB 0xce0 ++#define MTK_WED_RROQM_FDBK_ENQ_MIB 0xce4 ++#define MTK_WED_RROQM_FDBK_ANC_MIB 0xce8 ++#define MTK_WED_RROQM_FDBK_ANC2H_MIB 0xcec ++ ++#define MTK_WED_RX_BM_RX_DMAD 0xd80 ++#define MTK_WED_RX_BM_BASE 0xd84 ++#define MTK_WED_RX_BM_INIT_PTR 0xd88 ++#define MTK_WED_RX_BM_PTR 0xd8c ++#define MTK_WED_RX_BM_PTR_HEAD GENMASK(32, 16) ++#define MTK_WED_RX_BM_PTR_TAIL GENMASK(15, 0) ++ ++#define MTK_WED_RX_BM_BLEN 0xd90 ++#define MTK_WED_RX_BM_STS 0xd94 ++#define MTK_WED_RX_BM_INTF2 0xd98 ++#define MTK_WED_RX_BM_INTF 0xd9c ++#define MTK_WED_RX_BM_ERR_STS 0xda8 ++ ++#define MTK_WED_WOCPU_VIEW_MIOD_BASE 0x8000 ++#define MTK_WED_PCIE_INT_MASK 0x0 ++ + #endif +--- a/drivers/net/ethernet/mediatek/mtk_wed_wo.h ++++ b/drivers/net/ethernet/mediatek/mtk_wed_wo.h +@@ -49,6 +49,10 @@ enum { + MTK_WED_WARP_CMD_FLAG_FROM_TO_WO = BIT(2), + }; + ++#define MTK_WED_WO_CPU_MCUSYS_RESET_ADDR 0x15194050 ++#define MTK_WED_WO_CPU_WO0_MCUSYS_RESET_MASK 0x20 ++#define MTK_WED_WO_CPU_WO1_MCUSYS_RESET_MASK 0x1 ++ + enum { + MTK_WED_WO_REGION_EMI, + MTK_WED_WO_REGION_ILM, +@@ -57,6 +61,28 @@ enum { + __MTK_WED_WO_REGION_MAX, + }; + ++enum mtk_wed_wo_state { ++ MTK_WED_WO_STATE_UNDEFINED, ++ MTK_WED_WO_STATE_INIT, ++ MTK_WED_WO_STATE_ENABLE, ++ MTK_WED_WO_STATE_DISABLE, ++ MTK_WED_WO_STATE_HALT, ++ MTK_WED_WO_STATE_GATING, ++ MTK_WED_WO_STATE_SER_RESET, ++ MTK_WED_WO_STATE_WF_RESET, ++}; ++ ++enum mtk_wed_wo_done_state { ++ MTK_WED_WOIF_UNDEFINED, ++ MTK_WED_WOIF_DISABLE_DONE, ++ MTK_WED_WOIF_TRIGGER_ENABLE, ++ MTK_WED_WOIF_ENABLE_DONE, ++ MTK_WED_WOIF_TRIGGER_GATING, ++ MTK_WED_WOIF_GATING_DONE, ++ MTK_WED_WOIF_TRIGGER_HALT, ++ MTK_WED_WOIF_HALT_DONE, ++}; ++ + enum mtk_wed_dummy_cr_idx { + MTK_WED_DUMMY_CR_FWDL, + MTK_WED_DUMMY_CR_WO_STATUS, +@@ -245,6 +271,8 @@ void mtk_wed_mcu_rx_unsolicited_event(st + struct sk_buff *skb); + int mtk_wed_mcu_send_msg(struct mtk_wed_wo *wo, int id, int cmd, + const void *data, int len, bool wait_resp); ++int mtk_wed_mcu_msg_update(struct mtk_wed_device *dev, int id, void *data, ++ int len); + int mtk_wed_mcu_init(struct mtk_wed_wo *wo); + int mtk_wed_wo_init(struct mtk_wed_hw *hw); + void mtk_wed_wo_deinit(struct mtk_wed_hw *hw); +--- a/include/linux/soc/mediatek/mtk_wed.h ++++ b/include/linux/soc/mediatek/mtk_wed.h +@@ -5,10 +5,13 @@ + #include + #include + #include ++#include + + #define MTK_WED_TX_QUEUES 2 + #define