From 94816278383eab18c660652bfb7bc929c0d7aef1 Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Tue, 25 Sep 2018 12:05:22 +0300 Subject: [PATCH 1/8] net/mlx5e: RX, Make sure packet header does not cross page boundary In the non-linear SKB memory scheme of Striding RQ, a packet header could cross page boundary. This requires special care in fast path that costs LoC, additional runtime instructions and branches. It could happen when the header (up to 256B) does not fit in a single stride. Avoid this by working with a stride size that fits the maximum possible header. Stride is increased form 64B to 256B. Performance: Tested packet rate for UDP streams, single ring, on ConnectX-5. Configuration: Set Striding RQ and LRO ON (to enabled the non-linear SKB scheme). GRO OFF, early drop by TC rule. 64B: 4x worse memory utilization, no page-crossers headers - No degradation (5,887,305 pps). - The reduction in memory utilization is compensated by the saving of branches tests. 192B: 1.33x worse memory utilization, avoid page-crossers headers - Before: 5,727,252. After: 5,777,037. ~1% gain. 256B: Same memory util, no page-crossers - Before: 5,691,885. After: 5,748,007. ~1% gain. Signed-off-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 11 +++----- .../net/ethernet/mellanox/mlx5/core/en_main.c | 3 +-- .../net/ethernet/mellanox/mlx5/core/en_rx.c | 27 +++---------------- 3 files changed, 9 insertions(+), 32 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 8fa8fdd30b850..6def2c972bf0b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -76,15 +76,14 @@ struct page_pool; #define MLX5_SKB_FRAG_SZ(len) (SKB_DATA_ALIGN(len) + \ SKB_DATA_ALIGN(sizeof(struct skb_shared_info))) +#define MLX5E_RX_MAX_HEAD (256) + #define MLX5_MPWRQ_MIN_LOG_STRIDE_SZ(mdev) \ (6 + MLX5_CAP_GEN(mdev, cache_line_128byte)) /* HW restriction */ #define MLX5_MPWRQ_LOG_STRIDE_SZ(mdev, req) \ max_t(u32, MLX5_MPWRQ_MIN_LOG_STRIDE_SZ(mdev), req) -#define MLX5_MPWRQ_DEF_LOG_STRIDE_SZ(mdev) MLX5_MPWRQ_LOG_STRIDE_SZ(mdev, 6) -#define MLX5_MPWRQ_CQE_CMPRS_LOG_STRIDE_SZ(mdev) MLX5_MPWRQ_LOG_STRIDE_SZ(mdev, 8) -#define MLX5E_MPWQE_STRIDE_SZ(mdev, cqe_cmprs) \ - (cqe_cmprs ? MLX5_MPWRQ_CQE_CMPRS_LOG_STRIDE_SZ(mdev) : \ - MLX5_MPWRQ_DEF_LOG_STRIDE_SZ(mdev)) +#define MLX5_MPWRQ_DEF_LOG_STRIDE_SZ(mdev) \ + MLX5_MPWRQ_LOG_STRIDE_SZ(mdev, order_base_2(MLX5E_RX_MAX_HEAD)) #define MLX5_MPWRQ_LOG_WQE_SZ 18 #define MLX5_MPWRQ_WQE_PAGE_ORDER (MLX5_MPWRQ_LOG_WQE_SZ - PAGE_SHIFT > 0 ? \ @@ -119,8 +118,6 @@ struct page_pool; #define MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE_MPW 0x2 -#define MLX5E_RX_MAX_HEAD (256) - #define MLX5E_PARAMS_DEFAULT_LRO_WQE_SZ (64 * 1024) #define MLX5E_DEFAULT_LRO_TIMEOUT 32 #define MLX5E_LRO_TIMEOUT_ARR_SIZE 4 diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 8cfd2ec7c0a20..1bf547a2b905c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -171,8 +171,7 @@ static u8 mlx5e_mpwqe_get_log_stride_size(struct mlx5_core_dev *mdev, if (mlx5e_rx_mpwqe_is_linear_skb(mdev, params)) return order_base_2(mlx5e_rx_get_linear_frag_sz(params)); - return MLX5E_MPWQE_STRIDE_SZ(mdev, - MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_CQE_COMPRESS)); + return MLX5_MPWRQ_DEF_LOG_STRIDE_SZ(mdev); } static u8 mlx5e_mpwqe_get_log_num_strides(struct mlx5_core_dev *mdev, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index f86e4804e83e2..c025322544112 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -369,7 +369,7 @@ mlx5e_add_skb_frag(struct mlx5e_rq *rq, struct sk_buff *skb, static inline void mlx5e_copy_skb_header(struct device *pdev, struct sk_buff *skb, struct mlx5e_dma_info *dma_info, - int offset_from, int offset_to, u32 headlen) + int offset_from, u32 headlen) { const void *from = page_address(dma_info->page) + offset_from; /* Aligning len to sizeof(long) optimizes memcpy performance */ @@ -377,24 +377,7 @@ mlx5e_copy_skb_header(struct device *pdev, struct sk_buff *skb, dma_sync_single_for_cpu(pdev, dma_info->addr + offset_from, len, DMA_FROM_DEVICE); - skb_copy_to_linear_data_offset(skb, offset_to, from, len); -} - -static inline