From 17d00e839d3b592da9659c1977d45f85b77f986a Mon Sep 17 00:00:00 2001 From: Moshe Shemesh Date: Fri, 27 Dec 2019 07:01:53 +0200 Subject: [PATCH 01/13] net/mlx5: Add command entry handling completion When FW response to commands is very slow and all command entries in use are waiting for completion we can have a race where commands can get timeout before they get out of the queue and handled. Timeout completion on uninitialized command will cause releasing command's buffers before accessing it for initialization and then we will get NULL pointer exception while trying access it. It may also cause releasing buffers of another command since we may have timeout completion before even allocating entry index for this command. Add entry handling completion to avoid this race. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Moshe Shemesh Signed-off-by: Eran Ben Elisha Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 14 ++++++++++++++ include/linux/mlx5/driver.h | 1 + 2 files changed, 15 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index cede5bdfd598..d695b75bc0af 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -861,6 +861,7 @@ static void cmd_work_handler(struct work_struct *work) int alloc_ret; int cmd_mode; + complete(&ent->handling); sem = ent->page_queue ? &cmd->pages_sem : &cmd->sem; down(sem); if (!ent->page_queue) { @@ -978,6 +979,11 @@ static int wait_func(struct mlx5_core_dev *dev, struct mlx5_cmd_work_ent *ent) struct mlx5_cmd *cmd = &dev->cmd; int err; + if (!wait_for_completion_timeout(&ent->handling, timeout) && + cancel_work_sync(&ent->work)) { + ent->ret = -ECANCELED; + goto out_err; + } if (cmd->mode == CMD_MODE_POLLING || ent->polling) { wait_for_completion(&ent->done); } else if (!wait_for_completion_timeout(&ent->done, timeout)) { @@ -985,12 +991,17 @@ static int wait_func(struct mlx5_core_dev *dev, struct mlx5_cmd_work_ent *ent) mlx5_cmd_comp_handler(dev, 1UL << ent->idx, true); } +out_err: err = ent->ret; if (err == -ETIMEDOUT) { mlx5_core_warn(dev, "%s(0x%x) timeout. Will cause a leak of a command resource\n", mlx5_command_str(msg_to_opcode(ent->in)), msg_to_opcode(ent->in)); + } else if (err == -ECANCELED) { + mlx5_core_warn(dev, "%s(0x%x) canceled on out of queue timeout.\n", + mlx5_command_str(msg_to_opcode(ent->in)), + msg_to_opcode(ent->in)); } mlx5_core_dbg(dev, "err %d, delivery status %s(%d)\n", err, deliv_status_to_str(ent->status), ent->status); @@ -1026,6 +1037,7 @@ static int mlx5_cmd_invoke(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *in, ent->token = token; ent->polling = force_polling; + init_completion(&ent->handling); if (!callback) init_completion(&ent->done); @@ -1045,6 +1057,8 @@ static int mlx5_cmd_invoke(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *in, err = wait_func(dev, ent); if (err == -ETIMEDOUT) goto out; + if (err == -ECANCELED) + goto out_free; ds = ent->ts2 - ent->ts1; op = MLX5_GET(mbox_in, in->first.data, opcode); diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index 6f8f79ef829b..9b1f29f26c27 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -743,6 +743,7 @@ struct mlx5_cmd_work_ent { struct delayed_work cb_timeout_work; void *context; int idx; + struct completion handling; struct completion done; struct mlx5_cmd *cmd; struct work_struct work; From d43b7007dbd1195a5b6b83213e49b1516aaf6f5e Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Wed, 18 Mar 2020 21:44:32 +0200 Subject: [PATCH 02/13] net/mlx5: Fix a race when moving command interface to events mode After driver creates (via FW command) an EQ for commands, the driver will be informed on new commands completion by EQE. However, due to a