MTK_WED_RX_QUEUES 2 + ++#define WED_WO_STA_REC 0x6 ++ + struct mtk_wed_hw; + struct mtk_wdma_desc; + +@@ -41,21 +44,37 @@ enum mtk_wed_wo_cmd { + MTK_WED_WO_CMD_WED_END + }; + ++struct mtk_rxbm_desc { ++ __le32 buf0; ++ __le32 token; ++} __packed __aligned(4); ++ + enum mtk_wed_bus_tye { + MTK_WED_BUS_PCIE, + MTK_WED_BUS_AXI, + }; + ++#define MTK_WED_RING_CONFIGURED BIT(0) + struct mtk_wed_ring { + struct mtk_wdma_desc *desc; + dma_addr_t desc_phys; + u32 desc_size; + int size; ++ u32 flags; + + u32 reg_base; + void __iomem *wpdma; + }; + ++struct mtk_wed_wo_rx_stats { ++ __le16 wlan_idx; ++ __le16 tid; ++ __le32 rx_pkt_cnt; ++ __le32 rx_byte_cnt; ++ __le32 rx_err_cnt; ++ __le32 rx_drop_cnt; ++}; ++ + struct mtk_wed_device { + #ifdef CONFIG_NET_MEDIATEK_SOC_WED + const struct mtk_wed_ops *ops; +@@ -64,9 +83,12 @@ struct mtk_wed_device { + bool init_done, running; + int wdma_idx; + int irq; ++ u8 version; + + struct mtk_wed_ring tx_ring[MTK_WED_TX_QUEUES]; ++ struct mtk_wed_ring rx_ring[MTK_WED_RX_QUEUES]; + struct mtk_wed_ring txfree_ring; ++ struct mtk_wed_ring tx_wdma[MTK_WED_TX_QUEUES]; + struct mtk_wed_ring rx_wdma[MTK_WED_RX_QUEUES]; + + struct { +@@ -74,7 +96,20 @@ struct mtk_wed_device { + void **pages; + struct mtk_wdma_desc *desc; + dma_addr_t desc_phys; +- } buf_ring; ++ } tx_buf_ring; ++ ++ struct { ++ int size; ++ struct page_frag_cache rx_page; ++ struct mtk_rxbm_desc *desc; ++ dma_addr_t desc_phys; ++ } rx_buf_ring; ++ ++ struct { ++ struct mtk_wed_ring ring; ++ dma_addr_t miod_phys; ++ dma_addr_t fdbk_phys; ++ } rro; + + /* filled by driver: */ + struct { +@@ -83,22 +118,36 @@ struct mtk_wed_device { + struct pci_dev *pci_dev; + }; + enum mtk_wed_bus_tye bus_type; ++ void __iomem *base; ++ u32 phy_base; + + u32 wpdma_phys; + u32 wpdma_int; + u32 wpdma_mask; + u32 wpdma_tx; + u32 wpdma_txfree; ++ u32 wpdma_rx_glo; ++ u32 wpdma_rx; ++ ++ bool wcid_512; + + u16 token_start; + unsigned int nbuf; ++ unsigned int rx_nbuf; ++ unsigned int rx_npkt; ++ unsigned int rx_size; + + u8 tx_tbit[MTK_WED_TX_QUEUES]; ++ u8 rx_tbit[MTK_WED_RX_QUEUES]; + u8 txfree_tbit; + + u32 (*init_buf)(void *ptr, dma_addr_t phys, int token_id); + int (*offload_enable)(struct mtk_wed_device *wed); + void (*offload_disable)(struct mtk_wed_device *wed); ++ u32 (*init_rx_buf)(struct mtk_wed_device *wed, int size); ++ void (*release_rx_buf)(struct