Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752652AbdHBQCU (ORCPT ); Wed, 2 Aug 2017 12:02:20 -0400 Received: from szxga01-in.huawei.com ([45.249.212.187]:10751 "EHLO szxga01-in.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752286AbdHBQBF (ORCPT ); Wed, 2 Aug 2017 12:01:05 -0400 From: Salil Mehta To: CC: , , , , , , , , , Wei Hu Subject: [PATCH V6 net-next 5/8] net: hns3: Add support of TX Scheduler & Shaper to HNS3 driver Date: Wed, 2 Aug 2017 16:59:49 +0100 Message-ID: <20170802155952.185564-6-salil.mehta@huawei.com> X-Mailer: git-send-email 2.8.3 In-Reply-To: <20170802155952.185564-1-salil.mehta@huawei.com> References: <20170802155952.185564-1-salil.mehta@huawei.com> MIME-Version: 1.0 Content-Type: text/plain X-Originating-IP: [10.203.181.161] X-CFilter-Loop: Reflected X-Mirapoint-Virus-RAPID-Raw: score=unknown(0), refid=str=0001.0A0B0204.5981F73D.00D0,ss=1,re=0.000,recu=0.000,reip=0.000,cl=1,cld=1,fgs=0, ip=0.0.0.0, so=2014-11-16 11:51:01, dmn=2013-03-21 17:37:32 X-Mirapoint-Loop-Id: 3538dee0bc6a53554314dcad4ac2fa4e Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 28114 Lines: 1168 THis patch adds the support of the Scheduling and Shaping functionalities during the transmit leg. This also adds the support of Pause at MAC level. (Pause at per-priority level shall be added later along with the DCB feature). Hardware as such consists of two types of cofiguration of 6 level schedulers. Algorithms varies according to the level and type of scheduler being used. Current patch is used to initialize the mapping, algorithms(like SP, DWRR etc) and shaper(CIR, PIR etc) being used. Signed-off-by: Daode Huang Signed-off-by: lipeng Signed-off-by: Salil Mehta Signed-off-by: Yisen Zhuang Signed-off-by: Wei Hu (Xavier) --- Patch V6: Addressed some internal comments 1. Fixed the return values from ENOMEM -> EINVAL 2. renamed to __HCLGE_TM_H macro in header file Patch V5: Addressed comments by 1. Richard Cochran https://lkml.org/lkml/2017/7/23/13 Patch V2,V4: No Change Patch V1: Initial Submit --- .../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c | 1015 ++++++++++++++++++++ .../net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h | 106 ++ 2 files changed, 1121 insertions(+) create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c new file mode 100644 index 0000000..1c577d2 --- /dev/null +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.c @@ -0,0 +1,1015 @@ +/* + * Copyright (c) 2016~2017 Hisilicon Limited. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include + +#include "hclge_cmd.h" +#include "hclge_main.h" +#include "hclge_tm.h" + +enum hclge_shaper_level { + HCLGE_SHAPER_LVL_PRI = 0, + HCLGE_SHAPER_LVL_PG = 1, + HCLGE_SHAPER_LVL_PORT = 2, + HCLGE_SHAPER_LVL_QSET = 3, + HCLGE_SHAPER_LVL_CNT = 4, + HCLGE_SHAPER_LVL_VF = 0, + HCLGE_SHAPER_LVL_PF = 1, +}; + +#define HCLGE_SHAPER_BS_U_DEF 1 +#define HCLGE_SHAPER_BS_S_DEF 4 + +#define HCLGE_ETHER_MAX_RATE 100000 + +/* hclge_shaper_para_calc: calculate ir parameter for the shaper + * @ir: Rate to be config, its unit is Mbps + * @shaper_level: the shaper level. eg: port, pg, priority, queueset + * @ir_b: IR_B parameter of IR shaper + * @ir_u: IR_U parameter of IR shaper + * @ir_s: IR_S parameter of IR shaper + * + * the formula: + * + * IR_b * (2 ^ IR_u) * 8 + * IR(Mbps) = ------------------------- * CLOCK(1000Mbps) + * Tick * (2 ^ IR_s) + * + * @return: 0: calculate sucessful, negative: fail + */ +static int hclge_shaper_para_calc(u32 ir, u8 shaper_level, + u8 *ir_b, u8 *ir_u, u8 *ir_s) +{ + const u16 tick_array[HCLGE_SHAPER_LVL_CNT] = { + 6 * 256, /* Prioriy level */ + 6 * 32, /* Prioriy group level */ + 6 * 8, /* Port level */ + 6 * 256 /* Qset level */ + }; + u8 ir_u_calc = 0, ir_s_calc = 0; + u32 ir_calc; + u32 tick; + + /* Calc tick */ + if (shaper_level >= HCLGE_SHAPER_LVL_CNT) + return -EINVAL; + + tick = tick_array[shaper_level]; + + /** + * Calc the speed if ir_b = 126, ir_u = 0 and ir_s = 0 + * the formula is changed to: + * 126 * 1 * 8 + * ir_calc = ---------------- * 1000 + * tick * 1 + */ + ir_calc = (1008000 + (tick >> 1) - 1) / tick; + + if (ir_calc == ir) { + *ir_b = 126; + *ir_u = 0; + *ir_s = 0; + + return 0; + } else if (ir_calc > ir) { + /* Increasing the denominator to select ir_s value */ + while (ir_calc > ir) { + ir_s_calc++; + ir_calc = 1008000 / (tick * (1 << ir_s_calc)); + } + + if (ir_calc == ir) + *ir_b = 126; + else + *ir_b = (ir * tick * (1 << ir_s_calc) + 4000) / 8000; + } else { + /* Increasing the numerator to select ir_u value */ + u32 numerator; + + while (ir_calc < ir) { + ir_u_calc++; + numerator = 1008000 * (1 << ir_u_calc); + ir_calc = (numerator + (tick >> 1)) / tick; + } + + if (ir_calc == ir) { + *ir_b = 126; + } else { + u32 denominator = (8000 * (1 << --ir_u_calc)); + *ir_b = (ir * tick + (denominator >> 1)) / denominator; + } + } + + *ir_u = ir_u_calc; + *ir_s = ir_s_calc; + + return 0; +} + +static int hclge_mac_pause_en_cfg(struct hclge_dev *hdev, bool tx, bool rx) +{ + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_CFG_MAC_PAUSE_EN, false); + + desc.data[0] = cpu_to_le32((tx ? HCLGE_TX_MAC_PAUSE_EN_MSK : 0) | + (rx ? HCLGE_RX_MAC_PAUSE_EN_MSK : 0)); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_fill_pri_array(struct hclge_dev *hdev, u8 *pri, u8 pri_id) +{ + u8 tc; + + for (tc = 0; tc < hdev->tm_info.num_tc; tc++) + if (hdev->tm_info.tc_info[tc].up == pri_id) + break; + + if (tc >= hdev->tm_info.num_tc) + return -EINVAL; + + /** + * the register for priority has four bytes, the first bytes includes + * priority0 and priority1, the higher 4bit stands for priority1 + * while the lower 4bit stands for priority0, as below: + * first byte: | pri_1 | pri_0 | + * second byte: | pri_3 | pri_2 | + * third byte: | pri_5 | pri_4 | + * fourth byte: | pri_7 | pri_6 | + */ + pri[pri_id >> 1] |= tc << ((pri_id & 1) * 4); + + return 0; +} + +static int hclge_up_to_tc_map(struct hclge_dev *hdev) +{ + struct hclge_desc desc; + u8 *pri = (u8 *)desc.data; + u8 pri_id; + int ret; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_PRI_TO_TC_MAPPING, false); + + for (pri_id = 0; pri_id < hdev->tm_info.num_tc; pri_id++) { + ret = hclge_fill_pri_array(hdev, pri, pri_id); + if (ret) + return ret; + } + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pg_to_pri_map_cfg(struct hclge_dev *hdev, + u8 pg_id, u8 pri_bit_map) +{ + struct hclge_pg_to_pri_link_cmd *map; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_PG_TO_PRI_LINK, false); + + map = (struct hclge_pg_to_pri_link_cmd *)desc.data; + + map->pg_id = pg_id; + map->pri_bit_map = pri_bit_map; + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_qs_to_pri_map_cfg(struct hclge_dev *hdev, + u16 qs_id, u8 pri) +{ + struct hclge_qs_to_pri_link_cmd *map; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_QS_TO_PRI_LINK, false); + + map = (struct hclge_qs_to_pri_link_cmd *)desc.data; + + map->qs_id = cpu_to_le16(qs_id); + map->priority = pri; + map->link_vld = HCLGE_TM_QS_PRI_LINK_VLD_MSK; + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_q_to_qs_map_cfg(struct hclge_dev *hdev, + u8 q_id, u16 qs_id) +{ + struct hclge_nq_to_qs_link_cmd *map; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_NQ_TO_QS_LINK, false); + + map = (struct hclge_nq_to_qs_link_cmd *)desc.data; + + map->nq_id = cpu_to_le16(q_id); + map->qset_id = cpu_to_le16(qs_id | HCLGE_TM_Q_QS_LINK_VLD_MSK); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pg_weight_cfg(struct hclge_dev *hdev, u8 pg_id, + u8 dwrr) +{ + struct hclge_pg_weight_cmd *weight; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_PG_WEIGHT, false); + + weight = (struct hclge_pg_weight_cmd *)desc.data; + + weight->pg_id = pg_id; + weight->dwrr = dwrr; + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pri_weight_cfg(struct hclge_dev *hdev, u8 pri_id, + u8 dwrr) +{ + struct hclge_priority_weight_cmd *weight; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_PRI_WEIGHT, false); + + weight = (struct hclge_priority_weight_cmd *)desc.data; + + weight->pri_id = pri_id; + weight->dwrr = dwrr; + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_qs_weight_cfg(struct hclge_dev *hdev, u16 qs_id, + u8 dwrr) +{ + struct hclge_qs_weight_cmd *weight; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_QS_WEIGHT, false); + + weight = (struct hclge_qs_weight_cmd *)desc.data; + + weight->qs_id = cpu_to_le16(qs_id); + weight->dwrr = dwrr; + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pg_shapping_cfg(struct hclge_dev *hdev, + enum hclge_shap_bucket bucket, u8 pg_id, + u8 ir_b, u8 ir_u, u8 ir_s, u8 bs_b, u8 bs_s) +{ + struct hclge_pg_shapping_cmd *shap_cfg_cmd; + enum hclge_opcode_type opcode; + struct hclge_desc desc; + + opcode = bucket ? HCLGE_OPC_TM_PG_P_SHAPPING : + HCLGE_OPC_TM_PG_C_SHAPPING; + hclge_cmd_setup_basic_desc(&desc, opcode, false); + + shap_cfg_cmd = (struct hclge_pg_shapping_cmd *)desc.data; + + shap_cfg_cmd->pg_id = pg_id; + + hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, IR_B, ir_b); + hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, IR_U, ir_u); + hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, IR_S, ir_s); + hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, BS_B, bs_b); + hclge_tm_set_feild(shap_cfg_cmd->pg_shapping_para, BS_S, bs_s); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pri_shapping_cfg(struct hclge_dev *hdev, + enum hclge_shap_bucket bucket, u8 pri_id, + u8 ir_b, u8 ir_u, u8 ir_s, + u8 bs_b, u8 bs_s) +{ + struct hclge_pri_shapping_cmd *shap_cfg_cmd; + enum hclge_opcode_type opcode; + struct hclge_desc desc; + + opcode = bucket ? HCLGE_OPC_TM_PRI_P_SHAPPING : + HCLGE_OPC_TM_PRI_C_SHAPPING; + + hclge_cmd_setup_basic_desc(&desc, opcode, false); + + shap_cfg_cmd = (struct hclge_pri_shapping_cmd *)desc.data; + + shap_cfg_cmd->pri_id = pri_id; + + hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, IR_B, ir_b); + hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, IR_U, ir_u); + hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, IR_S, ir_s); + hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, BS_B, bs_b); + hclge_tm_set_feild(shap_cfg_cmd->pri_shapping_para, BS_S, bs_s); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pg_schd_mode_cfg(struct hclge_dev *hdev, u8 pg_id) +{ + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_PG_SCH_MODE_CFG, false); + + if (hdev->tm_info.pg_info[pg_id].pg_sch_mode == HCLGE_SCH_MODE_DWRR) + desc.data[1] = cpu_to_le32(HCLGE_TM_TX_SCHD_DWRR_MSK); + else + desc.data[1] = 0; + + desc.data[0] = cpu_to_le32(pg_id); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_pri_schd_mode_cfg(struct hclge_dev *hdev, u8 pri_id) +{ + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_PRI_SCH_MODE_CFG, false); + + if (hdev->tm_info.tc_info[pri_id].tc_sch_mode == HCLGE_SCH_MODE_DWRR) + desc.data[1] = cpu_to_le32(HCLGE_TM_TX_SCHD_DWRR_MSK); + else + desc.data[1] = 0; + + desc.data[0] = cpu_to_le32(pri_id); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_qs_schd_mode_cfg(struct hclge_dev *hdev, u16 qs_id) +{ + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_QS_SCH_MODE_CFG, false); + + if (hdev->tm_info.tc_info[qs_id].tc_sch_mode == HCLGE_SCH_MODE_DWRR) + desc.data[1] = cpu_to_le32(HCLGE_TM_TX_SCHD_DWRR_MSK); + else + desc.data[1] = 0; + + desc.data[0] = cpu_to_le32(qs_id); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +static int hclge_tm_qs_bp_cfg(struct hclge_dev *hdev, u8 tc) +{ + struct hclge_bp_to_qs_map_cmd *bp_to_qs_map_cmd; + struct hclge_desc desc; + + hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TM_BP_TO_QSET_MAPPING, + false); + + bp_to_qs_map_cmd = (struct hclge_bp_to_qs_map_cmd *)desc.data; + + bp_to_qs_map_cmd->tc_id = tc; + + /* Qset and tc is one by one mapping */ + bp_to_qs_map_cmd->qs_bit_map = cpu_to_le32(1 << tc); + + return hclge_cmd_send(&hdev->hw, &desc, 1); +} + +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; + u8 i; + + kinfo = &vport->nic.kinfo; + vport->bw_limit = hdev->tm_info.pg_info[0].bw_limit; + kinfo->num_tc = + min_t(u16, kinfo->num_tqps, hdev->tm_info.num_tc); + kinfo->rss_size + = min_t(u16, hdev->rss_size_max, + kinfo->num_tqps / kinfo->num_tc); + vport->qs_offset = hdev->tm_info.num_tc * vport->vport_id; + vport->dwrr = 100; /* 100 percent as init */ + + for (i = 0; i < kinfo->num_tc; i++) { + if (hdev->hw_tc_map & BIT(i)) { + 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; + kinfo->tc_info[i].up = hdev->tm_info.tc_info[i].up; + } 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; + kinfo->tc_info[i].up = 0; + } + } +} + +static void hclge_tm_vport_info_update(struct hclge_dev *hdev) +{ + struct hclge_vport *vport = hdev->vport; + u32 i; + + for (i = 0; i < hdev->num_alloc_vport; i++) { + hclge_tm_vport_tc_info_update(vport); + + vport++; + } +} + +static void hclge_tm_tc_info_init(struct hclge_dev *hdev) +{ + u8 i; + + for (i = 0; i < hdev->tm_info.num_tc; i++) { + hdev->tm_info.tc_info[i].tc_id = i; + hdev->tm_info.tc_info[i].tc_sch_mode = HCLGE_SCH_MODE_DWRR; + hdev->tm_info.tc_info[i].up = i; + hdev->tm_info.tc_info[i].pgid = 0; + hdev->tm_info.tc_info[i].bw_limit = + hdev->tm_info.pg_info[0].bw_limit; + } + + hdev->flag &= ~HCLGE_FLAG_DCB_ENABLE; +} + +static void hclge_tm_pg_info_init(struct hclge_dev *hdev) +{ + u8 i; + + for (i = 0; i < hdev->tm_info.num_pg; i++) { + int k; + + hdev->tm_info.pg_dwrr[i] = i ? 