race in driver's internal command mode metadata update, some new commands will still be miss-handled by driver as if we are in polling mode. Such commands can get two non forced completion, leading to already freed command entry access. CREATE_EQ command, that maps EQ to the command queue must be posted to the command queue while it is empty and no other command should be posted. Add SW mechanism that once the CREATE_EQ command is about to be executed, all other commands will return error without being sent to the FW. Allow sending other commands only after successfully changing the driver's internal command mode metadata. We can safely return error to all other commands while creating the command EQ, as all other commands might be sent from the user/application during driver load. Application can rerun them later after driver's load was finished. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha Signed-off-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 35 ++++++++++++++++--- drivers/net/ethernet/mellanox/mlx5/core/eq.c | 3 ++ include/linux/mlx5/driver.h | 6 ++++ 3 files changed, 40 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index d695b75bc0af..2f3cafdc3b1f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -848,6 +848,14 @@ static void free_msg(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *msg); static void mlx5_free_cmd_msg(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *msg); +static bool opcode_allowed(struct mlx5_cmd *cmd, u16 opcode) +{ + if (cmd->allowed_opcode == CMD_ALLOWED_OPCODE_ALL) + return true; + + return cmd->allowed_opcode == opcode; +} + static void cmd_work_handler(struct work_struct *work) { struct mlx5_cmd_work_ent *ent = container_of(work, struct mlx5_cmd_work_ent, work); @@ -914,7 +922,8 @@ static void cmd_work_handler(struct work_struct *work) /* Skip sending command to fw if internal error */ if (pci_channel_offline(dev->pdev) || - dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) { + dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR || + !opcode_allowed(&dev->cmd, ent->op)) { u8 status = 0; u32 drv_synd; @@ -1405,6 +1414,22 @@ static void create_debugfs_files(struct mlx5_core_dev *dev) mlx5_cmdif_debugfs_init(dev); } +void mlx5_cmd_allowed_opcode(struct mlx5_core_dev *dev, u16 opcode) +{ + struct mlx5_cmd *cmd = &dev->cmd; + int i; + + for (i = 0; i < cmd->max_reg_cmds; i++) + down(&cmd->sem); + down(&cmd->pages_sem); + + cmd->allowed_opcode = opcode; + + up(&cmd->pages_sem); + for (i = 0; i < cmd->max_reg_cmds; i++) + up(&cmd->sem); +} + static void mlx5_cmd_change_mod(struct mlx5_core_dev *dev, int mode) { struct mlx5_cmd *cmd = &dev->cmd; @@ -1681,12 +1706,13 @@ static int cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out, int err; u8 status = 0; u32 drv_synd; + u16 opcode; u8 token; + opcode = MLX5_GET(mbox_in, in, opcode); if (pci_channel_offline(dev->pdev) || - dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) { - u16 opcode = MLX5_GET(mbox_in, in, opcode); - + dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR || + !opcode_allowed(&dev->cmd, opcode)) { err = mlx5_internal_err_ret_value(dev, opcode, &drv_synd, &status); MLX5_SET(mbox_out, out, status, status); MLX5_SET(mbox_out, out, syndrome, drv_synd); @@ -1988,6 +2014,7 @@ int mlx5_cmd_init(struct mlx5_core_dev *dev) mlx5_core_dbg(dev, "descriptor at dma 0x%llx\n", (unsigned long long)(cmd->dma)); cmd->mode = CMD_MODE_POLLING; + cmd->allowed_opcode = CMD_ALLOWED_OPCODE_ALL; create_msg_cache(dev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index cccea3a8eddd..ce6c621af043 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -611,11 +611,13 @@ static int create_async_eqs(struct mlx5_core_dev *dev) .nent = MLX5_NUM_CMD_EQE, .mask[0] = 1ull << MLX5_EVENT_TYPE_CMD, }; + mlx5_cmd_allowed_opcode(dev, MLX5_CMD_OP_CREATE_EQ); err = setup_async_eq(dev, &table->cmd_eq, ¶m, "cmd"); if (err) goto err1; mlx5_cmd_use_events(dev); + mlx5_cmd_allowed_opcode(dev, CMD_ALLOWED_OPCODE_ALL); param = (struct mlx5_eq_param) { .irq_index = 0, @@ -645,6 +647,7 @@ err2: mlx5_cmd_use_polling(dev); cleanup_async_eq(dev, &table->cmd_eq, "cmd"); err1: + mlx5_cmd_allowed_opcode(dev, CMD_ALLOWED_OPCODE_ALL); mlx5_eq_notifier_unregister(dev, &table->cq_err_nb); return err; } diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index 9b1f29f26c27..c03778c75dfa 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -284,6 +284,7 @@ struct mlx5_cmd { struct semaphore sem; struct semaphore pages_sem; int mode; + u16 allowed_opcode; struct mlx5_cmd_work_ent *ent_arr[MLX5_MAX_COMMANDS]; struct dma_pool *pool; struct mlx5_cmd_debug dbg; @@ -875,10 +876,15 @@ mlx5_frag_buf_get_idx_last_contig_stride(struct mlx5_frag_buf_ctrl *fbc, u32 ix) return min_t(u32, last_frag_stride_idx - fbc->strides_offset, fbc->sz_m1); } +enum { + CMD_ALLOWED_OPCODE_ALL, +}; + int mlx5_cmd_init(struct mlx5_core_dev *dev); void mlx5_cmd_cleanup(struct mlx5_core_dev *dev); void mlx5_cmd_use_events(struct mlx5_core_dev *dev); void mlx5_cmd_use_polling(struct mlx5_core_dev *dev); +void mlx5_cmd_allowed_opcode(struct mlx5_core_dev *dev, u16 opcode); struct mlx5_async_ctx { struct mlx5_core_dev *dev; From f7936ddd35d8b849daf0372770c7c9dbe7910fca Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Thu, 19 Mar 2020 21:43:13 +0200 Subject: [PATCH 03/13] net/mlx5: Avoid processing commands before cmdif is ready When driver is reloading during recovery flow, it can't get new commands till command interface is up again. Otherwise we may get to null pointer trying to access non initialized command structures. Add cmdif state to avoid processing commands while cmdif is not ready. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha Signed-off-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 10 ++++++++++ drivers/net/ethernet/mellanox/mlx5/core/main.c | 4 ++++ include/linux/mlx5/driver.h | 9 +++++++++ 3 files changed, 23 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index 2f3cafdc3b1f..7a77fe40af3a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -923,6 +923,7 @@ static void cmd_work_handler(struct work_struct *work) /* Skip sending command to fw if internal error */ if (pci_channel_offline(dev->pdev) || dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR || + cmd->state != MLX5_CMDIF_STATE_UP || !opcode_allowed(&dev->cmd, ent->op)) { u8 status = 0; u32 drv_synd; @@ -1712,6 +1713,7 @@ static int cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out, opcode = MLX5_GET(mbox_in, in, opcode); if (pci_channel_offline(dev->pdev) || dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR || + dev->cmd.state != MLX5_CMDIF_STATE_UP || !opcode_allowed(&dev->cmd, opcode)) { err = mlx5_internal_err_ret_value(dev, opcode, &drv_synd, &status); MLX5_SET(mbox_out, out, status, status); @@ -1977,6 +1979,7 @@ int mlx5_cmd_init(struct mlx5_core_dev *dev) goto err_free_page; } + cmd->state = MLX5_CMDIF_STATE_DOWN; cmd->checksum_disabled = 1; cmd->max_reg_cmds = (1 << cmd->log_sz) - 1; cmd->bitmask = (1UL << cmd->max_reg_cmds) - 1; @@ -2054,3 +2057,10 @@ void mlx5_cmd_cleanup(struct mlx5_core_dev *dev) dma_pool_destroy(cmd->pool); } EXPORT_SYMBOL(mlx5_cmd_cleanup); + +void mlx5_cmd_set_state(struct mlx5_core_dev *dev, + enum mlx5_cmdif_state cmdif_state) +{ + dev->cmd.state = cmdif_state; +} +EXPORT_SYMBOL(mlx5_cmd_set_state); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index 7af4210c1b96..a61e473db7e1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -965,6 +965,8 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot) goto err_cmd_cleanup; } + mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_UP); + err = mlx5_core_enable_hca(dev, 0); if (err) { mlx5_core_err(dev, "enable hca failed\n"); @@ -1026,6 +1028,7 @@ reclaim_boot_pages: err_disable_hca: mlx5_core_disable_hca(dev, 0); err_cmd_cleanup: + mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN); mlx5_cmd_cleanup(dev); return err; @@ -1043,6 +1046,7 @@ static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot) } mlx5_reclaim_startup_pages(dev); mlx5_core_disable_hca(dev, 0); + mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN); mlx5_cmd_cleanup(dev); return 0; diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index c03778c75dfa..8397b6558dc7 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -213,6 +213,12 @@ enum mlx5_port_status { MLX5_PORT_DOWN = 2, }; +enum mlx5_cmdif_state { + MLX5_CMDIF_STATE_UNINITIALIZED, + MLX5_CMDIF_STATE_UP, + MLX5_CMDIF_STATE_DOWN, +}; + struct mlx5_cmd_first { __be32 data[4]; }; @@ -258,6 +264,7 @@ struct mlx5_cmd_stats { struct mlx5_cmd { struct mlx5_nb nb; + enum mlx5_cmdif_state state; void *cmd_alloc_buf; dma_addr_t alloc_dma; int alloc_size; @@ -882,6 +889,8 @@ enum { int mlx5_cmd_init(struct mlx5_core_dev *dev); void mlx5_cmd_cleanup(struct mlx5_core_dev *dev); +void mlx5_cmd_set_state(struct mlx5_core_dev *dev, + enum mlx5_cmdif_state cmdif_state); void mlx5_cmd_use_events(struct mlx5_core_dev *dev); void mlx5_cmd_use_polling(struct mlx5_core_dev *dev); void mlx5_cmd_allowed_opcode(struct mlx5_core_dev *dev, u16 opcode); From 321348475d544aa6705dcfac2135deeccb8dc0bb Mon Sep 17 00:00:00 2001 From: Maor Dickman Date: Thu, 23 Apr 2020 15:16:17 +0300 Subject: [PATCH 04/13] net/mlx5e: Fix allowed tc redirect merged eswitch offload cases After changing the parent_id to be the same for both NICs of same The cited commit wrongly allow offload of tc redirect flows from VF to uplink and vice versa when devcies are on different eswitch, these cases aren't supported by HW. Disallow the above offloads when devcies are on different eswitch and VF LAG is not configured. Fixes: f6dc1264f1c0 ("net/mlx5e: Disallow tc redirect offload cases we don't support") Signed-off-by: Maor Dickman Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_rep.c | 8 +--- .../net/ethernet/mellanox/mlx5/core/en_rep.h | 7 +++- .../net/ethernet/mellanox/mlx5/core/en_tc.c | 40 +++++++++++++++---- 3 files changed, 41 insertions(+), 14 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index f372e94948fd..cdecf4280e86 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -1484,13 +1484,9 @@ bool mlx5e_eswitch_uplink_rep(struct net_device *netdev) return netdev->netdev_ops == &mlx5e_netdev_ops_uplink_rep; } -bool mlx5e_eswitch_rep(struct net_device *netdev) +bool mlx5e_eswitch_vf_rep(struct net_device *netdev) { - if (netdev->netdev_ops == &mlx5e_netdev_ops_rep || - netdev->netdev_ops == &mlx5e_netdev_ops_uplink_rep) - return true; - - return false; + return netdev->netdev_ops == &mlx5e_netdev_ops_rep; } static void mlx5e_build_rep_params(struct net_device *netdev) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index 6a2337900420..612b5cf0673d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -210,8 +210,13 @@ void mlx5e_rep_encap_entry_detach(struct mlx5e_priv *priv, void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv); -bool mlx5e_eswitch_rep(struct net_device *netdev); +bool