From 35244430d62422e96b4c9c9b4c00c0ebece91d73 Mon Sep 17 00:00:00 2001 From: Jian Shen Date: Thu, 10 Dec 2020 11:42:06 +0800 Subject: [PATCH 1/7] net: hns3: refine the struct hane3_tc_info Currently, there are multiple members related to tc information in struct hnae3_knic_private_info. Merge them into a new struct hnae3_tc_info. Signed-off-by: Jian Shen Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 17 +++---- .../ethernet/hisilicon/hns3/hns3_debugfs.c | 3 +- .../net/ethernet/hisilicon/hns3/hns3_enet.c | 33 ++++++------- .../hisilicon/hns3/hns3pf/hclge_debugfs.c | 2 +- .../hisilicon/hns3/hns3pf/hclge_main.c | 4 +- .../hisilicon/hns3/hns3pf/hclge_mbx.c | 2 +- .../ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 49 +++++++++---------- .../hisilicon/hns3/hns3vf/hclgevf_main.c | 21 ++++---- 8 files changed, 64 insertions(+), 67 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 1a09b1fcb9a5d..0cb80ef4aba5c 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -659,15 +659,16 @@ struct hnae3_ae_algo { #define HNAE3_INT_NAME_LEN 32 #define HNAE3_ITR_COUNTDOWN_START 100 +#define HNAE3_MAX_TC 8 +#define HNAE3_MAX_USER_PRIO 8 struct hnae3_tc_info { - u16 tqp_offset; /* TQP offset from base TQP */ - u16 tqp_count; /* Total TQPs */ - u8 tc; /* TC index */ - bool enable; /* If this TC is enable or not */ + u8 prio_tc[HNAE3_MAX_USER_PRIO]; /* TC indexed by prio */ + u16 tqp_count[HNAE3_MAX_TC]; + u16 tqp_offset[HNAE3_MAX_TC]; + unsigned long tc_en; /* bitmap of TC enabled */ + u8 num_tc; /* Total number of enabled TCs */ }; -#define HNAE3_MAX_TC 8 -#define HNAE3_MAX_USER_PRIO 8 struct hnae3_knic_private_info { struct net_device *netdev; /* Set by KNIC client when init instance */ u16 rss_size; /* Allocated RSS queues */ @@ -676,9 +677,7 @@ struct hnae3_knic_private_info { u16 num_tx_desc; u16 num_rx_desc; - u8 num_tc; /* Total number of enabled TCs */ - u8 prio_tc[HNAE3_MAX_USER_PRIO]; /* TC indexed by prio */ - struct hnae3_tc_info tc_info[HNAE3_MAX_TC]; /* Idx of array is HW TC */ + struct hnae3_tc_info tc_info; u16 num_tqps; /* total number of TQPs in this handle */ struct hnae3_queue **tqp; /* array base of all TQPs in this instance */ diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c index cb26742e2ed8c..9d4e9c053a8fe 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c @@ -385,7 +385,8 @@ static void hns3_dbg_dev_specs(struct hnae3_handle *h) dev_info(priv->dev, "RX buffer length: %u\n", kinfo->rx_buf_len); dev_info(priv->dev, "Desc num per TX queue: %u\n", kinfo->num_tx_desc); dev_info(priv->dev, "Desc num per RX queue: %u\n", kinfo->num_rx_desc); - dev_info(priv->dev, "Total number of enabled TCs: %u\n", kinfo->num_tc); + dev_info(priv->dev, "Total number of enabled TCs: %u\n", + kinfo->tc_info.num_tc); dev_info(priv->dev, "MAX INT QL: %u\n", dev_specs->int_ql_max); dev_info(priv->dev, "MAX INT GL: %u\n", dev_specs->max_int_gl); } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index 4c2fb86889565..36e74ad5b612f 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -323,13 +323,14 @@ static int hns3_nic_set_real_num_queue(struct net_device *netdev) { struct hnae3_handle *h = hns3_get_handle(netdev); struct hnae3_knic_private_info *kinfo = &h->kinfo; - unsigned int queue_size = kinfo->rss_size * kinfo->num_tc; + struct hnae3_tc_info *tc_info = &kinfo->tc_info; + unsigned int queue_size = kinfo->rss_size * tc_info->num_tc; int i, ret; - if (kinfo->num_tc <= 1) { + if (tc_info->num_tc <= 1) { netdev_reset_tc(netdev); } else { - ret = netdev_set_num_tc(netdev, kinfo->num_tc); + ret = netdev_set_num_tc(netdev, tc_info->num_tc); if (ret) { netdev_err(netdev, "netdev_set_num_tc fail, ret=%d!\n", ret); @@ -337,13 +338,11 @@ static int hns3_nic_set_real_num_queue(struct net_device *netdev) } for (i = 0; i < HNAE3_MAX_TC; i++) { - if (!kinfo->tc_info[i].enable) + if (!test_bit(i, &tc_info->tc_en)) continue; - netdev_set_tc_queue(netdev, - kinfo->tc_info[i].tc, - kinfo->tc_info[i].tqp_count, - kinfo->tc_info[i].tqp_offset); + netdev_set_tc_queue(netdev, i, tc_info->tqp_count[i], + tc_info->tqp_offset[i]); } } @@ -369,7 +368,7 @@ static u16 hns3_get_max_available_channels(struct hnae3_handle *h) u16 alloc_tqps, max_rss_size, rss_size; h->ae_algo->ops->get_tqps_and_rss_info(h, &alloc_tqps, &max_rss_size); - rss_size = alloc_tqps / h->kinfo.num_tc; + rss_size = alloc_tqps / h->kinfo.tc_info.num_tc; return min_t(u16, rss_size, max_rss_size); } @@ -508,7 +507,7 @@ static int hns3_nic_net_open(struct net_device *netdev) kinfo = &h->kinfo; for (i = 0; i < HNAE3_MAX_USER_PRIO; i++) - netdev_set_prio_tc_map(netdev, i, kinfo->prio_tc[i]); + netdev_set_prio_tc_map(netdev, i, kinfo->tc_info.prio_tc[i]); if (h->ae_algo->ops->set_timer_task) h->ae_algo->ops->set_timer_task(priv->ae_handle, true); @@ -3980,21 +3979,20 @@ static void hns3_init_ring_hw(struct hns3_enet_ring *ring) static void hns3_init_tx_ring_tc(struct hns3_nic_priv *priv) { struct hnae3_knic_private_info *kinfo = &priv->ae_handle->kinfo; + struct hnae3_tc_info *tc_info = &kinfo->tc_info; int i; for (i = 0; i < HNAE3_MAX_TC; i++) { - struct hnae3_tc_info *tc_info = &kinfo->tc_info[i]; int j; - if (!tc_info->enable) + if (!test_bit(i, &tc_info->tc_en)) continue; - for (j = 0; j < tc_info->tqp_count; j++) { + for (j = 0; j < tc_info->tqp_count[i]; j++) { struct hnae3_queue *q; - q = priv->ring[tc_info->tqp_offset + j].tqp; - hns3_write_dev(q, HNS3_RING_TX_RING_TC_REG, - tc_info->tc); + q = priv->ring[tc_info->tqp_offset[i] + j].tqp; + hns3_write_dev(q, HNS3_RING_TX_RING_TC_REG, i); } } } @@ -4121,7 +4119,8 @@ static void hns3_info_show(struct hns3_nic_priv *priv) dev_info(priv->dev, "RX buffer length: %u\n", kinfo->rx_buf_len); dev_info(priv->dev, "Desc num per TX queue: %u\n", kinfo->num_tx_desc); dev_info(priv->dev, "Desc num per RX queue: %u\n", kinfo->num_rx_desc); - dev_info(priv->dev, "Total number of enabled TCs: %u\n", kinfo->num_tc); + dev_info(priv->dev, "Total number of enabled TCs: %u\n", + kinfo->tc_info.num_tc); dev_info(priv->dev, "Max mtu size: %u\n", priv->netdev->max_mtu); } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c index bedbc118c4a35..8f6dea5198cf5 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c @@ -1454,7 +1454,7 @@ static void hclge_dbg_dump_qs_shaper_all(struct hclge_dev *hdev) dev_info(&hdev->pdev->dev, "qs cfg of vport%d:\n", vport_id); - for (i = 0; i < kinfo->num_tc; i++) { + for (i = 0; i < kinfo->tc_info.num_tc; i++) { u16 qsid = vport->qs_offset + i; hclge_dbg_dump_qs_shaper_single(hdev, qsid); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 92cd4446a82eb..8bf1027612217 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -10699,7 +10699,7 @@ static u32 hclge_get_max_channels(struct hnae3_handle *handle) struct hclge_dev *hdev = vport->back; return min_t(u32, hdev->rss_size_max, - vport->alloc_tqps / kinfo->num_tc); + vport->alloc_tqps / kinfo->tc_info.num_tc); } static void hclge_get_channels(struct hnae3_handle *handle, @@ -10786,7 +10786,7 @@ static int hclge_set_channels(struct hnae3_handle *handle, u32 new_tqps_num, dev_info(&hdev->pdev->dev, "Channels changed, rss_size from %u to %u, tqps from %u to %u", cur_rss_size, kinfo->rss_size, - cur_tqps, kinfo->rss_size * kinfo->num_tc); + cur_tqps, kinfo->rss_size * kinfo->tc_info.num_tc); return ret; } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c index 0f6626b35959b..754c09ada9019 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c @@ -414,7 +414,7 @@ static void hclge_get_vf_tcinfo(struct hclge_vport *vport, struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; unsigned int i; - for (i = 0; i < kinfo->num_tc; i++) + for (i = 0; i < kinfo->tc_info.num_tc; i++) resp_msg->data[0] |= BIT(i); resp_msg->len = sizeof(u8); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c index b1026cd1ba0af..dd2b1000c9404 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c @@ -565,7 +565,7 @@ int hclge_tm_qs_shaper_cfg(struct hclge_vport *vport, int max_tx_rate) HCLGE_SHAPER_BS_U_DEF, HCLGE_SHAPER_BS_S_DEF); - for (i = 0; i < kinfo->num_tc; i++) { + for (i = 0; i < kinfo->tc_info.num_tc; i++) { hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_QCN_SHAPPING_CFG, false); @@ -599,13 +599,13 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) /* TC configuration is shared by PF/VF in one port, only allow * one tc for VF for simplicity. VF's vport_id is non zero. */ - kinfo->num_tc = vport->vport_id ? 