0 : 100; + + hdev->tm_info.pg_info[i].pg_id = i; + hdev->tm_info.pg_info[i].pg_sch_mode = HCLGE_SCH_MODE_DWRR; + + hdev->tm_info.pg_info[i].bw_limit = HCLGE_ETHER_MAX_RATE; + + if (i != 0) + continue; + + hdev->tm_info.pg_info[i].tc_bit_map = hdev->hw_tc_map; + for (k = 0; k < hdev->tm_info.num_tc; k++) + hdev->tm_info.pg_info[i].tc_dwrr[k] = 100; + } +} + +static int hclge_tm_schd_info_init(struct hclge_dev *hdev) +{ + if ((hdev->tx_sch_mode != HCLGE_FLAG_TC_BASE_SCH_MODE) && + (hdev->tm_info.num_pg != 1)) + return -EINVAL; + + hclge_tm_pg_info_init(hdev); + + hclge_tm_tc_info_init(hdev); + + hclge_tm_vport_info_update(hdev); + + hdev->tm_info.fc_mode = HCLGE_FC_NONE; + hdev->fc_mode_last_time = hdev->tm_info.fc_mode; + + return 0; +} + +static int hclge_tm_pg_to_pri_map(struct hclge_dev *hdev) +{ + int ret; + u32 i; + + if (hdev->tx_sch_mode != HCLGE_FLAG_TC_BASE_SCH_MODE) + return 0; + + for (i = 0; i < hdev->tm_info.num_pg; i++) { + /* Cfg mapping */ + ret = hclge_tm_pg_to_pri_map_cfg( + hdev, i, hdev->tm_info.pg_info[i].tc_bit_map); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pg_shaper_cfg(struct hclge_dev *hdev) +{ + u8 ir_u, ir_b, ir_s; + int ret; + u32 i; + + /* Cfg pg schd */ + if (hdev->tx_sch_mode != HCLGE_FLAG_TC_BASE_SCH_MODE) + return 0; + + /* Pg to pri */ + for (i = 0; i < hdev->tm_info.num_pg; i++) { + /* Calc shaper para */ + ret = hclge_shaper_para_calc( + hdev->tm_info.pg_info[i].bw_limit, + HCLGE_SHAPER_LVL_PG, + &ir_b, &ir_u, &ir_s); + if (ret) + return ret; + + ret = hclge_tm_pg_shapping_cfg(hdev, + HCLGE_TM_SHAP_C_BUCKET, i, + 0, 0, 0, HCLGE_SHAPER_BS_U_DEF, + HCLGE_SHAPER_BS_S_DEF); + if (ret) + return ret; + + ret = hclge_tm_pg_shapping_cfg(hdev, + HCLGE_TM_SHAP_P_BUCKET, i, + ir_b, ir_u, ir_s, + HCLGE_SHAPER_BS_U_DEF, + HCLGE_SHAPER_BS_S_DEF); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pg_dwrr_cfg(struct hclge_dev *hdev) +{ + int ret; + u32 i; + + /* cfg pg schd */ + if (hdev->tx_sch_mode != HCLGE_FLAG_TC_BASE_SCH_MODE) + return 0; + + /* pg to prio */ + for (i = 0; i < hdev->tm_info.num_pg; i++) { + /* Cfg dwrr */ + ret = hclge_tm_pg_weight_cfg(hdev, i, + hdev->tm_info.pg_dwrr[i]); + if (ret) + return ret; + } + + return 0; +} + +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_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]; + + ret = hclge_tm_q_to_qs_map_cfg(hdev, + hclge_get_queue_id(q), + vport->qs_offset + i); + if (ret) + return ret; + } + } + + return 0; +} + +static int hclge_tm_pri_q_qs_cfg(struct hclge_dev *hdev) +{ + struct hclge_vport *vport = hdev->vport; + int ret; + u32 i; + + if (hdev->tx_sch_mode == HCLGE_FLAG_TC_BASE_SCH_MODE) { + /* Cfg qs -> pri mapping, one by one mapping */ + for (i = 0; i < hdev->tm_info.num_tc; i++) { + ret = hclge_tm_qs_to_pri_map_cfg(hdev, i, i); + if (ret) + return ret; + } + } else if (hdev->tx_sch_mode == HCLGE_FLAG_VNET_BASE_SCH_MODE) { + int k; + /* Cfg qs -> pri mapping, qs = tc, pri = vf, 8 qs -> 1 pri */ + for (k = 0; k < hdev->num_alloc_vport; k++) + for (i = 0; i < HNAE3_MAX_TC; i++) { + ret = hclge_tm_qs_to_pri_map_cfg( + hdev, vport[k].qs_offset + i, k); + if (ret) + return