mtk_wed_device *wed); ++ void (*update_wo_rx_stats)(struct mtk_wed_device *wed, ++ struct mtk_wed_wo_rx_stats *stats); + } wlan; + #endif + }; +@@ -107,9 +156,15 @@ struct mtk_wed_ops { + int (*attach)(struct mtk_wed_device *dev); + int (*tx_ring_setup)(struct mtk_wed_device *dev, int ring, + void __iomem *regs); ++ int (*rx_ring_setup)(struct mtk_wed_device *dev, int ring, ++ void __iomem *regs); + int (*txfree_ring_setup)(struct mtk_wed_device *dev, + void __iomem *regs); ++ int (*msg_update)(struct mtk_wed_device *dev, int cmd_id, ++ void *data, int len); + void (*detach)(struct mtk_wed_device *dev); ++ void (*ppe_check)(struct mtk_wed_device *dev, struct sk_buff *skb, ++ u32 reason, u32 hash); + + void (*stop)(struct mtk_wed_device *dev); + void (*start)(struct mtk_wed_device *dev, u32 irq_mask); +@@ -144,6 +199,16 @@ mtk_wed_device_attach(struct mtk_wed_dev + return ret; + } + ++static inline bool ++mtk_wed_get_rx_capa(struct mtk_wed_device *dev) ++{ ++#ifdef CONFIG_NET_MEDIATEK_SOC_WED ++ return dev->version != 1; ++#else ++ return false; ++#endif ++} ++ + #ifdef CONFIG_NET_MEDIATEK_SOC_WED + #define mtk_wed_device_active(_dev) !!(_dev)->ops + #define mtk_wed_device_detach(_dev) (_dev)->ops->detach(_dev) +@@ -160,6 +225,12 @@ mtk_wed_device_attach(struct mtk_wed_dev + (_dev)->ops->irq_get(_dev, _mask) + #define mtk_wed_device_irq_set_mask(_dev, _mask) \ + (_dev)->ops->irq_set_mask(_dev, _mask) ++#define mtk_wed_device_rx_ring_setup(_dev, _ring, _regs) \ ++ (_dev)->ops->rx_ring_setup(_dev, _ring, _regs) ++#define mtk_wed_device_ppe_check(_dev, _skb, _reason, _hash) \ ++ (_dev)->ops->ppe_check(_dev, _skb, _reason, _hash) ++#define mtk_wed_device_update_msg(_dev, _id, _msg, _len) \ ++ (_dev)->ops->msg_update(_dev, _id, _msg, _len) + #else + static inline bool mtk_wed_device_active(struct mtk_wed_device *dev) + { +@@ -173,6 +244,9 @@ static inline bool mtk_wed_device_active + #define mtk_wed_device_reg_write(_dev, _reg, _val) do {} while (0) + #define mtk_wed_device_irq_get(_dev, _mask) 0 + #define mtk_wed_device_irq_set_mask(_dev, _mask) do {} while (0) ++#define mtk_wed_device_rx_ring_setup(_dev, _ring, _regs) -ENODEV ++#define mtk_wed_device_ppe_check(_dev, _skb, _reason, _hash) do {} while (0) ++#define mtk_wed_device_update_msg(_dev, _id, _msg, _len) -ENODEV + #endif + + #endif diff --git