Merge tag 'mlx5-fixes-2019-10-18' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
Mellanox, mlx5 kTLS fixes 18-10-2019

This series introduces kTLS related fixes to mlx5 driver from Tariq,
and two misc memory leak fixes form Navid Emamdoost.

Please pull and let me know if there is any problem.

I would appreciate it if you queue up kTLS fixes from the list below to
stable kernel v5.3 !

For -stable v4.13:
  nett/mlx5: prevent memory leak in mlx5_fpga_conn_create_cq
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/mellanox/mlx5/core/en.h

@@ -345,7 +345,7 @@ struct mlx5e_tx_wqe_info {
 	u8  num_wqebbs;
 	u8  num_dma;
 #ifdef CONFIG_MLX5_EN_TLS
-	skb_frag_t *resync_dump_frag;
+	struct page *resync_dump_frag_page;
 #endif
 };
 
@@ -410,6 +410,7 @@ struct mlx5e_txqsq {
 	struct device             *pdev;
 	__be32                     mkey_be;
 	unsigned long              state;
+	unsigned int               hw_mtu;
 	struct hwtstamp_config    *tstamp;
 	struct mlx5_clock         *clock;
 

drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h

@@ -15,15 +15,14 @@
 #else
 /* TLS offload requires additional stop_room for:
  *  - a resync SKB.
- * kTLS offload requires additional stop_room for:
- * - static params WQE,
- * - progress params WQE, and
- * - resync DUMP per frag.
+ * kTLS offload requires fixed additional stop_room for:
+ * - a static params WQE, and a progress params WQE.
+ * The additional MTU-depending room for the resync DUMP WQEs
+ * will be calculated and added in runtime.
  */
 #define MLX5E_SQ_TLS_ROOM  \
 	(MLX5_SEND_WQE_MAX_WQEBBS + \
-	 MLX5E_KTLS_STATIC_WQEBBS + MLX5E_KTLS_PROGRESS_WQEBBS + \
-	 MAX_SKB_FRAGS * MLX5E_KTLS_MAX_DUMP_WQEBBS)
+	 MLX5E_KTLS_STATIC_WQEBBS + MLX5E_KTLS_PROGRESS_WQEBBS)
 #endif
 
 #define INL_HDR_START_SZ (sizeof(((struct mlx5_wqe_eth_seg *)NULL)->inline_hdr.start))
@@ -92,7 +91,7 @@ mlx5e_fill_sq_frag_edge(struct mlx5e_txqsq *sq, struct mlx5_wq_cyc *wq,
 
 	/* fill sq frag edge with nops to avoid wqe wrapping two pages */
 	for (; wi < edge_wi; wi++) {
-		wi->skb        = NULL;
+		memset(wi, 0, sizeof(*wi));
 		wi->num_wqebbs = 1;
 		mlx5e_post_nop(wq, sq->sqn, &sq->pc);
 	}

drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c

@@ -38,7 +38,7 @@ static int mlx5e_ktls_add(struct net_device *netdev, struct sock *sk,
 		return -ENOMEM;
 
 	tx_priv->expected_seq = start_offload_tcp_sn;
-	tx_priv->crypto_info  = crypto_info;
+	tx_priv->crypto_info  = *(struct tls12_crypto_info_aes_gcm_128 *)crypto_info;
 	mlx5e_set_ktls_tx_priv_ctx(tls_ctx, tx_priv);
 
 	/* tc and underlay_qpn values are not in use for tls tis */

drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.h

@@ -21,7 +21,14 @@
 	 MLX5_ST_SZ_BYTES(tls_progress_params))
 #define MLX5E_KTLS_PROGRESS_WQEBBS \
 	(DIV_ROUND_UP(MLX5E_KTLS_PROGRESS_WQE_SZ, MLX5_SEND_WQE_BB))
-#define MLX5E_KTLS_MAX_DUMP_WQEBBS 2
+
+struct mlx5e_dump_wqe {
+	struct mlx5_wqe_ctrl_seg ctrl;
+	struct mlx5_wqe_data_seg data;
+};
+
+#define MLX5E_KTLS_DUMP_WQEBBS \
+	(DIV_ROUND_UP(sizeof(struct mlx5e_dump_wqe), MLX5_SEND_WQE_BB))
 