ret; + } + } else { + return -EINVAL; + } + + /* Cfg q -> qs mapping */ + for (i = 0; i < hdev->num_alloc_vport; i++) { + ret = hclge_vport_q_to_qs_map(hdev, vport); + if (ret) + return ret; + + vport++; + } + + return 0; +} + +static int hclge_tm_pri_tc_base_shaper_cfg(struct hclge_dev *hdev) +{ + u8 ir_u, ir_b, ir_s; + int ret; + u32 i; + + for (i = 0; i < hdev->tm_info.num_tc; i++) { + ret = hclge_shaper_para_calc( + hdev->tm_info.tc_info[i].bw_limit, + HCLGE_SHAPER_LVL_PRI, + &ir_b, &ir_u, &ir_s); + if (ret) + return ret; + + ret = hclge_tm_pri_shapping_cfg( + hdev, HCLGE_TM_SHAP_C_BUCKET, i, + 0, 0, 0, HCLGE_SHAPER_BS_U_DEF, + HCLGE_SHAPER_BS_S_DEF); + if (ret) + return ret; + + ret = hclge_tm_pri_shapping_cfg( + hdev, HCLGE_TM_SHAP_P_BUCKET, i, + ir_b, ir_u, ir_s, HCLGE_SHAPER_BS_U_DEF, + HCLGE_SHAPER_BS_S_DEF); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pri_vnet_base_shaper_pri_cfg(struct hclge_vport *vport) +{ + struct hclge_dev *hdev = vport->back; + u8 ir_u, ir_b, ir_s; + int ret; + + ret = hclge_shaper_para_calc(vport->bw_limit, HCLGE_SHAPER_LVL_VF, + &ir_b, &ir_u, &ir_s); + if (ret) + return ret; + + ret = hclge_tm_pri_shapping_cfg(hdev, HCLGE_TM_SHAP_C_BUCKET, + vport->vport_id, + 0, 0, 0, HCLGE_SHAPER_BS_U_DEF, + HCLGE_SHAPER_BS_S_DEF); + if (ret) + return ret; + + ret = hclge_tm_pri_shapping_cfg(hdev, HCLGE_TM_SHAP_P_BUCKET, + vport->vport_id, + ir_b, ir_u, ir_s, + HCLGE_SHAPER_BS_U_DEF, + HCLGE_SHAPER_BS_S_DEF); + if (ret) + return ret; + + return 0; +} + +static int hclge_tm_pri_vnet_base_shaper_qs_cfg(struct hclge_vport *vport) +{ + struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; + struct hclge_dev *hdev = vport->back; + struct hnae3_tc_info *v_tc_info; + u8 ir_u, ir_b, ir_s; + u32 i; + int ret; + + for (i = 0; i < kinfo->num_tc; i++) { + v_tc_info = &kinfo->tc_info[i]; + ret = hclge_shaper_para_calc( + hdev->tm_info.tc_info[i].bw_limit, + HCLGE_SHAPER_LVL_QSET, + &ir_b, &ir_u, &ir_s); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pri_vnet_base_shaper_cfg(struct hclge_dev *hdev) +{ + struct hclge_vport *vport = hdev->vport; + int ret; + u32 i; + + /* Need config vport shaper */ + for (i = 0; i < hdev->num_alloc_vport; i++) { + ret = hclge_tm_pri_vnet_base_shaper_pri_cfg(vport); + if (ret) + return ret; + + ret = hclge_tm_pri_vnet_base_shaper_qs_cfg(vport); + if (ret) + return ret; + + vport++; + } + + return 0; +} + +static int hclge_tm_pri_shaper_cfg(struct hclge_dev *hdev) +{ + int ret; + + if (hdev->tx_sch_mode == HCLGE_FLAG_TC_BASE_SCH_MODE) { + ret = hclge_tm_pri_tc_base_shaper_cfg(hdev); + if (ret) + return ret; + } else { + ret = hclge_tm_pri_vnet_base_shaper_cfg(hdev); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pri_tc_base_dwrr_cfg(struct hclge_dev *hdev) +{ + struct hclge_pg_info *pg_info; + u8 dwrr; + int ret; + u32 i; + + for (i = 0; i < hdev->tm_info.num_tc; i++) { + pg_info = + &hdev->tm_info.pg_info[hdev->tm_info.tc_info[i].pgid]; + dwrr = pg_info->tc_dwrr[i]; + + ret = hclge_tm_pri_weight_cfg(hdev, i, dwrr); + if (ret) + return ret; + + ret = hclge_tm_qs_weight_cfg(hdev, i, dwrr); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pri_vnet_base_dwrr_pri_cfg(struct hclge_vport *vport) +{ + struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; + struct hclge_dev *hdev = vport->back; + int ret; + u8 i; + + /* Vf dwrr */ + ret = hclge_tm_pri_weight_cfg(hdev, vport->vport_id, vport->dwrr); + if (ret) + return ret; + + /* Qset dwrr */ + for (i = 0; i < kinfo->num_tc; i++) { + ret = hclge_tm_qs_weight_cfg( + hdev, vport->qs_offset + i, + hdev->tm_info.pg_info[0].tc_dwrr[i]); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_pri_vnet_base_dwrr_cfg(struct hclge_dev *hdev) +{ + struct hclge_vport *vport = hdev->vport; + int ret; + u32 i; + + for (i = 0; i < hdev->num_alloc_vport; i++) { + ret = hclge_tm_pri_vnet_base_dwrr_pri_cfg(vport); + if (ret) + return ret; + + vport++; + } + + return 0; +} + +static int hclge_tm_pri_dwrr_cfg(struct hclge_dev *hdev) +{ + int ret; + + if (hdev->tx_sch_mode == HCLGE_FLAG_TC_BASE_SCH_MODE) { + ret = hclge_tm_pri_tc_base_dwrr_cfg(hdev); + if (ret) + return ret; + } else { + ret = hclge_tm_pri_vnet_base_dwrr_cfg(hdev); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_map_cfg(struct hclge_dev *hdev) +{ + int ret; + + ret = hclge_tm_pg_to_pri_map(hdev); + if (ret) + return ret; + + return hclge_tm_pri_q_qs_cfg(hdev); +} + +static int hclge_tm_shaper_cfg(struct hclge_dev *hdev) +{ + int ret; + + ret = hclge_tm_pg_shaper_cfg(hdev); + if (ret) + return ret; + + return hclge_tm_pri_shaper_cfg(hdev); +} + +int hclge_tm_dwrr_cfg(struct hclge_dev *hdev) +{ + int ret; + + ret = hclge_tm_pg_dwrr_cfg(hdev); + if (ret) + return ret; + + return hclge_tm_pri_dwrr_cfg(hdev); +} + +static int hclge_tm_lvl2_schd_mode_cfg(struct hclge_dev *hdev) +{ + int ret; + u8 i; + + /* Only being config on TC-Based scheduler mode */ + if (hdev->tx_sch_mode == HCLGE_FLAG_VNET_BASE_SCH_MODE) + return 0; + + for (i = 0; i < hdev->tm_info.num_pg; i++) { + ret = hclge_tm_pg_schd_mode_cfg(hdev, i); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_schd_mode_vnet_base_cfg(struct hclge_vport *vport) +{ + struct hnae3_knic_private_info *kinfo = &vport->nic.kinfo; + struct hclge_dev *hdev = vport->back; + int ret; + u8 i; + + ret = hclge_tm_pri_schd_mode_cfg(hdev, vport->vport_id); + if (ret) + return ret; + + for (i = 0; i < kinfo->num_tc; i++) { + ret = hclge_tm_qs_schd_mode_cfg(hdev, vport->qs_offset + i); + if (ret) + return ret; + } + + return 0; +} + +static int hclge_tm_lvl34_schd_mode_cfg(struct hclge_dev *hdev) +{ + struct hclge_vport *vport = hdev->vport; + int ret; + u8 i; + + if (hdev->tx_sch_mode == HCLGE_FLAG_TC_BASE_SCH_MODE) { + for (i = 0; i < hdev->tm_info.num_tc; i++) { + ret = hclge_tm_pri_schd_mode_cfg(hdev, i); + if (ret) + return ret; + + ret = hclge_tm_qs_schd_mode_cfg(hdev, i); + if (ret) + return ret; + } + } else { + for (i = 0; i < hdev->num_alloc_vport; i++) { + ret = hclge_tm_schd_mode_vnet_base_cfg(vport); + if (ret) + return ret; + + vport++; + } + } + + return 0; +} + +static int hclge_tm_schd_mode_hw(struct hclge_dev *hdev) +{ + int ret; + + ret = hclge_tm_lvl2_schd_mode_cfg(hdev); + if (ret) + return ret; + + return hclge_tm_lvl34_schd_mode_cfg(hdev); +} + +static int hclge_tm_schd_setup_hw(struct hclge_dev *hdev) +{ + int ret; + + /* Cfg tm mapping */ + ret = hclge_tm_map_cfg(hdev); + if (ret) + return ret; + + /* Cfg tm shaper */ + ret = hclge_tm_shaper_cfg(hdev); + if (ret) + return ret; + + /* Cfg dwrr */ + ret = hclge_tm_dwrr_cfg(hdev); + if (ret) + return ret; + + /* Cfg