mlx5e_eswitch_vf_rep(struct net_device *netdev); bool mlx5e_eswitch_uplink_rep(struct net_device *netdev); +static inline bool mlx5e_eswitch_rep(struct net_device *netdev) +{ + return mlx5e_eswitch_vf_rep(netdev) || + mlx5e_eswitch_uplink_rep(netdev); +} #else /* CONFIG_MLX5_ESWITCH */ static inline bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv) { return false; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index a574c588269a..5bcf95fcdd59 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -3073,6 +3073,11 @@ static bool actions_match_supported(struct mlx5e_priv *priv, return true; } +static bool same_port_devs(struct mlx5e_priv *priv, struct mlx5e_priv *peer_priv) +{ + return priv->mdev == peer_priv->mdev; +} + static bool same_hw_devs(struct mlx5e_priv *priv, struct mlx5e_priv *peer_priv) { struct mlx5_core_dev *fmdev, *pmdev; @@ -3291,7 +3296,7 @@ static inline int hash_encap_info(struct encap_key *key) } -static bool is_merged_eswitch_dev(struct mlx5e_priv *priv, +static bool is_merged_eswitch_vfs(struct mlx5e_priv *priv, struct net_device *peer_netdev) { struct mlx5e_priv *peer_priv; @@ -3299,13 +3304,11 @@ static bool is_merged_eswitch_dev(struct mlx5e_priv *priv, peer_priv = netdev_priv(peer_netdev); return (MLX5_CAP_ESW(priv->mdev, merged_eswitch) && - mlx5e_eswitch_rep(priv->netdev) && - mlx5e_eswitch_rep(peer_netdev) && + mlx5e_eswitch_vf_rep(priv->netdev) && + mlx5e_eswitch_vf_rep(peer_netdev) && same_hw_devs(priv, peer_priv)); } - - bool mlx5e_encap_take(struct mlx5e_encap_entry *e) { return refcount_inc_not_zero(&e->refcnt); @@ -3575,14 +3578,37 @@ static int add_vlan_pop_action(struct mlx5e_priv *priv, return err; } +static bool same_hw_reps(struct mlx5e_priv *priv, + struct net_device *peer_netdev) +{ + struct mlx5e_priv *peer_priv; + + peer_priv = netdev_priv(peer_netdev); + + return mlx5e_eswitch_rep(priv->netdev) && + mlx5e_eswitch_rep(peer_netdev) && + same_hw_devs(priv, peer_priv); +} + +static bool is_lag_dev(struct mlx5e_priv *priv, + struct net_device *peer_netdev) +{ + return ((mlx5_lag_is_sriov(priv->mdev) || + mlx5_lag_is_multipath(priv->mdev)) && + same_hw_reps(priv, peer_netdev)); +} + bool mlx5e_is_valid_eswitch_fwd_dev(struct mlx5e_priv *priv, struct net_device *out_dev) { - if (is_merged_eswitch_dev(priv, out_dev)) + if (is_merged_eswitch_vfs(priv, out_dev)) + return true; + + if (is_lag_dev(priv, out_dev)) return true; return mlx5e_eswitch_rep(out_dev) && - same_hw_devs(priv, netdev_priv(out_dev)); + same_port_devs(priv, netdev_priv(out_dev)); } static bool is_duplicated_output_device(struct net_device *dev, From 16736e11f43b80a38f98f6add54fab3b8c297df3 Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Mon, 27 Apr 2020 16:56:59 +0300 Subject: [PATCH 05/13] net/mlx5e: kTLS, Destroy key object after destroying the TIS The TLS TIS object contains the dek/key ID. By destroying the key first, the TIS would contain an invalid non-existing key ID. Reverse the destroy order, this also acheives the desired assymetry between the destroy and the create flows. Fixes: d2ead1f360e8 ("net/mlx5e: Add kTLS TX HW offload support") Signed-off-by: Tariq Toukan Reviewed-by: Boris Pismenny Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c index 46725cd743a3..7d1985fa0d4f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c @@ -69,8 +69,8 @@ static void mlx5e_ktls_del(struct net_device *netdev, struct mlx5e_ktls_offload_context_tx *tx_priv = mlx5e_get_ktls_tx_priv_ctx(tls_ctx); - mlx5_ktls_destroy_key(priv->mdev, tx_priv->key_id); mlx5e_destroy_tis(priv->mdev, tx_priv->tisn); + mlx5_ktls_destroy_key(priv->mdev, tx_priv->key_id); kvfree(tx_priv); } From