 enum {
 	MLX5E_TLS_PROGRESS_PARAMS_AUTH_STATE_NO_OFFLOAD     = 0,
@@ -37,7 +44,7 @@ enum {
 
 struct mlx5e_ktls_offload_context_tx {
 	struct tls_offload_context_tx *tx_ctx;
-	struct tls_crypto_info *crypto_info;
+	struct tls12_crypto_info_aes_gcm_128 crypto_info;
 	u32 expected_seq;
 	u32 tisn;
 	u32 key_id;
@@ -86,14 +93,28 @@ struct sk_buff *mlx5e_ktls_handle_tx_skb(struct net_device *netdev,
 					 struct mlx5e_tx_wqe **wqe, u16 *pi);
 void mlx5e_ktls_tx_handle_resync_dump_comp(struct mlx5e_txqsq *sq,
 					   struct mlx5e_tx_wqe_info *wi,
-					   struct mlx5e_sq_dma *dma);
-
+					   u32 *dma_fifo_cc);
+static inline u8
+mlx5e_ktls_dumps_num_wqebbs(struct mlx5e_txqsq *sq, unsigned int nfrags,
+			    unsigned int sync_len)
+{
+	/* Given the MTU and sync_len, calculates an upper bound for the
+	 * number of WQEBBs needed for the TX resync DUMP WQEs of a record.
+	 */
+	return MLX5E_KTLS_DUMP_WQEBBS *
+		(nfrags + DIV_ROUND_UP(sync_len, sq->hw_mtu));
+}
 #else
 
 static inline void mlx5e_ktls_build_netdev(struct mlx5e_priv *priv)
 {
 }
 
+static inline void
+mlx5e_ktls_tx_handle_resync_dump_comp(struct mlx5e_txqsq *sq,
+				      struct mlx5e_tx_wqe_info *wi,
+				      u32 *dma_fifo_cc) {}
+
 #endif
 
 #endif /* __MLX5E_TLS_H__ */

drivers/net/ethernet/mellanox/mlx5/core/en_main.c

@@ -1128,6 +1128,7 @@ static int mlx5e_alloc_txqsq(struct mlx5e_channel *c,
 	sq->txq_ix    = txq_ix;
 	sq->uar_map   = mdev->mlx5e_res.bfreg.map;
 	sq->min_inline_mode = params->tx_min_inline_mode;
+	sq->hw_mtu    = MLX5E_SW2HW_MTU(params, params->sw_mtu);
 	sq->stats     = &c->priv->channel_stats[c->ix].sq[tc];
 	sq->stop_room = MLX5E_SQ_STOP_ROOM;
 	INIT_WORK(&sq->recover_work, mlx5e_tx_err_cqe_work);
@@ -1135,10 +1136,14 @@ static int mlx5e_alloc_txqsq(struct mlx5e_channel *c,
 		set_bit(MLX5E_SQ_STATE_VLAN_NEED_L2_INLINE, &sq->state);
 	if (MLX5_IPSEC_DEV(c->priv->mdev))
 		set_bit(MLX5E_SQ_STATE_IPSEC, &sq->state);
+#ifdef CONFIG_MLX5_EN_TLS
 	if (mlx5_accel_is_tls_device(c->priv->mdev)) {
 		set_bit(MLX5E_SQ_STATE_TLS, &sq->state);
-		sq->stop_room += MLX5E_SQ_TLS_ROOM;
+		sq->stop_room += MLX5E_SQ_TLS_ROOM +
+			mlx5e_ktls_dumps_num_wqebbs(sq, MAX_SKB_FRAGS,
+						    TLS_MAX_PAYLOAD_SIZE);
 	}
+#endif
 
 	param->wq.db_numa_node = cpu_to_node(c->cpu);
 	err = mlx5_wq_cyc_create(mdev, &param->wq, sqc_wq, wq, &sq->wq_ctrl);
@@ -1349,9 +1354,13 @@ static void mlx5e_deactivate_txqsq(struct mlx5e_txqsq *sq)
 	/* last doorbell out, godspeed .. */
 	if (mlx5e_wqc_has_room_for(wq, sq->cc, sq->pc, 1)) {
 		u16 pi = mlx5_wq_cyc_ctr2ix(wq, sq->pc);
+		struct mlx5e_tx_wqe_info *wi;
 		struct mlx5e_tx_wqe *nop;
 