1 : + kinfo->tc_info.num_tc = vport->vport_id ? 1 : min_t(u16, vport->alloc_tqps, hdev->tm_info.num_tc); vport->qs_offset = (vport->vport_id ? HNAE3_MAX_TC : 0) + (vport->vport_id ? (vport->vport_id - 1) : 0); max_rss_size = min_t(u16, hdev->rss_size_max, - vport->alloc_tqps / kinfo->num_tc); + vport->alloc_tqps / kinfo->tc_info.num_tc); /* Set to user value, no larger than max_rss_size. */ if (kinfo->req_rss_size != kinfo->rss_size && kinfo->req_rss_size && @@ -622,34 +622,32 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) if (!kinfo->req_rss_size) max_rss_size = min_t(u16, max_rss_size, (hdev->num_nic_msi - 1) / - kinfo->num_tc); + kinfo->tc_info.num_tc); /* Set to the maximum specification value (max_rss_size). */ kinfo->rss_size = max_rss_size; } - kinfo->num_tqps = kinfo->num_tc * kinfo->rss_size; + kinfo->num_tqps = kinfo->tc_info.num_tc * kinfo->rss_size; vport->dwrr = 100; /* 100 percent as init */ vport->alloc_rss_size = kinfo->rss_size; vport->bw_limit = hdev->tm_info.pg_info[0].bw_limit; for (i = 0; i < HNAE3_MAX_TC; i++) { - if (hdev->hw_tc_map & BIT(i) && i < kinfo->num_tc) { - kinfo->tc_info[i].enable = true; - kinfo->tc_info[i].tqp_offset = i * kinfo->rss_size; - kinfo->tc_info[i].tqp_count = kinfo->rss_size; - kinfo->tc_info[i].tc = i; + if (hdev->hw_tc_map & BIT(i) && i < kinfo->tc_info.num_tc) { + set_bit(i, &kinfo->tc_info.tc_en); + kinfo->tc_info.tqp_offset[i] = i * kinfo->rss_size; + kinfo->tc_info.tqp_count[i] = kinfo->rss_size; } else { /* Set to default queue if TC is disable */ - kinfo->tc_info[i].enable = false; - kinfo->tc_info[i].tqp_offset = 0; - kinfo->tc_info[i].tqp_count = 1; - kinfo->tc_info[i].tc = 0; + clear_bit(i, &kinfo->tc_info.tc_en); + kinfo->tc_info.tqp_offset[i] = 0; + kinfo->tc_info.tqp_count[i] = 1; } } - memcpy(kinfo->prio_tc, hdev->tm_info.prio_tc, - sizeof_field(struct hnae3_knic_private_info, prio_tc)); + memcpy(kinfo->tc_info.prio_tc, hdev->tm_info.prio_tc, + sizeof_field(struct hnae3_tc_info, prio_tc)); } static void hclge_tm_vport_info_update(struct hclge_dev *hdev) @@ -854,15 +852,14 @@ static int hclge_vport_q_to_qs_map(struct hclge_dev *hdev, struct hclge_vport *vport) { struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; + struct hnae3_tc_info *tc_info = &kinfo->tc_info; struct hnae3_queue **tqp = kinfo->tqp; - struct hnae3_tc_info *v_tc_info; u32 i, j; int ret; - for (i = 0; i < kinfo->num_tc; i++) { - v_tc_info = &kinfo->tc_info[i]; - for (j = 0; j < v_tc_info->tqp_count; j++) { - struct hnae3_queue *q = tqp[v_tc_info->tqp_offset + j]; + for (i = 0; i < tc_info->num_tc; i++) { + for (j = 0; j < tc_info->tqp_count[i]; j++) { + struct hnae3_queue *q = tqp[tc_info->tqp_offset[i] + j]; ret = hclge_tm_q_to_qs_map_cfg(hdev, hclge_get_queue_id(q), @@ -887,7 +884,7 @@ static int hclge_tm_pri_q_qs_cfg(struct hclge_dev *hdev) struct hnae3_knic_private_info *kinfo = &vport[k].nic.kinfo; - for (i = 0; i < kinfo->num_tc; i++) { + for (i = 0; i < kinfo->tc_info.num_tc; i++) { ret = hclge_tm_qs_to_pri_map_cfg( hdev, vport[k].qs_offset + i, i); if (ret) @@ -1001,7 +998,7 @@ static int hclge_tm_pri_vnet_base_shaper_qs_cfg(struct hclge_vport *vport) u32 i; int ret; - for (i = 0; i < kinfo->num_tc; i++) { + for (i = 0; i < kinfo->tc_info.num_tc; i++) { ret = hclge_shaper_para_calc(hdev->tm_info.tc_info[i].bw_limit, HCLGE_SHAPER_LVL_QSET, &ir_para, max_tm_rate); @@ -1123,7 +1120,7 @@ static int hclge_tm_pri_vnet_base_dwrr_pri_cfg(struct hclge_vport *vport) return ret; /* Qset dwrr */ - for (i = 0; i < kinfo->num_tc; i++) { + for (i = 0; i < kinfo->tc_info.num_tc; i++) { ret = hclge_tm_qs_weight_cfg( hdev, vport->qs_offset + i, hdev->tm_info.pg_info[0].tc_dwrr[i]); @@ -1254,7 +1251,7 @@ static int hclge_tm_schd_mode_vnet_base_cfg(struct hclge_vport *vport) if (ret) return ret; - for (i = 0; i < kinfo->num_tc; i++) { + for (i = 0; i < kinfo->tc_info.num_tc; i++) { u8 sch_mode = hdev->tm_info.tc_info[i].tc_sch_mode; ret = hclge_tm_qs_schd_mode_cfg(hdev, vport->qs_offset + i, @@ -1484,7 +1481,7 @@ void hclge_tm_prio_tc_info_update(struct hclge_dev *hdev, u8 *prio_tc) for (k = 0; k < hdev->num_alloc_vport; k++) { kinfo = &vport[k].nic.kinfo; - kinfo->prio_tc[i] = prio_tc[i]; + kinfo->tc_info.prio_tc[i] = prio_tc[i]; } } } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c index b142ae6a8eecc..145757cb70f9f 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c @@ -433,19 +433,20 @@ static int hclgevf_knic_setup(struct hclgevf_dev *hdev) struct hnae3_knic_private_info *kinfo; u16 new_tqps = hdev->num_tqps; unsigned int i; + u8 num_tc = 0; kinfo = &nic->kinfo; - kinfo->num_tc = 0; kinfo->num_tx_desc = hdev->num_tx_desc; kinfo->num_rx_desc = hdev->num_rx_desc; kinfo->rx_buf_len = hdev->rx_buf_len; for (i = 0; i < HCLGEVF_MAX_TC_NUM; i++) if (hdev->hw_tc_map & BIT(i)) - kinfo->num_tc++; + num_tc++; - kinfo->rss_size - = min_t(u16, hdev->rss_size_max, new_tqps / kinfo->num_tc); - new_tqps = kinfo->rss_size * kinfo->num_tc; + num_tc = num_tc ? num_tc : 1; + kinfo->tc_info.num_tc = num_tc; + kinfo->rss_size = min_t(u16, hdev->rss_size_max, new_tqps / num_tc); + new_tqps = kinfo->rss_size * num_tc; kinfo->num_tqps = min(new_tqps, hdev->num_tqps); kinfo->tqp = devm_kcalloc(&hdev->pdev->dev, kinfo->num_tqps, @@ -463,7 +464,7 @@ static int hclgevf_knic_setup(struct hclgevf_dev *hdev) * and rss size with the actual vector numbers */ kinfo->num_tqps = min_t(u16, hdev->num_nic_msix - 1, kinfo->num_tqps); - kinfo->rss_size = min_t(u16, kinfo->num_tqps / kinfo->num_tc, + kinfo->rss_size = min_t(u16, kinfo->num_tqps / num_tc, kinfo->rss_size); return 0; @@ -3360,7 +3361,7 @@ static u32 hclgevf_get_max_channels(struct hclgevf_dev *hdev) struct hnae3_knic_private_info *kinfo = &nic->kinfo; return min_t(u32, hdev->rss_size_max, - hdev->num_tqps / kinfo->num_tc); + hdev->num_tqps / kinfo->tc_info.num_tc); } /** @@ -3403,7 +3404,7 @@ static void hclgevf_update_rss_size(struct hnae3_handle *handle, kinfo->req_rss_size = new_tqps_num; max_rss_size = min_t(u16, hdev->rss_size_max, - hdev->num_tqps / kinfo->num_tc); + hdev->num_tqps / kinfo->tc_info.num_tc); /* Use the user's configuration when it is not larger than * max_rss_size, otherwise, use the maximum specification value. @@ -3415,7 +3416,7 @@ static void hclgevf_update_rss_size(struct hnae3_handle *handle, (!kinfo->req_rss_size && kinfo->rss_size < max_rss_size)) kinfo->rss_size = max_rss_size; - kinfo->num_tqps = kinfo->num_tc * kinfo->rss_size; + kinfo->num_tqps = kinfo->tc_info.num_tc * kinfo->rss_size; } static int hclgevf_set_channels(struct hnae3_handle *handle, u32 new_tqps_num, @@ -3461,7 +3462,7 @@ static int hclgevf_set_channels(struct hnae3_handle *handle, u32 new_tqps_num, dev_info(&hdev->pdev->dev, "Channels changed, rss_size from %u to %u, tqps from %u to %u", cur_rss_size, kinfo->rss_size, - cur_tqps, kinfo->rss_size * kinfo->num_tc); + cur_tqps, kinfo->rss_size * kinfo->tc_info.num_tc); return ret; } From 5a5c9091746749d9e16ff807035a940ca0ae4941 Mon Sep 17 00:00:00 2001 From: Jian Shen Date: Thu, 10 Dec 2020 11:42:07 +0800 Subject: [PATCH 2/7] net: hns3: add support for tc mqprio offload Currently, the HNS3 driver only supports offload for tc number and prio_tc. This patch adds support for other qopts, including queues count and offset for each tc. When