void -mlx5e_copy_skb_header_mpwqe(struct device *pdev, - struct sk_buff *skb, - struct mlx5e_dma_info *dma_info, - u32 offset, u32 headlen) -{ - u16 headlen_pg = min_t(u32, headlen, PAGE_SIZE - offset); - - mlx5e_copy_skb_header(pdev, skb, dma_info, offset, 0, headlen_pg); - - if (unlikely(offset + headlen > PAGE_SIZE)) { - dma_info++; - mlx5e_copy_skb_header(pdev, skb, dma_info, 0, headlen_pg, - headlen - headlen_pg); - } + skb_copy_to_linear_data(skb, from, len); } static void @@ -973,8 +956,7 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe, } /* copy header */ - mlx5e_copy_skb_header(rq->pdev, skb, head_wi->di, head_wi->offset, - 0, headlen); + mlx5e_copy_skb_header(rq->pdev, skb, head_wi->di, head_wi->offset, headlen); /* skb linear part was allocated with headlen and aligned to long */ skb->tail += headlen; skb->len += headlen; @@ -1096,8 +1078,7 @@ mlx5e_skb_from_cqe_mpwrq_nonlinear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *w di++; } /* copy header */ - mlx5e_copy_skb_header_mpwqe(rq->pdev, skb, head_di, - head_offset, headlen); + mlx5e_copy_skb_header(rq->pdev, skb, head_di, head_offset, headlen); /* skb linear part was allocated with headlen and aligned to long */ skb->tail += headlen; skb->len += headlen; From 79d356ef2c10b19fdd3269b975579aa0be21710c Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Wed, 12 Sep 2018 18:32:49 +0300 Subject: [PATCH 2/8] net/mlx5e: Take CQ decompress fields into a separate structure Only the Receive CQ makes use of these fields. Take them out into a separate struct and save space in the generic CQ structure. Signed-off-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 15 ++- .../net/ethernet/mellanox/mlx5/core/en_rx.c | 127 ++++++++++-------- 2 files changed, 81 insertions(+), 61 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 6def2c972bf0b..587f94d68dc98 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -306,16 +306,18 @@ struct mlx5e_cq { struct mlx5_core_cq mcq; struct mlx5e_channel *channel; + /* control */ + struct mlx5_core_dev *mdev; + struct mlx5_wq_ctrl wq_ctrl; +} ____cacheline_aligned_in_smp; + +struct mlx5e_cq_decomp { /* cqe decompression */ struct mlx5_cqe64 title; struct mlx5_mini_cqe8 mini_arr[MLX5_MINI_CQE_ARRAY_SIZE]; u8 mini_arr_idx; - u16 decmprs_left; - u16 decmprs_wqe_counter; - - /* control */ - struct mlx5_core_dev *mdev; - struct mlx5_wq_ctrl wq_ctrl; + u16 left; + u16 wqe_counter; } ____cacheline_aligned_in_smp; struct mlx5e_tx_wqe_info { @@ -578,6 +580,7 @@ struct mlx5e_rq { struct net_device *netdev; struct mlx5e_rq_stats *stats; struct mlx5e_cq cq; + struct mlx5e_cq_decomp cqd; struct mlx5e_page_cache page_cache; struct hwtstamp_config *tstamp; struct mlx5_clock *clock; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index c025322544112..be396e5e4e39c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -52,40 +52,45 @@ static inline bool mlx5e_rx_hw_stamp(struct hwtstamp_config *config) return config->rx_filter == HWTSTAMP_FILTER_ALL; } -static inline void mlx5e_read_cqe_slot(struct mlx5e_cq *cq, u32 cqcc, - void *data) +static inline void mlx5e_read_cqe_slot(struct mlx5_cqwq *wq, + u32 cqcc, void *data) { - u32 ci = mlx5_cqwq_ctr2ix(&cq->wq, cqcc); + u32 ci = mlx5_cqwq_ctr2ix(wq, cqcc); - memcpy(data, mlx5_cqwq_get_wqe(&cq->wq, ci), sizeof(struct mlx5_cqe64)); + memcpy(data, mlx5_cqwq_get_wqe(wq, ci), sizeof(struct mlx5_cqe64)); } static inline void mlx5e_read_title_slot(struct mlx5e_rq *rq, - struct mlx5e_cq *cq, u32 cqcc) + struct mlx5_cqwq *wq, + u32 cqcc) { - mlx5e_read_cqe_slot(cq, cqcc, &cq->title); - cq->decmprs_left = be32_to_cpu(cq->title.byte_cnt); - cq->decmprs_wqe_counter = be16_to_cpu(cq->title.wqe_counter); + struct mlx5e_cq_decomp *cqd = &rq->cqd; + struct mlx5_cqe64 *title = &cqd->title; + + mlx5e_read_cqe_slot(wq, cqcc, title); + cqd->left = be32_to_cpu(title->byte_cnt); + cqd->wqe_counter = be16_to_cpu(title->wqe_counter); rq->stats->cqe_compress_blks++; } -static inline void mlx5e_read_mini_arr_slot(struct mlx5e_cq *cq, u32 cqcc) +static inline void mlx5e_read_mini_arr_slot(struct mlx5_cqwq *wq, + struct mlx5e_cq_decomp *cqd, + u32 cqcc) { - mlx5e_read_cqe_slot(cq, cqcc, cq->mini_arr); - cq->mini_arr_idx = 0; + mlx5e_read_cqe_slot(wq, cqcc, cqd->mini_arr); + cqd->mini_arr_idx = 0; } -static inline void mlx5e_cqes_update_owner(struct mlx5e_cq *cq, u32 cqcc, int n) +static inline void mlx5e_cqes_update_owner(struct mlx5_cqwq *wq, int n) { - struct mlx5_cqwq *wq = &cq->wq; - + u32 cqcc = wq->cc; u8 op_own = mlx5_cqwq_get_ctr_wrap_cnt(wq, cqcc) & 1; u32 ci = mlx5_cqwq_ctr2ix(wq, cqcc); u32 wq_sz = mlx5_cqwq_get_size(wq); u32 ci_top = min_t(u32, wq_sz, ci + n); for (; ci < ci_top; ci++, n--) { - struct mlx5_cqe64 *cqe = mlx5_cqwq_get_wqe(&cq->wq, ci); + struct mlx5_cqe64 *cqe = mlx5_cqwq_get_wqe(wq, ci); cqe->op_own = op_own; } @@ -93,7 +98,7 @@ static inline void mlx5e_cqes_update_owner(struct mlx5e_cq *cq, u32 cqcc, int n) if (unlikely(ci == wq_sz)) { op_own = !op_own; for (ci = 0; ci < n; ci++) { - struct mlx5_cqe64 *cqe = mlx5_cqwq_get_wqe(&cq->wq, ci); + struct mlx5_cqe64 *cqe = mlx5_cqwq_get_wqe(wq, ci); cqe->op_own = op_own; } @@ -101,68 +106,79 @@ static inline void mlx5e_cqes_update_owner(struct mlx5e_cq *cq, u32 cqcc, int n) } static inline void mlx5e_decompress_cqe(struct mlx5e_rq *rq, - struct mlx5e_cq *cq, u32 cqcc) + struct mlx5_cqwq *wq, + u32 cqcc) { - cq->title.byte_cnt = cq->mini_arr[cq->mini_arr_idx].byte_cnt; - cq->title.check_sum = cq->mini_arr[cq->mini_arr_idx].checksum; - cq->title.op_own &= 0xf0; - cq->title.op_own |= 0x01 & (cqcc >> cq->wq.fbc.log_sz); - cq->title.wqe_counter = cpu_to_be16(cq->decmprs_wqe_counter); + struct mlx5e_cq_decomp *cqd = &rq->cqd; + struct mlx5_mini_cqe8 *mini_cqe = &cqd->mini_arr[cqd->mini_arr_idx]; + struct mlx5_cqe64 *title = &cqd->title; + + title->byte_cnt = mini_cqe->byte_cnt; + title->check_sum = mini_cqe->checksum; + title->op_own &= 0xf0; + title->op_own |= 0x01 & (cqcc >> wq->fbc.log_sz); + title->wqe_counter = cpu_to_be16(cqd->wqe_counter); if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) - cq->decmprs_wqe_counter += - mpwrq_get_cqe_consumed_strides(&cq->title); + cqd->wqe_counter += mpwrq_get_cqe_consumed_strides(title); else - cq->decmprs_wqe_counter = - mlx5_wq_cyc_ctr2ix(&rq->wqe.wq, cq->decmprs_wqe_counter + 1); + cqd->wqe_counter = + mlx5_wq_cyc_ctr2ix(&rq->wqe.wq, cqd->wqe_counter + 1); } static inline void mlx5e_decompress_cqe_no_hash(struct mlx5e_rq *rq, - struct mlx5e_cq *cq, u32 cqcc) + struct mlx5_cqwq *wq, + u32 cqcc) { - mlx5e_decompress_cqe(rq, cq, cqcc); - cq->title.rss_hash_type = 0; - cq->title.rss_hash_result = 0; + struct mlx5e_cq_decomp *cqd = &rq->cqd; + + mlx5e_decompress_cqe(rq, wq, cqcc); + cqd->title.rss_hash_type = 0; + cqd->title.rss_hash_result = 0; } static inline u32 mlx5e_decompress_cqes_cont(struct mlx5e_rq *rq, - struct mlx5e_cq *cq, + struct mlx5_cqwq *wq, int update_owner_only, int budget_rem) { - u32 cqcc = cq->wq.cc + update_owner_only; + struct mlx5e_cq_decomp *cqd = &rq->cqd; + u32 cqcc = wq->cc + update_owner_only; u32 cqe_count; u32 i; - cqe_count = min_t(u32, cq->decmprs_left, budget_rem); + cqe_count = min_t(u32, cqd->left, budget_rem); for (i = update_owner_only; i < cqe_count; - i++, cq->mini_arr_idx++, cqcc++) { - if (cq->mini_arr_idx == MLX5_MINI_CQE_ARRAY_SIZE) - mlx5e_read_mini_arr_slot(cq, cqcc); + i++, cqd->mini_arr_idx++, cqcc++) { + if (cqd->mini_arr_idx == MLX5_MINI_CQE_ARRAY_SIZE) + mlx5e_read_mini_arr_slot(wq, cqd, cqcc); - mlx5e_decompress_cqe_no_hash(rq, cq, cqcc); - rq->handle_rx_cqe(rq, &cq->title); + mlx5e_decompress_cqe_no_hash(rq, wq, cqcc); + rq->handle_rx_cqe(rq, &cqd->title); } - mlx5e_cqes_update_owner(cq, cq->wq.cc, cqcc - cq->wq.cc); - cq->wq.cc = cqcc; - cq->decmprs_left -= cqe_count; + mlx5e_cqes_update_owner(wq, cqcc - wq->cc); + wq->cc = cqcc; + cqd->left -= cqe_count; rq->stats->cqe_compress_pkts += cqe_count; return cqe_count; } static inline u32 mlx5e_decompress_cqes_start(struct mlx5e_rq *rq, - struct mlx5e_cq *cq, + struct mlx5_cqwq *wq, int budget_rem) { - mlx5e_read_title_slot(rq, cq, cq->wq.cc); - mlx5e_read_mini_arr_slot(cq, cq->wq.cc + 