a/target/linux/generic/pending-5.15/733-05-net-ethernet-mtk_wed-add-rx-mib-counters.patch b/target/linux/generic/pending-5.15/733-05-net-ethernet-mtk_wed-add-rx-mib-counters.patch new file mode 100644 index 000000000..bb1066dec --- /dev/null +++ b/target/linux/generic/pending-5.15/733-05-net-ethernet-mtk_wed-add-rx-mib-counters.patch @@ -0,0 +1,149 @@ +From: Lorenzo Bianconi +Date: Sat, 5 Nov 2022 23:36:22 +0100 +Subject: [PATCH] net: ethernet: mtk_wed: add rx mib counters + +Introduce WED RX MIB counters support available on MT7986a SoC. + +Tested-by: Daniel Golle +Co-developed-by: Sujuan Chen +Signed-off-by: Sujuan Chen +Signed-off-by: Lorenzo Bianconi +Signed-off-by: David S. Miller +--- + +--- a/drivers/net/ethernet/mediatek/mtk_wed_debugfs.c ++++ b/drivers/net/ethernet/mediatek/mtk_wed_debugfs.c +@@ -2,6 +2,7 @@ + /* Copyright (C) 2021 Felix Fietkau */ + + #include ++#include + #include "mtk_wed.h" + #include "mtk_wed_regs.h" + +@@ -18,6 +19,8 @@ enum { + DUMP_TYPE_WDMA, + DUMP_TYPE_WPDMA_TX, + DUMP_TYPE_WPDMA_TXFREE, ++ DUMP_TYPE_WPDMA_RX, ++ DUMP_TYPE_WED_RRO, + }; + + #define DUMP_STR(_str) { _str, 0, DUMP_TYPE_STRING } +@@ -36,6 +39,9 @@ enum { + + #define DUMP_WPDMA_TX_RING(_n) DUMP_RING("WPDMA_TX" #_n, 0, DUMP_TYPE_WPDMA_TX, _n) + #define DUMP_WPDMA_TXFREE_RING DUMP_RING("WPDMA_RX1", 0, DUMP_TYPE_WPDMA_TXFREE) ++#define DUMP_WPDMA_RX_RING(_n) DUMP_RING("WPDMA_RX" #_n, 0, DUMP_TYPE_WPDMA_RX, _n) ++#define DUMP_WED_RRO_RING(_base)DUMP_RING("WED_RRO_MIOD", MTK_##_base, DUMP_TYPE_WED_RRO) ++#define DUMP_WED_RRO_FDBK(_base)DUMP_RING("WED_RRO_FDBK", MTK_##_base, DUMP_TYPE_WED_RRO) + + static void + print_reg_val(struct seq_file *s, const char *name, u32 val) +@@ -57,6 +63,7 @@ dump_wed_regs(struct seq_file *s, struct + cur > regs ? "\n" : "", + cur->name); + continue; ++ case DUMP_TYPE_WED_RRO: + case DUMP_TYPE_WED: + val = wed_r32(dev, cur->offset); + break; +@@ -69,6 +76,9 @@ dump_wed_regs(struct seq_file *s, struct + case DUMP_TYPE_WPDMA_TXFREE: + val = wpdma_txfree_r32(dev, cur->offset); + break; ++ case DUMP_TYPE_WPDMA_RX: ++ val = wpdma_rx_r32(dev, cur->base, cur->offset); ++ break; + } + print_reg_val(s, cur->name, val); + } +@@ -132,6 +142,80 @@ wed_txinfo_show(struct seq_file *s, void + } + DEFINE_SHOW_ATTRIBUTE(wed_txinfo); + ++static int ++wed_rxinfo_show(struct seq_file *s, void *data) ++{ ++ static const struct reg_dump regs[] = { ++ DUMP_STR("WPDMA