-		sq->db.wqe_info[pi].skb = NULL;
+		wi = &sq->db.wqe_info[pi];
+
+		memset(wi, 0, sizeof(*wi));
+		wi->num_wqebbs = 1;
 		nop = mlx5e_post_nop(wq, sq->sqn, &sq->pc);
 		mlx5e_notify_hw(wq, sq->pc, sq->uar_map, &nop->ctrl);
 	}

drivers/net/ethernet/mellanox/mlx5/core/en_stats.c

@@ -52,11 +52,12 @@ static const struct counter_desc sw_stats_desc[] = {
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_encrypted_bytes) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_ctx) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_ooo) },
+	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_dump_packets) },
+	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_dump_bytes) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_resync_bytes) },
+	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_skip_no_sync_data) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_drop_no_sync_data) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_drop_bypass_req) },
-	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_dump_packets) },
-	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_tls_dump_bytes) },
 #endif
 
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_lro_packets) },
@@ -288,11 +289,12 @@ static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv)
 			s->tx_tls_encrypted_bytes   += sq_stats->tls_encrypted_bytes;
 			s->tx_tls_ctx               += sq_stats->tls_ctx;
 			s->tx_tls_ooo               += sq_stats->tls_ooo;
+			s->tx_tls_dump_bytes        += sq_stats->tls_dump_bytes;
+			s->tx_tls_dump_packets      += sq_stats->tls_dump_packets;
 			s->tx_tls_resync_bytes      += sq_stats->tls_resync_bytes;
+			s->tx_tls_skip_no_sync_data += sq_stats->tls_skip_no_sync_data;
 			s->tx_tls_drop_no_sync_data += sq_stats->tls_drop_no_sync_data;
 			s->tx_tls_drop_bypass_req   += sq_stats->tls_drop_bypass_req;
-			s->tx_tls_dump_bytes        += sq_stats->tls_dump_bytes;
-			s->tx_tls_dump_packets      += sq_stats->tls_dump_packets;
 #endif
 			s->tx_cqes		+= sq_stats->cqes;
 		}
@@ -1472,10 +1474,12 @@ static const struct counter_desc sq_stats_desc[] = {
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_encrypted_bytes) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_ctx) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_ooo) },
-	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_drop_no_sync_data) },
-	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_drop_bypass_req) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_dump_packets) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_dump_bytes) },
+	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_resync_bytes) },
+	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_skip_no_sync_data) },
+	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_drop_no_sync_data) },
+	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tls_drop_bypass_req) },
 #endif
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, csum_none) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, stopped) },

drivers/net/ethernet/mellanox/mlx5/core/en_stats.h

@@ -129,11 +129,12 @@ struct mlx5e_sw_stats {
 	u64 tx_tls_encrypted_bytes;
 	u64 tx_tls_ctx;
 	u64 tx_tls_ooo;
+	u64 tx_tls_dump_packets;
+	u64 tx_tls_dump_bytes;
 	u64 tx_tls_resync_bytes;
+	u64 tx_tls_skip_no_sync_data;
 	u64 tx_tls_drop_no_sync_data;
 	u64 tx_tls_drop_bypass_req;
-	u64 tx_tls_dump_packets;
-	u64 tx_tls_dump_bytes;
 #endif
 
 	u64 rx_xsk_packets;
@@ -273,11 +274,12 @@ struct mlx5e_sq_stats {
 	u64 tls_encrypted_bytes;
 	u64 tls_ctx;
 	u64 tls_ooo;
+	u64 tls_dump_packets;
+	u64 tls_dump_bytes;
 	u64 tls_resync_bytes;
+	u64 tls_skip_no_sync_data;
 	u64 tls_drop_no_sync_data;
 	u64 tls_drop_bypass_req;
-	u64 tls_dump_packets;
-	u64 tls_dump_bytes;
 #endif
 	/* less likely accessed in data path */
 	u64 csum_none;