1); - mlx5e_decompress_cqe(rq, cq, cq->wq.cc); - rq->handle_rx_cqe(rq, &cq->title); - cq->mini_arr_idx++; + struct mlx5e_cq_decomp *cqd = &rq->cqd; + u32 cc = wq->cc; - return mlx5e_decompress_cqes_cont(rq, cq, 1, budget_rem) - 1; + mlx5e_read_title_slot(rq, wq, cc); + mlx5e_read_mini_arr_slot(wq, cqd, cc + 1); + mlx5e_decompress_cqe(rq, wq, cc); + rq->handle_rx_cqe(rq, &cqd->title); + cqd->mini_arr_idx++; + + return mlx5e_decompress_cqes_cont(rq, wq, 1, budget_rem) - 1; } static inline bool mlx5e_page_is_reserved(struct page *page) @@ -1184,16 +1200,17 @@ void mlx5e_handle_rx_cqe_mpwrq(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe) int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget) { struct mlx5e_rq *rq = container_of(cq, struct mlx5e_rq, cq); + struct mlx5_cqwq *cqwq = &cq->wq; struct mlx5_cqe64 *cqe; int work_done = 0; if (unlikely(!test_bit(MLX5E_RQ_STATE_ENABLED, &rq->state))) return 0; - if (cq->decmprs_left) - work_done += mlx5e_decompress_cqes_cont(rq, cq, 0, budget); + if (rq->cqd.left) + work_done += mlx5e_decompress_cqes_cont(rq, cqwq, 0, budget); - cqe = mlx5_cqwq_get_cqe(&cq->wq); + cqe = mlx5_cqwq_get_cqe(cqwq); if (!cqe) { if (unlikely(work_done)) goto out; @@ -1203,21 +1220,21 @@ int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget) do { if (mlx5_get_cqe_format(cqe) == MLX5_COMPRESSED) { work_done += - mlx5e_decompress_cqes_start(rq, cq, + mlx5e_decompress_cqes_start(rq, cqwq, budget - work_done); continue; } - mlx5_cqwq_pop(&cq->wq); + mlx5_cqwq_pop(cqwq); rq->handle_rx_cqe(rq, cqe); - } while ((++work_done < budget) && (cqe = mlx5_cqwq_get_cqe(&cq->wq))); + } while ((++work_done < budget) && (cqe = mlx5_cqwq_get_cqe(cqwq))); out: if (rq->xdp_prog) mlx5e_xdp_rx_poll_complete(rq); - mlx5_cqwq_update_db_record(&cq->wq); + mlx5_cqwq_update_db_record(cqwq); /* ensure cq space is freed before enabling more cqes */ wmb(); From 149e566fef8120a00c49f24e8f570186200e1e81 Mon Sep 17 00:00:00 2001 From: Moshe Shemesh Date: Thu, 25 Oct 2018 15:03:38 +0300 Subject: [PATCH 3/8] net/mlx5e: Expand XPS cpumask to cover all online cpus Currently we have one cpu in XPS cpumask per tx queue, this is good enough for default configuration where there is a tx queue per cpu. However, once configuration changes to use less tx queues, part of the cpus are not XPS-mapped and so the select queue decision falls back to hash calculation and balancing is not guaranteed. Expand XPS cpumask to enable using all cpus even when number of tx queues is smaller than number of cpus. Signed-off-by: Moshe Shemesh Reviewed-off-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 + .../net/ethernet/mellanox/mlx5/core/en_main.c | 34 +++++++++++++++++-- 2 files changed, 33 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 587f94d68dc98..3e35a90cf21ab 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -638,6 +638,7 @@ struct mlx5e_channel { struct hwtstamp_config *tstamp; int ix; int cpu; + cpumask_var_t xps_cpumask; }; struct mlx5e_channels { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 1bf547a2b905c..2cf6816f14262 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -1949,6 +1949,29 @@ static int mlx5e_set_tx_maxrate(struct net_device *dev, int index, u32 rate) return err; } +static int mlx5e_alloc_xps_cpumask(struct mlx5e_channel *c, + struct mlx5e_params *params) +{ + int num_comp_vectors = mlx5_comp_vectors_count(c->mdev); + int irq; + + if (!zalloc_cpumask_var(&c->xps_cpumask, GFP_KERNEL)) + return -ENOMEM; + + for (irq = c->ix; irq < num_comp_vectors; irq += params->num_channels) { + int cpu = cpumask_first(mlx5_comp_irq_get_affinity_mask(c->mdev, irq)); + + cpumask_set_cpu(cpu, c->xps_cpumask); + } + + return 0; +} + +static void mlx5e_free_xps_cpumask(struct mlx5e_channel *c) +{ + free_cpumask_var(c->xps_cpumask); +} + static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, struct mlx5e_params *params, struct mlx5e_channel_param *cparam, @@ -1981,9 +2004,12 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, c->num_tc = params->num_tc; c->xdp = !!params->xdp_prog; c->stats = &priv->channel_stats[ix].ch; - c->irq_desc = irq_to_desc(irq); + err = mlx5e_alloc_xps_cpumask(c, params); + if (err) + goto err_free_channel; + netif_napi_add(netdev, &c->napi, mlx5e_napi_poll, 64); err = mlx5e_open_cq(c, icocq_moder, &cparam->icosq_cq, &c->icosq.cq); @@ -2066,6 +2092,9 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, err_napi_del: netif_napi_del(&c->napi); + mlx5e_free_xps_cpumask(c); + +err_free_channel: kvfree(c); return err; @@ -2078,7 +2107,7 @@ static void mlx5e_activate_channel(struct mlx5e_channel *c) for (tc = 0; tc < c->num_tc; tc++) mlx5e_activate_txqsq(&c->sq[tc]); mlx5e_activate_rq(&c->rq); - netif_set_xps_queue(c->netdev, get_cpu_mask(c->cpu), c->ix); + netif_set_xps_queue(c->netdev, c->xps_cpumask, c->ix); } static void mlx5e_deactivate_channel(struct mlx5e_channel *c) @@ -2106,6 +2135,7 @@ static void mlx5e_close_channel(struct mlx5e_channel *c) mlx5e_close_tx_cqs(c); mlx5e_close_cq(&c->icosq.cq); netif_napi_del(&c->napi); + mlx5e_free_xps_cpumask(c); kvfree(c); } From 71129676ab59dc63ba9e48a09a3386242a516751 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Sun, 20 Jan 2019 10:30:25 +0200 Subject: [PATCH 4/8] net/mlx5e: Return the allocated flow directly from __mlx5e_add_fdb_flow This confusing construction confuses the compiler which can't see that flow is initialized if !err: drivers/net/ethernet/mellanox/mlx5/core/en_tc.c: In function `mlx5e_configure_flower` drivers/net/ethernet/mellanox/mlx5/core/en_tc.c:2727:28: warning: `flow` may be used uninitialized in this function [-Wmaybe-uninitialized] There is no reason for two function outputs, just return the pointer directly and use ERR_PTR to encode a failure. Signed-off-by: Jason Gunthorpe Signed-off-by: Tariq Toukan Reviewed-by: Or Gerlitz Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 29 +++++++++---------- 1 file changed, 14 insertions(+), 15 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index cae6c6d489847..74159d39dd666 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -2767,14 +2767,13 @@ mlx5e_alloc_flow(struct mlx5e_priv *priv, int attr_size, return err; } -static int +static struct mlx5e_tc_flow * __mlx5e_add_fdb_flow(struct mlx5e_priv *priv, struct tc_cls_flower_offload *f, u16 flow_flags, struct net_device *filter_dev, struct mlx5_eswitch_rep *in_rep, - struct mlx5_core_dev *in_mdev, - struct mlx5e_tc_flow **__flow) + struct mlx5_core_dev *in_mdev) { struct netlink_ext_ack *extack = f->common.extack; struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; @@ -2814,15 +2813,13 @@ __mlx5e_add_fdb_flow(struct mlx5e_priv *priv, if (err) goto err_free; - *__flow = flow; - - return 0; + return flow; err_free: kfree(flow); kvfree(parse_attr); out: - return err; + return ERR_PTR(err); } static int mlx5e_tc_add_fdb_peer_flow(struct tc_cls_flower_offload *f, @@ -2855,11 +2852,13 @@ static int mlx5e_tc_add_fdb_peer_flow(struct tc_cls_flower_offload *f, in_mdev = priv->mdev; parse_attr = flow->esw_attr->parse_attr; - err = __mlx5e_add_fdb_flow(peer_priv, f, flow->flags, - parse_attr->filter_dev, - flow->esw_attr->in_rep, in_mdev, &peer_flow); - if (err) + peer_flow = __mlx5e_add_fdb_flow(peer_priv, f, flow->flags, + parse_attr->filter_dev, + flow->esw_attr->in_rep, in_mdev); + if (IS_ERR(peer_flow)) { + err = PTR_ERR(peer_flow); goto out; + } flow->peer_flow = peer_flow; flow->flags |= MLX5E_TC_FLOW_DUP; @@ -2885,10 +2884,10 @@ mlx5e_add_fdb_flow(struct mlx5e_priv *priv, struct mlx5e_tc_flow *flow; int err; - err = __mlx5e_add_fdb_flow(priv, f, flow_flags, filter_dev, in_rep, - in_mdev, &flow); - if (err) - goto out; + flow = __mlx5e_add_fdb_flow(priv, f, flow_flags, filter_dev, in_rep, + in_mdev); + if (IS_ERR(flow)) + return PTR_ERR(flow); if (is_peer_flow_needed(flow)) { err = mlx5e_tc_add_fdb_peer_flow(f, flow); From 8e4ca98609a55d696c96dd260df7a2c5f7398299 Mon Sep 17 00:00:00 2001 From: Or Gerlitz Date: Tue, 18 Dec 2018 21:18:20 +0200 Subject: [PATCH 5/8] net/mlx5: Add trace points for flow tables create/destroy We were not tracking flow tables so far, add it up. Signed-off-by: Or Gerlitz Signed-off-by: Saeed Mahameed --- .../mellanox/mlx5/core/diag/fs_tracepoint.c | 2 ++ .../mellanox/mlx5/core/diag/fs_tracepoint.h | 35 +++++++++++++++++++ .../net/ethernet/mellanox/mlx5/core/fs_core.c | 2 ++ 3 files changed, 39 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c index 424457ff9759f..8ecac81a385dd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c @@ -258,6 +258,8 @@ const char *parse_fs_dst(struct trace_seq *p, return ret; } +EXPORT_TRACEPOINT_SYMBOL(mlx5_fs_add_ft); +EXPORT_TRACEPOINT_SYMBOL(mlx5_fs_del_ft); EXPORT_TRACEPOINT_SYMBOL(mlx5_fs_add_fg); EXPORT_TRACEPOINT_SYMBOL(mlx5_fs_del_fg); EXPORT_TRACEPOINT_SYMBOL(mlx5_fs_set_fte); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h index d027ce00c8ce4..a4cf123e3f177 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h @@ -61,6 +61,41 @@ const char *parse_fs_dst(struct trace_seq *p, const struct mlx5_flow_destination *dst, u32 counter_id); +TRACE_EVENT(mlx5_fs_add_ft, + TP_PROTO(const struct mlx5_flow_table *ft), + TP_ARGS(ft), + TP_STRUCT__entry( + __field(const struct mlx5_flow_table *, ft) + __field(u32, id) + __field(u32, level) + __field(u32, type) + ), + TP_fast_assign( + __entry->ft = ft; + __entry->id = ft->id; + __entry->level = ft->level; + __entry->type = ft->type; + ), + TP_printk("ft=%p id=%u level=%u type=%u \n", + __entry->ft, __entry->id, __entry->level, __entry->type) + ); + +TRACE_EVENT(mlx5_fs_del_ft, + TP_PROTO(const struct mlx5_flow_table *ft), + TP_ARGS(ft), + TP_STRUCT__entry( + __field(const struct mlx5_flow_table *, ft) + __field(u32, id) + ), + TP_fast_assign( + __entry->ft = ft; + __entry->id = ft->id; + + ), + TP_printk("ft=%p id=%u\n", + __entry->ft, __entry->id) + ); + TRACE_EVENT(mlx5_fs_add_fg, TP_PROTO(const struct mlx5_flow_group *fg), TP_ARGS(fg), diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 79f122b45def1..b2df7c3749fb3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -397,6 +397,7 @@ static void del_hw_flow_table(struct fs_node *node) fs_get_obj(ft, node); dev = get_dev(&ft->node); root = find_root(&ft->node); + trace_mlx5_fs_del_ft(ft); if (node->active) { err = root->cmds->destroy_flow_table(dev, ft); @@ -1019,6 +1020,7 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa fs_prio->num_ft++; up_write_ref_node(&fs_prio->node); mutex_unlock(&root->chain_lock); + trace_mlx5_fs_add_ft(ft); return ft; destroy_ft: root->cmds->destroy_flow_table(root->dev, ft); From 9659e49a6db45474000e4df1a2875bd0d5ab2390 Mon Sep 17 00:00:00 2001 From: Saeed Mahameed Date: Fri, 2 Nov 2018 18:21:27 -0700 Subject: [PATCH 6/8] net/mlx5e: Separate between ethtool and netdev software stats folding mlx5e_grp_sw_update_stats can be called from two threads, 1) ndo_get_stats64 2) get_ethtool_stats For this reason and to minimize concurrency issue impact on 64bit machines mlx5e_grp_sw_update_stats folds the software stats into a temporary variable then copies it to the global driver stats, both ethtool and ndo statistics callbacks will use the global software stats variable to report whatever stats they need. Actually ndo_get_stats64 doesn't need to fold the whole software stats (mlx5e_grp_sw_update_stats), all it needs is five counters to fill the rtnl_link_stats64 relevant stats parameter. Hence this patch introduces a simpler helper function to fold software stats for ndo_get_stats64 which will work directly on rtnl_link_stats64 stats parameter and not on the global or even temporary mlx5e_sw_stats variable. Since now mlx5e_grp_sw_update_stats is not called by ndo_get_stats64 we can make it static and remove the temp var. Unlike mlx5e_grp_sw_update_stats the new fold stats function doesn't need to zero out the output statistics parameter since it is already done by the stack @dev_get_stats(). This patch is fixing stack usage of mlx5e_grp_sw_update_stats on x86 gcc-4.9 and higher, the concurrency issue between mlx5's ndo_get_stats64 and get_ethtool_stats is resolved as well. Fixes: 8bfaf07f7806 ("net/mlx5e: Present SW stats when state is not opened") Reported-by: Arnd Bergmann Reported-by: Andrew Morton Reviewed-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_main.c | 30 ++++++++++++++----- .../ethernet/mellanox/mlx5/core/en_stats.c | 6 ++-- .../ethernet/mellanox/mlx5/core/en_stats.h | 1 - 3 files changed, 25 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 2cf6816f14262..01d0895b62641 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -3521,11 +3521,32 @@ static int mlx5e_setup_tc(struct net_device *dev, enum tc_setup_type type, } } +static void mlx5e_fold_sw_stats64(struct mlx5e_priv *priv, struct rtnl_link_stats64 *s) +{ + int i; + + for (i = 0; i < mlx5e_get_netdev_max_channels(priv->netdev); i++) { + struct mlx5e_channel_stats *channel_stats = &priv->channel_stats[i]; + struct mlx5e_rq_stats *rq_stats = &channel_stats->rq; + int j; + + s->rx_packets += rq_stats->packets; + s->rx_bytes += rq_stats->bytes; + + for (j = 0; j < priv->max_opened_tc; j++) { + struct mlx5e_sq_stats *sq_stats = &channel_stats->sq[j]; + + s->tx_packets += sq_stats->packets; + s->tx_bytes += sq_stats->bytes; + s->tx_dropped += sq_stats->dropped; + } + } +} + void mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats) { struct mlx5e_priv *priv = netdev_priv(dev); - struct mlx5e_sw_stats *sstats = &priv->stats.sw; struct mlx5e_vport_stats *vstats = &priv->stats.vport; struct mlx5e_pport_stats *pstats = &priv->stats.pport; @@ -3540,12 +3561,7 @@ mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats) stats->tx_packets = PPORT_802_3_GET(pstats, a_frames_transmitted_ok); stats->tx_bytes = PPORT_802_3_GET(pstats, a_octets_transmitted_ok); } else { - mlx5e_grp_sw_update_stats(priv); - stats->rx_packets = sstats->rx_packets; - stats->rx_bytes = sstats->rx_bytes; - stats->tx_packets = sstats->tx_packets; - stats->tx_bytes = sstats->tx_bytes; - stats->tx_dropped = sstats->tx_queue_dropped; + mlx5e_fold_sw_stats64(priv, stats); } stats->rx_dropped = priv->stats.qcnt.rx_out_of_buffer; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index d3fe48ff9da97..1a78e05cbba81 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -127,9 +127,9 @@ static int mlx5e_grp_sw_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } -void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) +static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) { - struct mlx5e_sw_stats temp, *s = &temp; + struct mlx5e_sw_stats *s = &priv->stats.sw; int i; memset(s, 0, sizeof(*s)); @@ -212,8 +212,6 @@ void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) s->tx_cqes += sq_stats->cqes; } } - - memcpy(&priv->stats.sw, s, sizeof(*s)); } static const struct counter_desc q_stats_desc[] = { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h index fe91ec06e3c79..4640d4f986f8c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h @@ -277,7 +277,6 @@ struct mlx5e_stats_grp { extern const struct mlx5e_stats_grp mlx5e_stats_grps[]; extern const int mlx5e_num_stats_grps; -void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv); void mlx5e_grp_802_3_update_stats(struct mlx5e_priv *priv); #endif /* __MLX5_EN_STATS_H__ */ From 168af00a3b7b4cb9522687ea08bc2dfa5e9b5774 Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Thu, 8 Nov 2018 11:37:24 +0200 Subject: [PATCH 7/8] net/mlx5e: Present the representors SW stats when state is not opened This behavior is already adopted for all other cases in the cited patch. The representor's functions were missed, here we modify the them to behave similarly. Fixes: 8bfaf07f7806 ("net/mlx5e: Present SW stats when state is not opened") Signed-off-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_rep.