schd mode for each level schd */ + return hclge_tm_schd_mode_hw(hdev); +} + +int hclge_pause_setup_hw(struct hclge_dev *hdev) +{ + bool en = hdev->tm_info.fc_mode != HCLGE_FC_PFC; + int ret; + u8 i; + + ret = hclge_mac_pause_en_cfg(hdev, en, en); + if (ret) + return ret; + + for (i = 0; i < hdev->tm_info.num_tc; i++) { + ret = hclge_tm_qs_bp_cfg(hdev, i); + if (ret) + return ret; + } + + return hclge_up_to_tc_map(hdev); +} + +int hclge_tm_init_hw(struct hclge_dev *hdev) +{ + int ret; + + if ((hdev->tx_sch_mode != HCLGE_FLAG_TC_BASE_SCH_MODE) && + (hdev->tx_sch_mode != HCLGE_FLAG_VNET_BASE_SCH_MODE)) + return -ENOTSUPP; + + ret = hclge_tm_schd_setup_hw(hdev); + if (ret) + return ret; + + ret = hclge_pause_setup_hw(hdev); + if (ret) + return ret; + + return 0; +} + +int hclge_tm_schd_init(struct hclge_dev *hdev) +{ + int ret = hclge_tm_schd_info_init(hdev); + + if (ret) + return ret; + + return hclge_tm_init_hw(hdev); +} diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h new file mode 100644 index 0000000..7e67337 --- /dev/null +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_tm.h @@ -0,0 +1,106 @@ +/* + * Copyright (c) 2016~2017 Hisilicon Limited. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#ifndef __HCLGE_TM_H +#define __HCLGE_TM_H + +#include + +/* MAC Pause */ +#define HCLGE_TX_MAC_PAUSE_EN_MSK BIT(0) +#define HCLGE_RX_MAC_PAUSE_EN_MSK BIT(1) + +#define HCLGE_TM_PORT_BASE_MODE_MSK BIT(0) + +/* SP or DWRR */ +#define HCLGE_TM_TX_SCHD_DWRR_MSK BIT(0) +#define HCLGE_TM_TX_SCHD_SP_MSK (0xFE) + +struct hclge_pg_to_pri_link_cmd { + u8 pg_id; + u8 rsvd1[3]; + u8 pri_bit_map; +}; + +struct hclge_qs_to_pri_link_cmd { + __le16 qs_id; + __le16 rsvd; + u8 priority; +#define HCLGE_TM_QS_PRI_LINK_VLD_MSK BIT(0) + u8 link_vld; +}; + +struct hclge_nq_to_qs_link_cmd { + __le16 nq_id; + __le16 rsvd; +#define HCLGE_TM_Q_QS_LINK_VLD_MSK BIT(10) + __le16 qset_id; +}; + +struct hclge_pg_weight_cmd { + u8 pg_id; + u8 dwrr; +}; + +struct hclge_priority_weight_cmd { + u8 pri_id; + u8 dwrr; +}; + +struct hclge_qs_weight_cmd { + __le16 qs_id; + u8 dwrr; +}; + +#define HCLGE_TM_SHAP_IR_B_MSK GENMASK(7, 0) +#define HCLGE_TM_SHAP_IR_B_LSH 0 +#define HCLGE_TM_SHAP_IR_U_MSK GENMASK(11, 8) +#define HCLGE_TM_SHAP_IR_U_LSH 8 +#define HCLGE_TM_SHAP_IR_S_MSK GENMASK(15, 12) +#define HCLGE_TM_SHAP_IR_S_LSH 12 +#define HCLGE_TM_SHAP_BS_B_MSK GENMASK(20, 16) +#define HCLGE_TM_SHAP_BS_B_LSH 16 +#define HCLGE_TM_SHAP_BS_S_MSK GENMASK(25, 21) +#define HCLGE_TM_SHAP_BS_S_LSH 21 + +enum hclge_shap_bucket { + HCLGE_TM_SHAP_C_BUCKET = 0, + HCLGE_TM_SHAP_P_BUCKET, +}; + +struct hclge_pri_shapping_cmd { + u8 pri_id; + u8 rsvd[3]; + __le32 pri_shapping_para; +}; + +struct hclge_pg_shapping_cmd { + u8 pg_id; + u8 rsvd[3]; + __le32 pg_shapping_para; +}; + +struct hclge_bp_to_qs_map_cmd { + u8 tc_id; + u8 rsvd[2]; + u8 qs_group_id; + __le32 qs_bit_map; + u32 rsvd1; +}; + +#define hclge_tm_set_feild(dest, string, val) \ + hnae_set_field((dest), (HCLGE_TM_SHAP_##string##_MSK), \ + (HCLGE_TM_SHAP_##string##_LSH), val) +#define hclge_tm_get_feild(src, string) \ + hnae_get_field((src), (HCLGE_TM_SHAP_##string##_MSK), \ + (HCLGE_TM_SHAP_##string##_LSH)) + +int hclge_tm_schd_init(struct hclge_dev *hdev); +int hclge_pause_setup_hw(struct hclge_dev *hdev); +#endif -- 2.7.4