drivers/net/ethernet/mellanox/mlx5/core/en_tx.c

@@ -403,7 +403,10 @@ netdev_tx_t mlx5e_xmit(struct sk_buff *skb, struct net_device *dev)
 static void mlx5e_dump_error_cqe(struct mlx5e_txqsq *sq,
 				 struct mlx5_err_cqe *err_cqe)
 {
-	u32 ci = mlx5_cqwq_get_ci(&sq->cq.wq);
+	struct mlx5_cqwq *wq = &sq->cq.wq;
+	u32 ci;
+
+	ci = mlx5_cqwq_ctr2ix(wq, wq->cc - 1);
 
 	netdev_err(sq->channel->netdev,
 		   "Error cqe on cqn 0x%x, ci 0x%x, sqn 0x%x, opcode 0x%x, syndrome 0x%x, vendor syndrome 0x%x\n",
@@ -479,14 +482,7 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq, int napi_budget)
 			skb = wi->skb;
 
 			if (unlikely(!skb)) {
-#ifdef CONFIG_MLX5_EN_TLS
-				if (wi->resync_dump_frag) {
-					struct mlx5e_sq_dma *dma =
-						mlx5e_dma_get(sq, dma_fifo_cc++);
-
-					mlx5e_ktls_tx_handle_resync_dump_comp(sq, wi, dma);
-				}
-#endif
+				mlx5e_ktls_tx_handle_resync_dump_comp(sq, wi, &dma_fifo_cc);
 				sqcc += wi->num_wqebbs;
 				continue;
 			}
@@ -542,29 +538,38 @@ void mlx5e_free_txqsq_descs(struct mlx5e_txqsq *sq)
 {
 	struct mlx5e_tx_wqe_info *wi;
 	struct sk_buff *skb;
+	u32 dma_fifo_cc;
+	u16 sqcc;
 	u16 ci;
 	int i;
 
-	while (sq->cc != sq->pc) {
-		ci = mlx5_wq_cyc_ctr2ix(&sq->wq, sq->cc);
+	sqcc = sq->cc;
+	dma_fifo_cc = sq->dma_fifo_cc;
+
+	while (sqcc != sq->pc) {
+		ci = mlx5_wq_cyc_ctr2ix(&sq->wq, sqcc);
 		wi = &sq->db.wqe_info[ci];
 		skb = wi->skb;
 
-		if (!skb) { /* nop */
-			sq->cc++;
+		if (!skb) {
+			mlx5e_ktls_tx_handle_resync_dump_comp(sq, wi, &dma_fifo_cc);
+			sqcc += wi->num_wqebbs;
 			continue;
 		}
 
 		for (i = 0; i < wi->num_dma; i++) {
 			struct mlx5e_sq_dma *dma =
-				mlx5e_dma_get(sq, sq->dma_fifo_cc++);
+				mlx5e_dma_get(sq, dma_fifo_cc++);
 
 			mlx5e_tx_dma_unmap(sq->pdev, dma);
 		}
 
 		dev_kfree_skb_any(skb);
-		sq->cc += wi->num_wqebbs;
+		sqcc += wi->num_wqebbs;
 	}
+
+	sq->dma_fifo_cc = dma_fifo_cc;
+	sq->cc = sqcc;
 }
 
 #ifdef CONFIG_MLX5_CORE_IPOIB

drivers/net/ethernet/mellanox/mlx5/core/fpga/conn.c

@@ -464,8 +464,10 @@ static int mlx5_fpga_conn_create_cq(struct mlx5_fpga_conn *conn, int cq_size)
 	}
 
 	err = mlx5_vector2eqn(mdev, smp_processor_id(), &eqn, &irqn);
-	if (err)
+	if (err) {
+		kvfree(in);
 		goto err_cqwq;
+	}
 
 	cqc = MLX5_ADDR_OF(create_cq_in, in, cq_context);
 	MLX5_SET(cqc, cqc, log_cq_size, ilog2(cq_size));

drivers/net/ethernet/mellanox/mlx5/core/health.c

@@ -572,7 +572,7 @@ mlx5_fw_fatal_reporter_dump(struct devlink_health_reporter *reporter,
 		return -ENOMEM;
 	err = mlx5_crdump_collect(dev, cr_data);
 	if (err)
-		return err;
+		goto free_data;
 
 	if (priv_ctx) {
 		struct mlx5_fw_reporter_ctx *fw_reporter_ctx = priv_ctx;