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 04736212a21c9..22f686b779efa 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -162,21 +162,19 @@ static void mlx5e_rep_update_hw_counters(struct mlx5e_priv *priv) static void mlx5e_rep_update_sw_counters(struct mlx5e_priv *priv) { struct mlx5e_sw_stats *s = &priv->stats.sw; - struct mlx5e_rq_stats *rq_stats; - struct mlx5e_sq_stats *sq_stats; int i, j; memset(s, 0, sizeof(*s)); - for (i = 0; i < priv->channels.num; i++) { - struct mlx5e_channel *c = priv->channels.c[i]; - - rq_stats = c->rq.stats; + for (i = 0; i < mlx5e_get_netdev_max_channels(priv->netdev); i++) { + struct mlx5e_channel_stats *channel_stats = + &priv->channel_stats[i]; + struct mlx5e_rq_stats *rq_stats = &channel_stats->rq; s->rx_packets += rq_stats->packets; s->rx_bytes += rq_stats->bytes; - for (j = 0; j < priv->channels.params.num_tc; j++) { - sq_stats = c->sq[j].stats; + for (j = 0; j < priv->max_opened_tc; j++) { + struct mlx5e_sq_stats *sq_stats = &channel_stats->sq[j]; s->tx_packets += sq_stats->packets; s->tx_bytes += sq_stats->bytes; @@ -195,8 +193,7 @@ static void mlx5e_rep_get_ethtool_stats(struct net_device *dev, return; mutex_lock(&priv->state_lock); - if (test_bit(MLX5E_STATE_OPENED, &priv->state)) - mlx5e_rep_update_sw_counters(priv); + mlx5e_rep_update_sw_counters(priv); mlx5e_rep_update_hw_counters(priv); mutex_unlock(&priv->state_lock); From b832d4fdf105b5464d786e321e3c9e012e67cdfb Mon Sep 17 00:00:00 2001 From: Saeed Mahameed Date: Wed, 12 Dec 2018 01:42:30 -0800 Subject: [PATCH 8/8] net/mlx5e: Reuse fold sw stats in representors Representors software stats are basic, this patch is reusing the mlx5e_fold_sw_stats in representors, which sums up the basic stats64 for a mlx5e netdevice. Fixes: 8bfaf07f7806 ("net/mlx5e: Present SW stats when state is not opened") Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 + .../net/ethernet/mellanox/mlx5/core/en_main.c | 2 +- .../net/ethernet/mellanox/mlx5/core/en_rep.c | 34 +++++-------------- 3 files changed, 10 insertions(+), 27 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 3e35a90cf21ab..6dd74ef693896 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -804,6 +804,7 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe, void mlx5e_update_stats(struct mlx5e_priv *priv); void mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats); +void mlx5e_fold_sw_stats64(struct mlx5e_priv *priv, struct rtnl_link_stats64 *s); void mlx5e_init_l2_addr(struct mlx5e_priv *priv); int mlx5e_self_test_num(struct mlx5e_priv *priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 01d0895b62641..17b6babbed8a6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -3521,7 +3521,7 @@ static int mlx5e_setup_tc(struct net_device *dev, enum tc_setup_type type, } } -static void mlx5e_fold_sw_stats64(struct mlx5e_priv *priv, struct rtnl_link_stats64 *s) +void mlx5e_fold_sw_stats64(struct mlx5e_priv *priv, struct rtnl_link_stats64 *s) { int i; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 22f686b779efa..edb34b397c538 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -162,25 +162,16 @@ static void mlx5e_rep_update_hw_counters(struct mlx5e_priv *priv) static void mlx5e_rep_update_sw_counters(struct mlx5e_priv *priv) { struct mlx5e_sw_stats *s = &priv->stats.sw; - int i, j; + struct rtnl_link_stats64 stats64 = {}; memset(s, 0, sizeof(*s)); - for (i = 0; i < mlx5e_get_netdev_max_channels(priv->netdev); i++) { - struct mlx5e_channel_stats *channel_stats = - &priv->channel_stats[i]; - struct mlx5e_rq_stats *rq_stats = &channel_stats->rq; - - s->rx_packets += rq_stats->packets; - s->rx_bytes += rq_stats->bytes; - - for (j = 0; j < priv->max_opened_tc; j++) { - struct mlx5e_sq_stats *sq_stats = &channel_stats->sq[j]; + mlx5e_fold_sw_stats64(priv, &stats64); - s->tx_packets += sq_stats->packets; - s->tx_bytes += sq_stats->bytes; - s->tx_queue_dropped += sq_stats->dropped; - } - } + s->rx_packets = stats64.rx_packets; + s->rx_bytes = stats64.rx_bytes; + s->tx_packets = stats64.tx_packets; + s->tx_bytes = stats64.tx_bytes; + s->tx_queue_dropped = stats64.tx_dropped; } static void mlx5e_rep_get_ethtool_stats(struct net_device *dev, @@ -1226,17 +1217,8 @@ mlx5e_get_sw_stats64(const struct net_device *dev, struct rtnl_link_stats64 *stats) { struct mlx5e_priv *priv = netdev_priv(dev); - struct mlx5e_sw_stats *sstats = &priv->stats.sw; - - mlx5e_rep_update_sw_counters(priv); - - stats->rx_packets = sstats->rx_packets; - stats->rx_bytes = sstats->rx_bytes; - stats->tx_packets = sstats->tx_packets; - stats->tx_bytes = sstats->tx_bytes; - - stats->tx_dropped = sstats->tx_queue_dropped; + mlx5e_fold_sw_stats64(priv, stats); return 0; }