RX"), ++ DUMP_WPDMA_RX_RING(0), ++ DUMP_WPDMA_RX_RING(1), ++ ++ DUMP_STR("WPDMA RX"), ++ DUMP_WED(WED_WPDMA_RX_D_MIB(0)), ++ DUMP_WED_RING(WED_WPDMA_RING_RX_DATA(0)), ++ DUMP_WED(WED_WPDMA_RX_D_PROCESSED_MIB(0)), ++ DUMP_WED(WED_WPDMA_RX_D_MIB(1)), ++ DUMP_WED_RING(WED_WPDMA_RING_RX_DATA(1)), ++ DUMP_WED(WED_WPDMA_RX_D_PROCESSED_MIB(1)), ++ DUMP_WED(WED_WPDMA_RX_D_COHERENT_MIB), ++ ++ DUMP_STR("WED RX"), ++ DUMP_WED_RING(WED_RING_RX_DATA(0)), ++ DUMP_WED_RING(WED_RING_RX_DATA(1)), ++ ++ DUMP_STR("WED RRO"), ++ DUMP_WED_RRO_RING(WED_RROQM_MIOD_CTRL0), ++ DUMP_WED(WED_RROQM_MID_MIB), ++ DUMP_WED(WED_RROQM_MOD_MIB), ++ DUMP_WED(WED_RROQM_MOD_COHERENT_MIB), ++ DUMP_WED_RRO_FDBK(WED_RROQM_FDBK_CTRL0), ++ DUMP_WED(WED_RROQM_FDBK_IND_MIB), ++ DUMP_WED(WED_RROQM_FDBK_ENQ_MIB), ++ DUMP_WED(WED_RROQM_FDBK_ANC_MIB), ++ DUMP_WED(WED_RROQM_FDBK_ANC2H_MIB), ++ ++ DUMP_STR("WED Route QM"), ++ DUMP_WED(WED_RTQM_R2H_MIB(0)), ++ DUMP_WED(WED_RTQM_R2Q_MIB(0)), ++ DUMP_WED(WED_RTQM_Q2H_MIB(0)), ++ DUMP_WED(WED_RTQM_R2H_MIB(1)), ++ DUMP_WED(WED_RTQM_R2Q_MIB(1)), ++ DUMP_WED(WED_RTQM_Q2H_MIB(1)), ++ DUMP_WED(WED_RTQM_Q2N_MIB), ++ DUMP_WED(WED_RTQM_Q2B_MIB), ++ DUMP_WED(WED_RTQM_PFDBK_MIB), ++ ++ DUMP_STR("WED WDMA TX"), ++ DUMP_WED(WED_WDMA_TX_MIB), ++ DUMP_WED_RING(WED_WDMA_RING_TX), ++ ++ DUMP_STR("WDMA TX"), ++ DUMP_WDMA(WDMA_GLO_CFG), ++ DUMP_WDMA_RING(WDMA_RING_TX(0)), ++ DUMP_WDMA_RING(WDMA_RING_TX(1)), ++ ++ DUMP_STR("WED RX BM"), ++ DUMP_WED(WED_RX_BM_BASE), ++ DUMP_WED(WED_RX_BM_RX_DMAD), ++ DUMP_WED(WED_RX_BM_PTR), ++ DUMP_WED(WED_RX_BM_TKID_MIB), ++ DUMP_WED(WED_RX_BM_BLEN), ++ DUMP_WED(WED_RX_BM_STS), ++ DUMP_WED(WED_RX_BM_INTF2), ++ DUMP_WED(WED_RX_BM_INTF), ++ DUMP_WED(WED_RX_BM_ERR_STS), ++ }; ++ struct mtk_wed_hw *hw = s->private; ++ struct mtk_wed_device *dev = hw->wed_dev; ++ ++ if (!dev) ++ return 0; ++ ++ dump_wed_regs(s, dev, regs, ARRAY_SIZE(regs)); ++ ++ return 0; ++} ++DEFINE_SHOW_ATTRIBUTE(wed_rxinfo); + + static int + mtk_wed_reg_set(void *data, u64 val) +@@ -175,4 +259,7 @@ void mtk_wed_hw_add_debugfs(struct mtk_w + debugfs_create_u32("regidx", 0600, dir, &hw->debugfs_reg); + debugfs_create_file_unsafe("regval", 0600, dir, hw, &fops_regval); + debugfs_create_file_unsafe("txinfo", 0400, dir, hw, &wed_txinfo_fops); ++ if (hw->version != 1) ++ debugfs_create_file_unsafe("rxinfo", 0400, dir, hw, ++ &wed_rxinfo_fops); + }