enable tc mqprio offload, it's not allowed to change queue numbers by ethtool. For hardware limitation, the queue number of each tc should be power of 2. For the queues is not assigned to each tc by average, so it's should return vport->alloc_tqps for hclge_get_max_channels(). Signed-off-by: Jian Shen Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 6 +- .../net/ethernet/hisilicon/hns3/hns3_enet.c | 13 +- .../hisilicon/hns3/hns3pf/hclge_dcb.c | 126 ++++++++++++++++-- .../hisilicon/hns3/hns3pf/hclge_main.c | 80 ++++++----- .../ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 48 ++++++- 5 files changed, 220 insertions(+), 53 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 0cb80ef4aba5c..a7ff9c7dbd2af 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -29,7 +29,9 @@ #include #include #include +#include #include +#include #define HNAE3_MOD_VERSION "1.0" @@ -647,7 +649,8 @@ struct hnae3_dcb_ops { u8 (*getdcbx)(struct hnae3_handle *); u8 (*setdcbx)(struct hnae3_handle *, u8); - int (*setup_tc)(struct hnae3_handle *, u8, u8 *); + int (*setup_tc)(struct hnae3_handle *handle, + struct tc_mqprio_qopt_offload *mqprio_qopt); }; struct hnae3_ae_algo { @@ -667,6 +670,7 @@ struct hnae3_tc_info { u16 tqp_offset[HNAE3_MAX_TC]; unsigned long tc_en; /* bitmap of TC enabled */ u8 num_tc; /* Total number of enabled TCs */ + bool mqprio_active; }; struct hnae3_knic_private_info { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index 36e74ad5b612f..d6dd4bc2fba1f 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -324,10 +324,10 @@ static int hns3_nic_set_real_num_queue(struct net_device *netdev) struct hnae3_handle *h = hns3_get_handle(netdev); struct hnae3_knic_private_info *kinfo = &h->kinfo; struct hnae3_tc_info *tc_info = &kinfo->tc_info; - unsigned int queue_size = kinfo->rss_size * tc_info->num_tc; + unsigned int queue_size = kinfo->num_tqps; int i, ret; - if (tc_info->num_tc <= 1) { + if (tc_info->num_tc <= 1 && !tc_info->mqprio_active) { netdev_reset_tc(netdev); } else { ret = netdev_set_num_tc(netdev, tc_info->num_tc); @@ -1793,7 +1793,6 @@ static void hns3_nic_get_stats64(struct net_device *netdev, static int hns3_setup_tc(struct net_device *netdev, void *type_data) { struct tc_mqprio_qopt_offload *mqprio_qopt = type_data; - u8 *prio_tc = mqprio_qopt->qopt.prio_tc_map; struct hnae3_knic_private_info *kinfo; u8 tc = mqprio_qopt->qopt.num_tc; u16 mode = mqprio_qopt->mode; @@ -1816,7 +1815,7 @@ static int hns3_setup_tc(struct net_device *netdev, void *type_data) netif_dbg(h, drv, netdev, "setup tc: num_tc=%u\n", tc); return (kinfo->dcb_ops && kinfo->dcb_ops->setup_tc) ? - kinfo->dcb_ops->setup_tc(h, tc ? tc : 1, prio_tc) : -EOPNOTSUPP; + kinfo->dcb_ops->setup_tc(h, mqprio_qopt) : -EOPNOTSUPP; } static int hns3_nic_setup_tc(struct net_device *dev, enum tc_setup_type type, @@ -4691,6 +4690,12 @@ int hns3_set_channels(struct net_device *netdev, if (ch->rx_count || ch->tx_count) return -EINVAL; + if (kinfo->tc_info.mqprio_active) { + dev_err(&netdev->dev, + "it's not allowed to set channels via ethtool when MQPRIO mode is on\n"); + return -EINVAL; + } + if (new_tqp_num > hns3_get_max_available_channels(h) || new_tqp_num < 1) { dev_err(&netdev->dev, diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c index f990f6915226f..a7f4c6a85d852 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c @@ -397,32 +397,130 @@ static u8 hclge_setdcbx(struct hnae3_handle *h, u8 mode) return 0; } +static int hclge_mqprio_qopt_check(struct hclge_dev *hdev, + struct tc_mqprio_qopt_offload *mqprio_qopt) +{ + u16 queue_sum = 0; + int ret; + int i; + + if (!mqprio_qopt->qopt.num_tc) { + mqprio_qopt->qopt.num_tc = 1; + return 0; + } + + ret = hclge_dcb_common_validate(hdev, mqprio_qopt->qopt.num_tc, + mqprio_qopt->qopt.prio_tc_map); + if (ret) + return ret; + + for (i = 0; i < mqprio_qopt->qopt.num_tc; i++) { + if (!is_power_of_2(mqprio_qopt->qopt.count[i])) { + dev_err(&hdev->pdev->dev, + "qopt queue count must be power of 2\n"); + return -EINVAL; + } + + if (mqprio_qopt->qopt.count[i] > hdev->rss_size_max) { + dev_err(&hdev->pdev->dev, + "qopt queue count should be no more than %u\n", + hdev->rss_size_max); + return -EINVAL; + } + + if (mqprio_qopt->qopt.offset[i] != queue_sum) { + dev_err(&hdev->pdev->dev, + "qopt queue offset must start from 0, and being continuous\n"); + return -EINVAL; + } + + if (mqprio_qopt->min_rate[i] || mqprio_qopt->max_rate[i]) { + dev_err(&hdev->pdev->dev, + "qopt tx_rate is not supported\n"); + return -EOPNOTSUPP; + } + + queue_sum = mqprio_qopt->qopt.offset[i]; + queue_sum += mqprio_qopt->qopt.count[i]; + } + if (hdev->vport[0].alloc_tqps < queue_sum) { + dev_err(&hdev->pdev->dev, + "qopt queue count sum should be less than %u\n", + hdev->vport[0].alloc_tqps); + return -EINVAL; + } + + return 0; +} + +static void hclge_sync_mqprio_qopt(struct hnae3_tc_info *tc_info, + struct tc_mqprio_qopt_offload *mqprio_qopt) +{ + int i; + + memset(tc_info, 0, sizeof(*tc_info)); + tc_info->num_tc = mqprio_qopt->qopt.num_tc; + memcpy(tc_info->prio_tc, mqprio_qopt->qopt.prio_tc_map, + sizeof_field(struct hnae3_tc_info, prio_tc)); + memcpy(tc_info->tqp_count, mqprio_qopt->qopt.count, + sizeof_field(struct hnae3_tc_info, tqp_count)); + memcpy(tc_info->tqp_offset, mqprio_qopt->qopt.offset, + sizeof_field(struct hnae3_tc_info, tqp_offset)); + + for (i = 0; i < HNAE3_MAX_USER_PRIO; i++) + set_bit(tc_info->prio_tc[i], &tc_info->tc_en); +} + +static int hclge_config_tc(struct hclge_dev *hdev, + struct hnae3_tc_info *tc_info) +{ + int i; + + hclge_tm_schd_info_update(hdev, tc_info->num_tc); + for (i = 0; i < HNAE3_MAX_USER_PRIO; i++) + hdev->tm_info.prio_tc[i] = tc_info->prio_tc[i]; + + return hclge_map_update(hdev); +} + /* Set up TC for hardware offloaded mqprio in channel mode */ -static int hclge_setup_tc(struct hnae3_handle *h, u8 tc, u8 *prio_tc) +static int hclge_setup_tc(struct hnae3_handle *h, + struct tc_mqprio_qopt_offload *mqprio_qopt) { struct hclge_vport *vport = hclge_get_vport(h); + struct hnae3_knic_private_info *kinfo; struct hclge_dev *hdev = vport->back; + struct hnae3_tc_info old_tc_info; + u8 tc = mqprio_qopt->qopt.num_tc; int ret; + /* if client unregistered, it's not allowed to change + * mqprio configuration, which may cause uninit ring + * fail. + */ + if (!test_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state)) + return -EBUSY; + if (hdev->flag & HCLGE_FLAG_DCB_ENABLE) return -EINVAL; - ret = hclge_dcb_common_validate(hdev, tc, prio_tc); - if (ret) - return -EINVAL; + ret = hclge_mqprio_qopt_check(hdev, mqprio_qopt); + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to check mqprio qopt params, ret = %d\n", ret); + return ret; + } ret = hclge_notify_down_uinit(hdev); if (ret) return ret; - hclge_tm_schd_info_update(hdev, tc); - hclge_tm_prio_tc_info_update(hdev, prio_tc); - - ret = hclge_tm_init_hw(hdev, false); - if (ret) - goto err_out; + kinfo = &vport->nic.kinfo; + memcpy(&old_tc_info, &kinfo->tc_info, sizeof(old_tc_info)); + hclge_sync_mqprio_qopt(&kinfo->tc_info, mqprio_qopt); + kinfo->tc_info.mqprio_active = tc > 0; - ret = hclge_client_setup_tc(hdev); + ret = hclge_config_tc(hdev, &kinfo->tc_info); if (ret) goto err_out; @@ -436,6 +534,12 @@ static int hclge_setup_tc(struct hnae3_handle *h, u8 tc, u8 *prio_tc) return hclge_notify_init_up(hdev); err_out: + /* roll-back */ + memcpy(&kinfo->tc_info, &old_tc_info, sizeof(old_tc_info)); + if (hclge_config_tc(hdev, &kinfo->tc_info)) + dev_err(&hdev->pdev->dev, + "failed to roll back tc configuration\n"); + hclge_notify_init_up(hdev); return ret; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 8bf1027612217..366920bd5d21d 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -4604,18 +4604,55 @@ static int hclge_get_tc_size(struct hnae3_handle *handle) return hdev->rss_size_max; } -int hclge_rss_init_hw(struct hclge_dev *hdev) +static int hclge_init_rss_tc_mode(struct hclge_dev *hdev) { + struct hnae3_ae_dev *ae_dev = hdev->ae_dev; struct hclge_vport *vport = hdev->vport; - u8 *rss_indir = vport[0].rss_indirection_tbl; - u16 rss_size = vport[0].alloc_rss_size; u16 tc_offset[HCLGE_MAX_TC_NUM] = {0}; + u16 tc_valid[HCLGE_MAX_TC_NUM] = {0}; u16 tc_size[HCLGE_MAX_TC_NUM] = {0}; + struct hnae3_tc_info *tc_info; + u16 roundup_size; + u16 rss_size; + int i; + + tc_info = &vport->nic.kinfo.tc_info; + for (i = 0; i < HCLGE_MAX_TC_NUM; i++) { + rss_size = tc_info->tqp_count[i]; + tc_valid[i] = 0; + + if (!