298977887/lede
1
0
Fork 0
mirror of https://github.com/coolsnowwolf/lede.git synced 2025-04-16 04:13:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

kernel: add WED rx support for mediatek on mt76 with MT7986 and MT7915

Browse Source
This commit is contained in:
coolsnowwolf 2022-11-27 01:49:13 +08:00
parent da4a9aaee0
commit b1e4d3d413
20 changed files with 4724 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
target/linux/generic/pending-5.15/731-net-ethernet-mediatek-ppe-add-support-for-flow-accou.patch
View File

@ -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(&eth->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;

22
target/linux/generic/pending-5.15/732-01-net-ethernet-mtk_eth_soc-account-for-vlan-in-rx-head.patch Normal file
View File

@ -0,0 +1,22 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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 | \

143
target/linux/generic/pending-5.15/732-02-net-ethernet-mtk_eth_soc-increase-tx-ring-side-for-Q.patch Normal file
View File

@ -0,0 +1,143 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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 = &eth->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)

52
target/linux/generic/pending-5.15/732-03-net-ethernet-mtk_eth_soc-avoid-port_mg-assignment-on.patch Normal file
View File

@ -0,0 +1,52 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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);
}

654
target/linux/generic/pending-5.15/732-04-net-ethernet-mtk_eth_soc-implement-multi-queue-suppo.patch Normal file
View File

@ -0,0 +1,654 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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(&eth->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(&eth->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 = &eth->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 = &eth->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 = &eth->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 = &eth->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(&eth->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 <linux/bpf_trace.h>
#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 */

20
target/linux/generic/pending-5.15/732-05-net-dsa-tag_mtk-assign-per-port-queues.patch Normal file
View File

@ -0,0 +1,20 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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

93
target/linux/generic/pending-5.15/732-06-net-ethernet-mediatek-ppe-assign-per-port-queues-for.patch Normal file
View File

@ -0,0 +1,93 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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);

28
target/linux/generic/pending-5.15/732-07-net-ethernet-mtk_eth_soc-compile-out-netsys-v2-code-.patch Normal file
View File

@ -0,0 +1,28 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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 | \

72
target/linux/generic/pending-5.15/732-08-net-dsa-add-support-for-DSA-rx-offloading-via-metada.patch Normal file
View File

@ -0,0 +1,72 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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 <linux/phy_fixed.h>
#include <linux/ptp_classify.h>
#include <linux/etherdevice.h>
+#include <net/dst_metadata.h>
#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;

192
target/linux/generic/pending-5.15/732-09-net-ethernet-mtk_eth_soc-fix-VLAN-rx-hardware-accele.patch Normal file
View File

@ -0,0 +1,192 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -23,6 +23,7 @@
#include <linux/jhash.h>
#include <linux/bitfield.h>
#include <net/dsa.h>
+#include <net/dst_metadata.h>
#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, &eth->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 <linux/bpf_trace.h>
#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;

78
target/linux/generic/pending-5.15/732-10-net-ethernet-mtk_eth_soc-work-around-issue-with-send.patch Normal file
View File

@ -0,0 +1,78 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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 = &eth->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);

21
target/linux/generic/pending-5.15/732-11-net-ethernet-mtk_eth_soc-set-NETIF_F_ALL_TSO.patch Normal file
View File

@ -0,0 +1,21 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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)

37
target/linux/generic/pending-5.15/732-12-net-ethernet-mtk_eth_soc-drop-packets-to-WDMA-if-the.patch Normal file
View File

@ -0,0 +1,37 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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))

52
target/linux/generic/pending-5.15/732-13-net-ethernet-mtk_eth_soc-fix-flow_offload-related-re.patch Normal file
View File

@ -0,0 +1,52 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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);

181
target/linux/generic/pending-5.15/732-14-net-ethernet-mtk_eth_soc-drop-generic-vlan-rx-offloa.patch Normal file
View File

@ -0,0 +1,181 @@
From: Felix Fietkau <nbd@nbd.name>
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 <nbd@nbd.name>
---
--- 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, &eth->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(&eth->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)

591
target/linux/generic/pending-5.15/733-01-net-ethernet-mtk_wed-introduce-wed-mcu-support.patch Normal file
View File

@ -0,0 +1,591 @@
From: Sujuan Chen <sujuan.chen@mediatek.com>
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 <daniel@makrotopia.org>
Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
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 <lorenzo@kernel.org>
+ * Sujuan Chen <sujuan.chen@mediatek.com>
+ */
+
+#include <linux/firmware.h>
+#include <linux/of_address.h>
+#include <linux/of_reserved_mem.h>
+#include <linux/mfd/syscon.h>
+#include <linux/soc/mediatek/mtk_wed.h>
+
+#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 <lorenzo@kernel.org> */
+
+#ifndef __MTK_WED_WO_H
+#define __MTK_WED_WO_H
+
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+
+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,

737
target/linux/generic/pending-5.15/733-02-net-ethernet-mtk_wed-introduce-wed-wo-support.patch Normal file
View File

@ -0,0 +1,737 @@
From: Lorenzo Bianconi <lorenzo@kernel.org>
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 <daniel@makrotopia.org>
Co-developed-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
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 <linux/netdevice.h>
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 <lorenzo@kernel.org>
+ * Sujuan Chen <sujuan.chen@mediatek.com>
+ */
+
+#include <linux/kernel.h>
+#include <linux/dma-mapping.h>
+#include <linux/of_platform.h>
+#include <linux/interrupt.h>
+#include <linux/of_address.h>
+#include <linux/mfd/syscon.h>
+#include <linux/of_irq.h>
+#include <linux/bitfield.h>
+
+#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,
+ &regs);
+ 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,
+ &regs);
+ 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 */

79
target/linux/generic/pending-5.15/733-03-net-ethernet-mtk_wed-rename-tx_wdma-array-in-rx_wdma.patch Normal file
View File

@ -0,0 +1,79 @@
From: Lorenzo Bianconi <lorenzo@kernel.org>
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 <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
--- 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 <linux/pci.h>
#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;

1521
target/linux/generic/pending-5.15/733-04-net-ethernet-mtk_wed-add-configure-wed-wo-support.patch Normal file
View File

File diff suppressed because it is too large Load Diff

149
target/linux/generic/pending-5.15/733-05-net-ethernet-mtk_wed-add-rx-mib-counters.patch Normal file
View File

@ -0,0 +1,149 @@
From: Lorenzo Bianconi <lorenzo@kernel.org>
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 <daniel@makrotopia.org>
Co-developed-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Sujuan Chen <sujuan.chen@mediatek.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
--- 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 <nbd@nbd.name> */
#include <linux/seq_file.h>
+#include <linux/soc/mediatek/mtk_wed.h>
#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);
}