(hdev->hw_tc_map & BIT(i))) + continue; + + /* tc_size set to hardware is the log2 of roundup power of two + * of rss_size, the acutal queue size is limited by indirection + * table. + */ + if (rss_size > ae_dev->dev_specs.rss_ind_tbl_size || + rss_size == 0) { + dev_err(&hdev->pdev->dev, + "Configure rss tc size failed, invalid TC_SIZE = %u\n", + rss_size); + return -EINVAL; + } + + roundup_size = roundup_pow_of_two(rss_size); + roundup_size = ilog2(roundup_size); + + tc_valid[i] = 1; + tc_size[i] = roundup_size; + tc_offset[i] = tc_info->tqp_offset[i]; + } + + return hclge_set_rss_tc_mode(hdev, tc_valid, tc_size, tc_offset); +} + +int hclge_rss_init_hw(struct hclge_dev *hdev) +{ + struct hclge_vport *vport = hdev->vport; + u8 *rss_indir = vport[0].rss_indirection_tbl; u8 *key = vport[0].rss_hash_key; u8 hfunc = vport[0].rss_algo; - u16 tc_valid[HCLGE_MAX_TC_NUM]; - u16 roundup_size; - unsigned int i; int ret; ret = hclge_set_rss_indir_table(hdev, rss_indir); @@ -4630,32 +4667,7 @@ int hclge_rss_init_hw(struct hclge_dev *hdev) if (ret) return ret; - /* Each TC have the same queue size, and tc_size set to hardware is - * the log2 of roundup power of two of rss_size, the acutal queue - * size is limited by indirection table. - */ - if (rss_size > HCLGE_RSS_TC_SIZE_7 || rss_size == 0) { - dev_err(&hdev->pdev->dev, - "Configure rss tc size failed, invalid TC_SIZE = %u\n", - rss_size); - return -EINVAL; - } - - roundup_size = roundup_pow_of_two(rss_size); - roundup_size = ilog2(roundup_size); - - for (i = 0; i < HCLGE_MAX_TC_NUM; i++) { - tc_valid[i] = 0; - - if (!(hdev->hw_tc_map & BIT(i))) - continue; - - tc_valid[i] = 1; - tc_size[i] = roundup_size; - tc_offset[i] = rss_size * i; - } - - return hclge_set_rss_tc_mode(hdev, tc_valid, tc_size, tc_offset); + return hclge_init_rss_tc_mode(hdev); } void hclge_rss_indir_init_cfg(struct hclge_dev *hdev) @@ -10694,12 +10706,10 @@ static void hclge_uninit_ae_dev(struct hnae3_ae_dev *ae_dev) static u32 hclge_get_max_channels(struct hnae3_handle *handle) { - struct hnae3_knic_private_info *kinfo = &handle->kinfo; struct hclge_vport *vport = hclge_get_vport(handle); struct hclge_dev *hdev = vport->back; - return min_t(u32, hdev->rss_size_max, - vport->alloc_tqps / kinfo->tc_info.num_tc); + return min_t(u32, hdev->rss_size_max, vport->alloc_tqps); } static void hclge_get_channels(struct hnae3_handle *handle, diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c index dd2b1000c9404..c3dcf94b2f6bf 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c @@ -589,6 +589,46 @@ int hclge_tm_qs_shaper_cfg(struct hclge_vport *vport, int max_tx_rate) return 0; } +static u16 hclge_vport_get_max_rss_size(struct hclge_vport *vport) +{ + struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; + struct hnae3_tc_info *tc_info = &kinfo->tc_info; + struct hclge_dev *hdev = vport->back; + u16 max_rss_size = 0; + int i; + + if (!tc_info->mqprio_active) + return vport->alloc_tqps / tc_info->num_tc; + + for (i = 0; i < HNAE3_MAX_TC; i++) { + if (!(hdev->hw_tc_map & BIT(i)) || i >= tc_info->num_tc) + continue; + if (max_rss_size < tc_info->tqp_count[i]) + max_rss_size = tc_info->tqp_count[i]; + } + + return max_rss_size; +} + +static u16 hclge_vport_get_tqp_num(struct hclge_vport *vport) +{ + struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; + struct hnae3_tc_info *tc_info = &kinfo->tc_info; + struct hclge_dev *hdev = vport->back; + int sum = 0; + int i; + + if (!tc_info->mqprio_active) + return kinfo->rss_size * tc_info->num_tc; + + for (i = 0; i < HNAE3_MAX_TC; i++) { + if (hdev->hw_tc_map & BIT(i) && i < tc_info->num_tc) + sum += tc_info->tqp_count[i]; + } + + return sum; +} + static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) { struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; @@ -605,7 +645,7 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) (vport->vport_id ? (vport->vport_id - 1) : 0); max_rss_size = min_t(u16, hdev->rss_size_max, - vport->alloc_tqps / kinfo->tc_info.num_tc); + hclge_vport_get_max_rss_size(vport)); /* Set to user value, no larger than max_rss_size. */ if (kinfo->req_rss_size != kinfo->rss_size && kinfo->req_rss_size && @@ -628,11 +668,15 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) kinfo->rss_size = max_rss_size; } - kinfo->num_tqps = kinfo->tc_info.num_tc * kinfo->rss_size; + kinfo->num_tqps = hclge_vport_get_tqp_num(vport); vport->dwrr = 100; /* 100 percent as init */ vport->alloc_rss_size = kinfo->rss_size; vport->bw_limit = hdev->tm_info.pg_info[0].bw_limit; + /* when enable mqprio, the tc_info has been updated. */ + if (kinfo->tc_info.mqprio_active) + return; + for (i = 0; i < HNAE3_MAX_TC; i++) { if (hdev->hw_tc_map & BIT(i) && i < kinfo->tc_info.num_tc) { set_bit(i, &kinfo->tc_info.tc_en); From 0f993fe2b89db2342833856cc0597a309f278e55 Mon Sep 17 00:00:00 2001 From: Jian Shen Date: Thu, 10 Dec 2020 11:42:08 +0800 Subject: [PATCH 3/7] net: hns3: add support for forwarding packet to queues of specified TC when flow director rule hit For some new device, it supports forwarding packet to queues of specified TC when flow director rule hit. So extend the command handle to support it. Signed-off-by: Jian Shen Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- .../hisilicon/hns3/hns3pf/hclge_cmd.c | 2 ++ .../hisilicon/hns3/hns3pf/hclge_cmd.h | 3 +++ .../hisilicon/hns3/hns3pf/hclge_main.c | 21 +++++++++++++++---- .../hisilicon/hns3/hns3pf/hclge_main.h | 6 +++++- 4 files changed, 27 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c index 85986c7d71fad..b728be4737f88 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c @@ -359,6 +359,8 @@ static void hclge_parse_capability(struct hclge_dev *hdev, set_bit(HNAE3_DEV_SUPPORT_HW_TX_CSUM_B, ae_dev->caps); if (hnae3_get_bit(caps, HCLGE_CAP_UDP_TUNNEL_CSUM_B)) set_bit(HNAE3_DEV_SUPPORT_UDP_TUNNEL_CSUM_B, ae_dev->caps); + if (hnae3_get_bit(caps, HCLGE_CAP_FD_FORWARD_TC_B)) + set_bit(HNAE3_DEV_SUPPORT_FD_FORWARD_TC_B, ae_dev->caps); } static enum hclge_cmd_status diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h index 52a6f9b112007..df5417d9698df 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h @@ -1051,6 +1051,9 @@ struct hclge_fd_tcam_config_3_cmd { #define HCLGE_FD_AD_WR_RULE_ID_B 0 #define HCLGE_FD_AD_RULE_ID_S 1 #define HCLGE_FD_AD_RULE_ID_M GENMASK(13, 1) +#define HCLGE_FD_AD_TC_OVRD_B 16 +#define HCLGE_FD_AD_TC_SIZE_S 17 +#define HCLGE_FD_AD_TC_SIZE_M GENMASK(20, 17) struct hclge_fd_ad_config_cmd { u8 stage; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 366920bd5d21d..0b6102e228d27 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -5099,6 +5099,7 @@ static int hclge_fd_tcam_config(struct hclge_dev *hdev, u8 stage, bool sel_x, static int hclge_fd_ad_config(struct hclge_dev *hdev, u8 stage, int loc, struct hclge_fd_ad_data *action) { + struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); struct hclge_fd_ad_config_cmd *req; struct hclge_desc desc; u64 ad_data = 0; @@ -5114,6 +5115,12 @@ static int hclge_fd_ad_config(struct hclge_dev *hdev, u8 stage, int loc, action->write_rule_id_to_bd); hnae3_set_field(ad_data, HCLGE_FD_AD_RULE_ID_M, HCLGE_FD_AD_RULE_ID_S, action->rule_id); + if (test_bit(HNAE3_DEV_SUPPORT_FD_FORWARD_TC_B, ae_dev->caps)) { + hnae3_set_bit(ad_data, HCLGE_FD_AD_TC_OVRD_B, + action->override_tc); + hnae3_set_field(ad_data, HCLGE_FD_AD_TC_SIZE_M, + HCLGE_FD_AD_TC_SIZE_S, (u32)action->tc_size); + } ad_data <<= 32; hnae3_set_bit(ad_data, HCLGE_FD_AD_DROP_B, action->drop_packet); hnae3_set_bit(ad_data, HCLGE_FD_AD_DIRECT_QID_B, @@ -5357,16 +5364,22 @@ static int hclge_config_key(struct hclge_dev *hdev, u8 stage, static int hclge_config_action(struct hclge_dev *hdev, u8 stage, struct hclge_fd_rule *rule) { + struct hclge_vport *vport = hdev->vport; + struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; struct hclge_fd_ad_data ad_data; + memset(&ad_data, 0, sizeof(struct hclge_fd_ad_data)); ad_data.ad_id = rule->location; if (rule->action == HCLGE_FD_ACTION_DROP_PACKET) { ad_data.drop_packet = true; - ad_data.forward_to_direct_queue = false; - ad_data.queue_id = 0; + } else if (rule->action == HCLGE_FD_ACTION_SELECT_TC) { + ad_data.override_tc = true; + ad_data.queue_id = + kinfo->tc_info.tqp_offset[rule->tc]; + ad_data.tc_size = + ilog2(kinfo->tc_info.tqp_count[rule->tc]); } else { - ad_data.drop_packet = false; ad_data.forward_to_direct_queue = true; ad_data.queue_id = rule->queue_id; } @@ -5937,7 +5950,7 @@ static int hclge_add_fd_entry(struct hnae3_handle *handle, return -EINVAL; } - action = HCLGE_FD_ACTION_ACCEPT_PACKET; + action = HCLGE_FD_ACTION_SELECT_QUEUE; q_index = ring; } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h index b3c1301f7dbc6..a48106455f316 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h @@ -572,8 +572,9 @@ enum HCLGE_FD_PACKET_TYPE { }; enum HCLGE_FD_ACTION { - HCLGE_FD_ACTION_ACCEPT_PACKET, + HCLGE_FD_ACTION_SELECT_QUEUE, HCLGE_FD_ACTION_DROP_PACKET, + HCLGE_FD_ACTION_SELECT_TC, }; struct hclge_fd_key_cfg { @@ -619,6 +620,7 @@ struct hclge_fd_rule { u32 unused_tuple; u32 flow_type; u8 action; + u8 tc; u16 vf_id; u16 queue_id; u16 location; @@ -637,6 +639,8 @@ struct hclge_fd_ad_data { u8 write_rule_id_to_bd; u8 next_input_key; u16 rule_id; + u16 tc_size; + u8 override_tc; }; enum HCLGE_MAC_NODE_STATE { From 0205ec041ec61838f80f092e4232a8e5de4ee55f Mon Sep 17 00:00:00 2001 From: Jian Shen Date: Thu, 10 Dec 2020 11:42:09 +0800 Subject: [PATCH 4/7] net: hns3: add support for hw tc offload of tc flower Some new device supports forwarding packet to queues of specified TC when flow director rule hit. So add support to configure flow director rule by tc flower. To avoid rule conflict, add a new flow director mode HCLGE_FD_TC_FLOWER_ACTIVE, and only one mode can be active at the same time. Signed-off-by: Jian Shen Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 11 + .../net/ethernet/hisilicon/hns3/hns3_enet.c | 70 +++- .../hisilicon/hns3/hns3pf/hclge_main.c | 313 +++++++++++++++++- .../hisilicon/hns3/hns3pf/hclge_main.h | 16 +- 4 files changed, 397 insertions(+), 13 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index a7ff9c7dbd2af..a7daf6d4511ee 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -459,6 +459,12 @@ struct hnae3_ae_dev { * Configure the default MAC for specified VF * get_module_eeprom * Get the optical module eeprom info. + * add_cls_flower + * Add clsflower rule + * del_cls_flower + * Delete clsflower rule + * cls_flower_active + * Check if any cls flower rule exist */ struct hnae3_ae_ops { int (*init_ae_dev)(struct hnae3_ae_dev *ae_dev); @@ -636,6 +642,11 @@ struct hnae3_ae_ops { int (*get_module_eeprom)(struct hnae3_handle *handle, u32 offset, u32 len, u8 *data); bool (*get_cmdq_stat)(struct hnae3_handle *handle); + int (*add_cls_flower)(struct hnae3_handle *handle, + struct flow_cls_offload *cls_flower, int tc); + int (*del_cls_flower)(struct hnae3_handle *handle, + struct flow_cls_offload *cls_flower); + bool (*cls_flower_active)(struct hnae3_handle *handle); }; struct hnae3_dcb_ops { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index d6dd4bc2fba1f..405e490334178 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -1668,6 +1668,13 @@ static int hns3_nic_set_features(struct net_device *netdev, h->ae_algo->ops->enable_fd(h, enable); } + if ((netdev->features & NETIF_F_HW_TC) > (features & NETIF_F_HW_TC) && + h->ae_algo->ops->cls_flower_active(h)) { + netdev_err(netdev, + "there are offloaded TC filters active, cannot disable HW TC offload"); + return -EINVAL; + } + netdev->features = features; return 0; } @@ -1818,13 +1825,67 @@ static int hns3_setup_tc(struct net_device *netdev, void *type_data) kinfo->dcb_ops->setup_tc(h, mqprio_qopt) : -EOPNOTSUPP; } +static int hns3_setup_tc_cls_flower(struct hns3_nic_priv *priv, + struct flow_cls_offload *flow) +{ + int tc = tc_classid_to_hwtc(priv->netdev, flow->classid); + struct hnae3_handle *h = hns3_get_handle(priv->netdev); + + switch (flow->command) { + case FLOW_CLS_REPLACE: + if (h->ae_algo->ops->add_cls_flower) + return h->ae_algo->ops->add_cls_flower(h, flow, tc); + break; + case FLOW_CLS_DESTROY: + if (h->ae_algo->ops->del_cls_flower) + return h->ae_algo->ops->del_cls_flower(h, flow); + break; + default: + break; + } + + return -EOPNOTSUPP; +} + +static int hns3_setup_tc_block_cb(enum tc_setup_type type, void *type_data, + void *cb_priv) +{ + struct hns3_nic_priv *priv = cb_priv; + + if (!tc_cls_can_offload_and_chain0(priv->netdev, type_data)) + return -EOPNOTSUPP; + + switch (type) { + case TC_SETUP_CLSFLOWER: + return hns3_setup_tc_cls_flower(priv, type_data); + default: + return -EOPNOTSUPP; + } +} + +static LIST_HEAD(hns3_block_cb_list); + static int hns3_nic_setup_tc(struct net_device *dev, enum tc_setup_type type, void *type_data) { - if (type != TC_SETUP_QDISC_MQPRIO) + struct hns3_nic_priv *priv = netdev_priv(dev); + int ret; + + switch (type) { + case TC_SETUP_QDISC_MQPRIO: + ret = hns3_setup_tc(dev, type_data); + break; + case TC_SETUP_BLOCK: + ret = flow_block_cb_setup_simple(type_data, + &hns3_block_cb_list, + hns3_setup_tc_block_cb, + priv, priv, true); + break; + default: return -EOPNOTSUPP; + } - return hns3_setup_tc(dev, type_data); + return ret; } static int hns3_vlan_rx_add_vid(struct net_device *netdev, @@ -2421,6 +2482,11 @@ static void hns3_set_default_feature(struct net_device *netdev) netdev->vlan_features |= NETIF_F_GSO_UDP_TUNNEL_CSUM; netdev->hw_enc_features |= NETIF_F_GSO_UDP_TUNNEL_CSUM; } + + if (test_bit(HNAE3_DEV_SUPPORT_FD_FORWARD_TC_B, ae_dev->caps)) { + netdev->hw_features |= NETIF_F_HW_TC; + netdev->features |= NETIF_F_HW_TC; + } } static int hns3_alloc_buffer(struct hns3_enet_ring *ring, diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 0b6102e228d27..455ee6cd46563 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -5376,9 +5376,9 @@ static int hclge_config_action(struct hclge_dev *hdev, u8 stage, } else if (rule->action == HCLGE_FD_ACTION_SELECT_TC) { ad_data.override_tc = true; ad_data.queue_id = - kinfo->tc_info.tqp_offset[rule->tc]; + kinfo->tc_info.tqp_offset[rule->cls_flower.tc]; ad_data.tc_size = - ilog2(kinfo->tc_info.tqp_count[rule->tc]); + ilog2(kinfo->tc_info.tqp_count[rule->cls_flower.tc]); } else { ad_data.forward_to_direct_queue = true; ad_data.queue_id = rule->queue_id; @@ -5896,6 +5896,14 @@ static int hclge_fd_config_rule(struct hclge_dev *hdev, return ret; } +static bool hclge_is_cls_flower_active(struct hnae3_handle *handle) +{ + struct hclge_vport *vport = hclge_get_vport(handle); + struct hclge_dev *hdev = vport->back; + + return hdev->fd_active_type == HCLGE_FD_TC_FLOWER_ACTIVE; +} + static int hclge_add_fd_entry(struct hnae3_handle *handle, struct ethtool_rxnfc *cmd) { @@ -5920,6 +5928,12 @@ static int hclge_add_fd_entry(struct hnae3_handle *handle, return -EOPNOTSUPP; } + if (hclge_is_cls_flower_active(handle)) { + dev_err(&hdev->pdev->dev, + "please delete all exist cls flower rules first\n"); + return -EINVAL; + } + fs = (struct ethtool_rx_flow_spec *)&cmd->fs; ret = hclge_fd_check_spec(hdev, fs, &unused); @@ -6001,7 +6015,8 @@ static int hclge_del_fd_entry(struct hnae3_handle *handle, if (fs->location >= hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1]) return -EINVAL; - if (!hclge_fd_rule_exist(hdev, fs->location)) { + if (hclge_is_cls_flower_active(handle) || !hdev->hclge_fd_rule_num || + !hclge_fd_rule_exist(hdev, fs->location)) { dev_err(&hdev->pdev->dev, "Delete fail, rule %u is inexistent\n", fs->location); return -ENOENT; @@ -6101,7 +6116,7 @@ static int hclge_get_fd_rule_cnt(struct hnae3_handle *handle, struct hclge_vport *vport = hclge_get_vport(handle); struct hclge_dev *hdev = vport->back; - if (!hnae3_dev_fd_supported(hdev)) + if (!hnae3_dev_fd_supported(hdev) || hclge_is_cls_flower_active(handle)) return -EOPNOTSUPP; cmd->rule_cnt = hdev->hclge_fd_rule_num; @@ -6444,7 +6459,8 @@ static int hclge_add_fd_entry_by_arfs(struct hnae3_handle *handle, u16 queue_id, * arfs should not work */ spin_lock_bh(&hdev->fd_rule_lock); - if (hdev->fd_active_type == HCLGE_FD_EP_ACTIVE) { + if (hdev->fd_active_type != HCLGE_FD_ARFS_ACTIVE || + hdev->fd_active_type != HCLGE_FD_RULE_NONE) { spin_unlock_bh(&hdev->fd_rule_lock); return -EOPNOTSUPP; } @@ -6472,7 +6488,7 @@ static int hclge_add_fd_entry_by_arfs(struct hnae3_handle *handle, u16 queue_id, set_bit(bit_id, hdev->fd_bmap); rule->location = bit_id; - rule->flow_id = flow_id; + rule->arfs.flow_id = flow_id; rule->queue_id = queue_id; hclge_fd_build_arfs_rule(&new_tuples, rule); ret = hclge_fd_config_rule(hdev, rule); @@ -6516,7 +6532,7 @@ static void hclge_rfs_filter_expire(struct hclge_dev *hdev) } hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list, rule_node) { if (rps_may_expire_flow(handle->netdev, rule->queue_id, - rule->flow_id, rule->location)) { + rule->arfs.flow_id, rule->location)) { hlist_del_init(&rule->rule_node); hlist_add_head(&rule->rule_node, &del_list); hdev->hclge_fd_rule_num--; @@ -6545,6 +6561,286 @@ static void hclge_clear_arfs_rules(struct hnae3_handle *handle) #endif } +static void hclge_get_cls_key_basic(const struct flow_rule *flow, + struct hclge_fd_rule *rule) +{ + if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_BASIC)) { + struct flow_match_basic match; + u16 ethtype_key, ethtype_mask; + + flow_rule_match_basic(flow, &match); + ethtype_key = ntohs(match.key->n_proto); + ethtype_mask = ntohs(match.mask->n_proto); + + if (ethtype_key == ETH_P_ALL) { + ethtype_key = 0; + ethtype_mask = 0; + } + rule->tuples.ether_proto = ethtype_key; + rule->tuples_mask.ether_proto = ethtype_mask; + rule->tuples.ip_proto = match.key->ip_proto; + rule->tuples_mask.ip_proto = match.mask->ip_proto; + } else { + rule->unused_tuple |= BIT(INNER_IP_PROTO); + rule->unused_tuple |= BIT(INNER_ETH_TYPE); + } +} + +static void hclge_get_cls_key_mac(const struct flow_rule *flow, + struct hclge_fd_rule *rule) +{ + if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_ETH_ADDRS)) { + struct flow_match_eth_addrs match; + + flow_rule_match_eth_addrs(flow, &match); + ether_addr_copy(rule->tuples.dst_mac, match.key->dst); + ether_addr_copy(rule->tuples_mask.dst_mac, match.mask->dst); + ether_addr_copy(rule->tuples.src_mac, match.key->src); + ether_addr_copy(rule->tuples_mask.src_mac, match.mask->src); + } else { + rule->unused_tuple |= BIT(INNER_DST_MAC); + rule->unused_tuple |= BIT(INNER_SRC_MAC); + } +} + +static void hclge_get_cls_key_vlan(const struct flow_rule *flow, + struct hclge_fd_rule *rule) +{ + if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_VLAN)) { + struct flow_match_vlan match; + + flow_rule_match_vlan(flow, &match); + rule->tuples.vlan_tag1 = match.key->vlan_id | + (match.key->vlan_priority << VLAN_PRIO_SHIFT); + rule->tuples_mask.vlan_tag1 = match.mask->vlan_id | + (match.mask->vlan_priority << VLAN_PRIO_SHIFT); + } else { + rule->unused_tuple |= BIT(INNER_VLAN_TAG_FST); + } +} + +static void hclge_get_cls_key_ip(const struct flow_rule *flow, + struct hclge_fd_rule *rule) +{ + u16 addr_type = 0; + + if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_CONTROL)) { + struct flow_match_control match; + + flow_rule_match_control(flow, &match); + addr_type = match.key->addr_type; + } + + if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) { + struct flow_match_ipv4_addrs match; + + flow_rule_match_ipv4_addrs(flow, &match); + rule->tuples.src_ip[IPV4_INDEX] = be32_to_cpu(match.key->src); + rule->tuples_mask.src_ip[IPV4_INDEX] = + be32_to_cpu(match.mask->src); + rule->tuples.dst_ip[IPV4_INDEX] = be32_to_cpu(match.key->dst); + rule->tuples_mask.dst_ip[IPV4_INDEX] = + be32_to_cpu(match.mask->dst); + } else if (addr_type == FLOW_DISSECTOR_KEY_IPV6_ADDRS) { + struct flow_match_ipv6_addrs match; + + flow_rule_match_ipv6_addrs(flow, &match); + be32_to_cpu_array(rule->tuples.src_ip, match.key->src.s6_addr32, + IPV6_SIZE); + be32_to_cpu_array(rule->tuples_mask.src_ip, + match.mask->src.s6_addr32, IPV6_SIZE); + be32_to_cpu_array(rule->tuples.dst_ip, match.key->dst.s6_addr32, + IPV6_SIZE); + be32_to_cpu_array(rule->tuples_mask.dst_ip, + match.mask->dst.s6_addr32, IPV6_SIZE); + } else { + rule->unused_tuple |= BIT(INNER_SRC_IP); + rule->unused_tuple |= BIT(INNER_DST_IP); + } +} + +static void hclge_get_cls_key_port(const struct flow_rule *flow, + struct hclge_fd_rule *rule) +{ + if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_PORTS)) { + struct flow_match_ports match; + + flow_rule_match_ports(flow, &match); + + rule->tuples.src_port = be16_to_cpu(match.key->src); + rule->tuples_mask.src_port = be16_to_cpu(match.mask->src); + rule->tuples.dst_port = be16_to_cpu(match.key->dst); + rule->tuples_mask.dst_port = be16_to_cpu(match.mask->dst); + } else { + rule->unused_tuple |= BIT(INNER_SRC_PORT); + rule->unused_tuple |= BIT(INNER_DST_PORT); + } +} + +static int hclge_parse_cls_flower(struct hclge_dev *hdev, + struct flow_cls_offload *cls_flower, + struct hclge_fd_rule *rule) +{ + struct flow_rule *flow = flow_cls_offload_flow_rule(cls_flower); + struct flow_dissector *dissector = flow->match.dissector; + + if (dissector->used_keys & + ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) | + BIT(FLOW_DISSECTOR_KEY_BASIC) | + BIT(FLOW_DISSECTOR_KEY_ETH_ADDRS) | + BIT(FLOW_DISSECTOR_KEY_VLAN) | + BIT(FLOW_DISSECTOR_KEY_IPV4_ADDRS) | + BIT(FLOW_DISSECTOR_KEY_IPV6_ADDRS) | + BIT(FLOW_DISSECTOR_KEY_PORTS))) { + dev_err(&hdev->pdev->dev, "unsupported key set: %#x\n", + dissector->used_keys); + return -EOPNOTSUPP; + } + + hclge_get_cls_key_basic(flow, rule); + hclge_get_cls_key_mac(flow, rule); + hclge_get_cls_key_vlan(flow, rule); + hclge_get_cls_key_ip(flow, rule); + hclge_get_cls_key_port(flow, rule); + + return 0; +} + +static int hclge_check_cls_flower(struct hclge_dev *hdev, + struct flow_cls_offload *cls_flower, int tc) +{ + u32 prio = cls_flower->common.prio; + + if (tc < 0 || tc > hdev->tc_max) { + dev_err(&hdev->pdev->dev, "invalid traffic class\n"); + return -EINVAL; + } + + if (prio == 0 || + prio > hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1]) { + dev_err(&hdev->pdev->dev, + "prio %u should be in range[1, %u]\n", + prio, hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1]); + return -EINVAL; + } + + if (test_bit(prio - 1, hdev->fd_bmap)) { + dev_err(&hdev->pdev->dev, "prio %u is already used\n", prio); + return -EINVAL; + } + return 0; +} + +static int hclge_add_cls_flower(struct hnae3_handle *handle, + struct flow_cls_offload *cls_flower, + int tc) +{ + struct hclge_vport *vport = hclge_get_vport(handle); + struct hclge_dev *hdev = vport->back; + struct hclge_fd_rule *rule; + int ret; + + if (hdev->fd_active_type == HCLGE_FD_EP_ACTIVE) { + dev_err(&hdev->pdev->dev, + "please remove all exist fd rules via ethtool first\n"); + return -EINVAL; + } + + ret = hclge_check_cls_flower(hdev, cls_flower, tc); + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to check cls flower params, ret = %d\n", ret); + return ret; + } + + rule = kzalloc(sizeof(*rule), GFP_KERNEL); + if (!rule) + return -ENOMEM; + + ret = hclge_parse_cls_flower(hdev, cls_flower, rule); + if (ret) + goto err; + + rule->action = HCLGE_FD_ACTION_SELECT_TC; + rule->cls_flower.tc = tc; + rule->location = cls_flower->common.prio - 1; + rule->vf_id = 0; + rule->cls_flower.cookie = cls_flower->cookie; + rule->rule_type = HCLGE_FD_TC_FLOWER_ACTIVE; + + spin_lock_bh(&hdev->fd_rule_lock); + hclge_clear_arfs_rules(handle); + + ret = hclge_fd_config_rule(hdev, rule); + + spin_unlock_bh(&hdev->fd_rule_lock); + + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to add cls flower rule, ret = %d\n", ret); + goto err; + } + + return 0; +err: + kfree(rule); + return ret; +} + +static struct hclge_fd_rule *hclge_find_cls_flower(struct hclge_dev *hdev, + unsigned long cookie) +{ + struct hclge_fd_rule *rule; + struct hlist_node *node; + + hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list, rule_node) { + if (rule->cls_flower.cookie == cookie) + return rule; + } + + return NULL; +} + +static int hclge_del_cls_flower(struct hnae3_handle *handle, + struct flow_cls_offload *cls_flower) +{ + struct hclge_vport *vport = hclge_get_vport(handle); + struct hclge_dev *hdev = vport->back; + struct hclge_fd_rule *rule; + int ret; + + spin_lock_bh(&hdev->fd_rule_lock); + + rule = hclge_find_cls_flower(hdev, cls_flower->cookie); + if (!rule) { + spin_unlock_bh(&hdev->fd_rule_lock); + return -EINVAL; + } + + ret = hclge_fd_tcam_config(hdev, HCLGE_FD_STAGE_1, true, rule->location, + NULL, false); + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to delete cls flower rule %u, ret = %d\n", + rule->location, ret); + spin_unlock_bh(&hdev->fd_rule_lock); + return ret; + } + + ret = hclge_fd_update_rule_list(hdev, NULL, rule->location, false); + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to delete cls flower rule %u in list, ret = %d\n", + rule->location, ret); + spin_unlock_bh(&hdev->fd_rule_lock); + return ret; + } + + spin_unlock_bh(&hdev->fd_rule_lock); + + return 0; +} + static bool hclge_get_hw_reset_stat(struct hnae3_handle *handle) { struct hclge_vport *vport = hclge_get_vport(handle); @@ -11542,6 +11838,9 @@ static const struct hnae3_ae_ops hclge_ops = { .set_vf_mac = hclge_set_vf_mac, .get_module_eeprom = hclge_get_module_eeprom, .get_cmdq_stat = hclge_get_cmdq_stat, + .add_cls_flower = hclge_add_cls_flower, + .del_cls_flower = hclge_del_cls_flower, + .cls_flower_active = hclge_is_cls_flower_active, }; static struct hnae3_ae_algo ae_algo = { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h index a48106455f316..008005efe3134 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h @@ -564,6 +564,7 @@ enum HCLGE_FD_ACTIVE_RULE_TYPE { HCLGE_FD_RULE_NONE, HCLGE_FD_ARFS_ACTIVE, HCLGE_FD_EP_ACTIVE, + HCLGE_FD_TC_FLOWER_ACTIVE, }; enum HCLGE_FD_PACKET_TYPE { @@ -619,13 +620,20 @@ struct hclge_fd_rule { struct hclge_fd_rule_tuples tuples_mask; u32 unused_tuple; u32 flow_type; - u8 action; - u8 tc; - u16 vf_id; + union { + struct { + unsigned long cookie; + u8 tc; + } cls_flower; + struct { + u16 flow_id; /* only used for arfs */ + } arfs; + }; u16 queue_id; + u16 vf_id; u16 location; - u16 flow_id; /* only used for arfs */ enum HCLGE_FD_ACTIVE_RULE_TYPE rule_type; + u8 action; }; struct hclge_fd_ad_data { From f1c2e66d7f38cb1e919bd594af4ad4a9ae8344cc Mon Sep 17 00:00:00 2001 From: Guojia Liao Date: Thu, 10 Dec 2020 11:42:10 +0800 Subject: [PATCH 5/7] net: hns3: add support for max 512 rss size Currently, the driver gets the max rss size from configuration file when initialization. Both the PF and VF share the same max rss size, and no more than 128. For DEVICE_VERSION_V3, the max rss size for PF can be up to 512, so there is a new field in configuration file to store it, the old filed is used for VF. To be compatible with boards using old configure file, the PF will use the old filed if the one is zero. For the rss size may be larger than 256, so the type of rss_indirection_tbl of struct hclge_vport should be changed to u16 as well. Signed-off-by: Guojia Liao Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- .../hisilicon/hns3/hns3pf/hclge_cmd.h | 2 ++ .../hisilicon/hns3/hns3pf/hclge_dcb.c | 4 +-- .../hisilicon/hns3/hns3pf/hclge_main.c | 36 +++++++++++++------ .../hisilicon/hns3/hns3pf/hclge_main.h | 8 +++-- .../ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 5 ++- 5 files changed, 39 insertions(+), 16 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h index df5417d9698df..f5a620c3e66db 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h @@ -518,6 +518,8 @@ struct hclge_pf_res_cmd { #define HCLGE_CFG_SPEED_ABILITY_EXT_M GENMASK(15, 10) #define HCLGE_CFG_UMV_TBL_SPACE_S 16 #define HCLGE_CFG_UMV_TBL_SPACE_M GENMASK(31, 16) +#define HCLGE_CFG_PF_RSS_SIZE_S 0 +#define HCLGE_CFG_PF_RSS_SIZE_M GENMASK(3, 0) #define HCLGE_CFG_CMD_CNT 4 diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c index a7f4c6a85d852..e08d11b8ecf1e 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_dcb.c @@ -421,10 +421,10 @@ static int hclge_mqprio_qopt_check(struct hclge_dev *hdev, return -EINVAL; } - if (mqprio_qopt->qopt.count[i] > hdev->rss_size_max) { + if (mqprio_qopt->qopt.count[i] > hdev->pf_rss_size_max) { dev_err(&hdev->pdev->dev, "qopt queue count should be no more than %u\n", - hdev->rss_size_max); + hdev->pf_rss_size_max); return -EINVAL; } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 455ee6cd46563..f361226b35007 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -1285,9 +1285,9 @@ static void hclge_parse_cfg(struct hclge_cfg *cfg, struct hclge_desc *desc) cfg->default_speed = hnae3_get_field(__le32_to_cpu(req->param[3]), HCLGE_CFG_DEFAULT_SPEED_M, HCLGE_CFG_DEFAULT_SPEED_S); - cfg->rss_size_max = hnae3_get_field(__le32_to_cpu(req->param[3]), - HCLGE_CFG_RSS_SIZE_M, - HCLGE_CFG_RSS_SIZE_S); + cfg->vf_rss_size_max = hnae3_get_field(__le32_to_cpu(req->param[3]), + HCLGE_CFG_RSS_SIZE_M, + HCLGE_CFG_RSS_SIZE_S); for (i = 0; i < ETH_ALEN; i++) cfg->mac_addr[i] = (mac_addr_tmp >> (8 * i)) & 0xff; @@ -1308,6 +1308,21 @@ static void hclge_parse_cfg(struct hclge_cfg *cfg, struct hclge_desc *desc) HCLGE_CFG_UMV_TBL_SPACE_S); if (!cfg->umv_space) cfg->umv_space = HCLGE_DEFAULT_UMV_SPACE_PER_PF; + + cfg->pf_rss_size_max = hnae3_get_field(__le32_to_cpu(req->param[2]), + HCLGE_CFG_PF_RSS_SIZE_M, + HCLGE_CFG_PF_RSS_SIZE_S); + + /* HCLGE_CFG_PF_RSS_SIZE_M is the PF max rss size, which is a + * power of 2, instead of reading out directly. This would + * be more flexible for future changes and expansions. + * When VF max rss size field is HCLGE_CFG_RSS_SIZE_S, + * it does not make sense if PF's field is 0. In this case, PF and VF + * has the same max rss size filed: HCLGE_CFG_RSS_SIZE_S. + */ + cfg->pf_rss_size_max = cfg->pf_rss_size_max ? + 1U << cfg->pf_rss_size_max : + cfg->vf_rss_size_max; } /* hclge_get_cfg: query the static parameter from flash @@ -1469,7 +1484,8 @@ static int hclge_configure(struct hclge_dev *hdev) hdev->num_vmdq_vport = cfg.vmdq_vport_num; hdev->base_tqp_pid = 0; - hdev->rss_size_max = cfg.rss_size_max; + hdev->vf_rss_size_max = cfg.vf_rss_size_max; + hdev->pf_rss_size_max = cfg.pf_rss_size_max; hdev->rx_buf_len = cfg.rx_buf_len; ether_addr_copy(hdev->hw.mac.mac_addr, cfg.mac_addr); hdev->hw.mac.media_type = cfg.media_type; @@ -1652,7 +1668,7 @@ static int hclge_assign_tqp(struct hclge_vport *vport, u16 num_tqps) } } vport->alloc_tqps = alloced; - kinfo->rss_size = min_t(u16, hdev->rss_size_max, + kinfo->rss_size = min_t(u16, hdev->pf_rss_size_max, vport->alloc_tqps / hdev->tm_info.num_tc); /* ensure one to one mapping between irq and queue at default */ @@ -4262,7 +4278,7 @@ static int hclge_set_rss_algo_key(struct hclge_dev *hdev, return 0; } -static int hclge_set_rss_indir_table(struct hclge_dev *hdev, const u8 *indir) +static int hclge_set_rss_indir_table(struct hclge_dev *hdev, const u16 *indir) { struct hclge_rss_indirection_table_cmd *req; struct hclge_desc desc; @@ -4601,7 +4617,7 @@ static int hclge_get_tc_size(struct hnae3_handle *handle) struct hclge_vport *vport = hclge_get_vport(handle); struct hclge_dev *hdev = vport->back; - return hdev->rss_size_max; + return hdev->pf_rss_size_max; } static int hclge_init_rss_tc_mode(struct hclge_dev *hdev) @@ -4650,7 +4666,7 @@ static int hclge_init_rss_tc_mode(struct hclge_dev *hdev) int hclge_rss_init_hw(struct hclge_dev *hdev) { struct hclge_vport *vport = hdev->vport; - u8 *rss_indir = vport[0].rss_indirection_tbl; + u16 *rss_indir = vport[0].rss_indirection_tbl; u8 *key = vport[0].rss_hash_key; u8 hfunc = vport[0].rss_algo; int ret; @@ -11018,7 +11034,7 @@ static u32 hclge_get_max_channels(struct hnae3_handle *handle) struct hclge_vport *vport = hclge_get_vport(handle); struct hclge_dev *hdev = vport->back; - return min_t(u32, hdev->rss_size_max, vport->alloc_tqps); + return min_t(u32, hdev->pf_rss_size_max, vport->alloc_tqps); } static void hclge_get_channels(struct hnae3_handle *handle, @@ -11037,7 +11053,7 @@ static void hclge_get_tqps_and_rss_info(struct hnae3_handle *handle, struct hclge_dev *hdev = vport->back; *alloc_tqps = vport->alloc_tqps; - *max_rss_size = hdev->rss_size_max; + *max_rss_size = hdev->pf_rss_size_max; } static int hclge_set_channels(struct hnae3_handle *handle, u32 new_tqps_num, diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h index 008005efe3134..50a294dfaff50 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h @@ -348,7 +348,8 @@ struct hclge_cfg { u8 tc_num; u16 tqp_desc_num; u16 rx_buf_len; - u16 rss_size_max; + u16 vf_rss_size_max; + u16 pf_rss_size_max; u8 phy_addr; u8 media_type; u8 mac_addr[ETH_ALEN]; @@ -757,7 +758,8 @@ struct hclge_dev { u16 base_tqp_pid; /* Base task tqp physical id of this PF */ u16 alloc_rss_size; /* Allocated RSS task queue */ - u16 rss_size_max; /* HW defined max RSS task queue */ + u16 vf_rss_size_max; /* HW defined VF max RSS task queue */ + u16 pf_rss_size_max; /* HW defined PF max RSS task queue */ u16 fdir_pf_filter_count; /* Num of guaranteed filters for this PF */ u16 num_alloc_vport; /* Num vports this driver supports */ @@ -918,7 +920,7 @@ struct hclge_vport { u8 rss_hash_key[HCLGE_RSS_KEY_SIZE]; /* User configured hash keys */ /* User configured lookup table entries */ - u8 rss_indirection_tbl[HCLGE_RSS_IND_TBL_SIZE]; + u16 rss_indirection_tbl[HCLGE_RSS_IND_TBL_SIZE]; int rss_algo; /* User configured hash algorithm */ /* User configured rss tuple sets */ struct hclge_rss_tuple_cfg rss_tuple_sets; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c index c3dcf94b2f6bf..82742a64f3b70 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c @@ -633,6 +633,7 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) { struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; struct hclge_dev *hdev = vport->back; + u16 vport_max_rss_size; u16 max_rss_size; u8 i; @@ -644,7 +645,9 @@ static void hclge_tm_vport_tc_info_update(struct hclge_vport *vport) vport->qs_offset = (vport->vport_id ? HNAE3_MAX_TC : 0) + (vport->vport_id ? (vport->vport_id - 1) : 0); - max_rss_size = min_t(u16, hdev->rss_size_max, + vport_max_rss_size = vport->vport_id ? hdev->vf_rss_size_max : + hdev->pf_rss_size_max; + max_rss_size = min_t(u16, vport_max_rss_size, hclge_vport_get_max_rss_size(vport)); /* Set to user value, no larger than max_rss_size. */ From 8eeb1f4bce68e91dea52d1832b89fed3a984cb58 Mon Sep 17 00:00:00 2001 From: Guojia Liao Date: Thu, 10 Dec 2020 11:42:11 +0800 Subject: [PATCH 6/7] net: hns3: adjust rss indirection table configure command For the max rss size of PF may be up to 512, so adjust the command of configuring rss indirection table to support queue id larger than 255. The width of queue id is extended from 8 bits to 10 bits. The high 2 bits are stored in filed rss_qid_h when the queue id is larger than 255. Signed-off-by: Guojia Liao Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- .../hisilicon/hns3/hns3pf/hclge_cmd.h | 7 +++++-- .../hisilicon/hns3/hns3pf/hclge_main.c | 20 +++++++++++++------ 2 files changed, 19 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h index f5a620c3e66db..a6c306b3b9b3d 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h @@ -560,12 +560,15 @@ struct hclge_rss_input_tuple_cmd { }; #define HCLGE_RSS_CFG_TBL_SIZE 16 +#define HCLGE_RSS_CFG_TBL_SIZE_H 4 +#define HCLGE_RSS_CFG_TBL_BW_H 2U +#define HCLGE_RSS_CFG_TBL_BW_L 8U struct hclge_rss_indirection_table_cmd { __le16 start_table_index; __le16 rss_set_bitmap; - u8 rsv[4]; - u8 rss_result[HCLGE_RSS_CFG_TBL_SIZE]; + u8 rss_qid_h[HCLGE_RSS_CFG_TBL_SIZE_H]; + u8 rss_qid_l[HCLGE_RSS_CFG_TBL_SIZE]; }; #define HCLGE_RSS_TC_OFFSET_S 0 diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index f361226b35007..5de45a935e4de 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -4282,8 +4282,12 @@ static int hclge_set_rss_indir_table(struct hclge_dev *hdev, const u16 *indir) { struct hclge_rss_indirection_table_cmd *req; struct hclge_desc desc; - int i, j; + u8 rss_msb_oft; + u8 rss_msb_val; int ret; + u16 qid; + int i; + u32 j; req = (struct hclge_rss_indirection_table_cmd *)desc.data; @@ -4294,11 +4298,15 @@ static int hclge_set_rss_indir_table(struct hclge_dev *hdev, const u16 *indir) req->start_table_index = cpu_to_le16(i * HCLGE_RSS_CFG_TBL_SIZE); req->rss_set_bitmap = cpu_to_le16(HCLGE_RSS_SET_BITMAP_MSK); - - for (j = 0; j < HCLGE_RSS_CFG_TBL_SIZE; j++) - req->rss_result[j] = - indir[i * HCLGE_RSS_CFG_TBL_SIZE + j]; - + for (j = 0; j < HCLGE_RSS_CFG_TBL_SIZE; j++) { + qid = indir[i * HCLGE_RSS_CFG_TBL_SIZE + j]; + req->rss_qid_l[j] = qid & 0xff; + rss_msb_oft = + j * HCLGE_RSS_CFG_TBL_BW_H / BITS_PER_BYTE; + rss_msb_val = (qid >> HCLGE_RSS_CFG_TBL_BW_L & 0x1) << + (j * HCLGE_RSS_CFG_TBL_BW_H % BITS_PER_BYTE); + req->rss_qid_h[rss_msb_oft] |= rss_msb_val; + } ret = hclge_cmd_send(&hdev->hw, &desc, 1); if (ret) { dev_err(&hdev->pdev->dev, From cdab7c9779709755de90b9d017df2682f797e2bb Mon Sep 17 00:00:00 2001 From: Guojia Liao Date: Thu, 10 Dec 2020 11:42:12 +0800 Subject: [PATCH 7/7] net: hns3: adjust rss tc mode configure command For the max rss size of PF may be up to 512, the max queue number of single tc may be up to 512 too. For the total queue numbers may be up to 1280, so the queue offset of each tc may be more than 1024. So adjust the rss tc mode configuration command, including extend tc size field from 10 bits to 11 bits, and extend tc size field from 3 bits to 4 bits. Signed-off-by: Guojia Liao Signed-off-by: Huazhong Tan Signed-off-by: David S. Miller --- drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h | 4 +++- drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 2 ++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h index a6c306b3b9b3d..edfadb5cb1c34 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h @@ -572,9 +572,11 @@ struct hclge_rss_indirection_table_cmd { }; #define HCLGE_RSS_TC_OFFSET_S 0 -#define HCLGE_RSS_TC_OFFSET_M GENMASK(9, 0) +#define HCLGE_RSS_TC_OFFSET_M GENMASK(10, 0) +#define HCLGE_RSS_TC_SIZE_MSB_B 11 #define HCLGE_RSS_TC_SIZE_S 12 #define HCLGE_RSS_TC_SIZE_M GENMASK(14, 12) +#define HCLGE_RSS_TC_SIZE_MSB_OFFSET 3 #define HCLGE_RSS_TC_VALID_B 15 struct hclge_rss_tc_mode_cmd { __le16 rss_tc_mode[HCLGE_MAX_TC_NUM]; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 5de45a935e4de..7a164115c845f 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -4335,6 +4335,8 @@ static int hclge_set_rss_tc_mode(struct hclge_dev *hdev, u16 *tc_valid, hnae3_set_bit(mode, HCLGE_RSS_TC_VALID_B, (tc_valid[i] & 0x1)); hnae3_set_field(mode, HCLGE_RSS_TC_SIZE_M, HCLGE_RSS_TC_SIZE_S, tc_size[i]); + hnae3_set_bit(mode, HCLGE_RSS_TC_SIZE_MSB_B, + tc_size[i] >> HCLGE_RSS_TC_SIZE_MSB_OFFSET & 0x1); hnae3_set_field(mode, HCLGE_RSS_TC_OFFSET_M, HCLGE_RSS_TC_OFFSET_S, tc_offset[i]);