a16b8e0dcf7043bee46174bed0553cc9e36b63a5 Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Thu, 30 Apr 2020 09:16:01 +0300 Subject: [PATCH 06/13] net/mlx5e: Fix inner tirs handling In the cited commit inner_tirs argument was added to create and destroy inner tirs, and no indication was added to mlx5e_modify_tirs_hash() function. In order to have a consistent handling, use inner_indir_tir[0].tirn in tirs destroy/modify function as an indication to whether inner tirs are created. Inner tirs are not created for representors and before this commit, a call to mlx5e_modify_tirs_hash() was sending HW commands to modify non-existent inner tirs. Fixes: 46dc933cee82 ("net/mlx5e: Provide explicit directive if to create inner indirect tirs") Signed-off-by: Roi Dayan Reviewed-by: Vlad Buslov Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 2 +- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 12 +++++++----- drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 4 ++-- .../net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c | 4 ++-- 4 files changed, 12 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 23701c0e36ec..59745402747b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -1121,7 +1121,7 @@ void mlx5e_close_drop_rq(struct mlx5e_rq *drop_rq); int mlx5e_create_indirect_rqt(struct mlx5e_priv *priv); int mlx5e_create_indirect_tirs(struct mlx5e_priv *priv, bool inner_ttc); -void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv, bool inner_ttc); +void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv); int mlx5e_create_direct_rqts(struct mlx5e_priv *priv, struct mlx5e_tir *tirs); void mlx5e_destroy_direct_rqts(struct mlx5e_priv *priv, struct mlx5e_tir *tirs); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index b314adf438da..c6b83042d431 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -2717,7 +2717,8 @@ void mlx5e_modify_tirs_hash(struct mlx5e_priv *priv, void *in, int inlen) mlx5_core_modify_tir(mdev, priv->indir_tir[tt].tirn, in, inlen); } - if (!mlx5e_tunnel_inner_ft_supported(priv->mdev)) + /* Verify inner tirs resources allocated */ + if (!priv->inner_indir_tir[0].tirn) return; for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) { @@ -3408,14 +3409,15 @@ out: return err; } -void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv, bool inner_ttc) +void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv) { int i; for (i = 0; i < MLX5E_NUM_INDIR_TIRS; i++) mlx5e_destroy_tir(priv->mdev, &priv->indir_tir[i]); - if (!inner_ttc || !mlx5e_tunnel_inner_ft_supported(priv->mdev)) + /* Verify inner tirs resources allocated */ + if (!priv->inner_indir_tir[0].tirn) return; for (i = 0; i < MLX5E_NUM_INDIR_TIRS; i++) @@ -5123,7 +5125,7 @@ err_destroy_xsk_rqts: err_destroy_direct_tirs: mlx5e_destroy_direct_tirs(priv, priv->direct_tir); err_destroy_indirect_tirs: - mlx5e_destroy_indirect_tirs(priv, true); + mlx5e_destroy_indirect_tirs(priv); err_destroy_direct_rqts: mlx5e_destroy_direct_rqts(priv, priv->direct_tir); err_destroy_indirect_rqts: @@ -5142,7 +5144,7 @@ static void mlx5e_cleanup_nic_rx(struct mlx5e_priv *priv) mlx5e_destroy_direct_tirs(priv, priv->xsk_tir); mlx5e_destroy_direct_rqts(priv, priv->xsk_tir); mlx5e_destroy_direct_tirs(priv, priv->direct_tir); - mlx5e_destroy_indirect_tirs(priv, true); + mlx5e_destroy_indirect_tirs(priv); mlx5e_destroy_direct_rqts(priv, priv->direct_tir); mlx5e_destroy_rqt(priv, &priv->indir_rqt); mlx5e_close_drop_rq(&priv->drop_rq); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index cdecf4280e86..4a8e0dfdc5f2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -1743,7 +1743,7 @@ err_destroy_ttc_table: err_destroy_direct_tirs: mlx5e_destroy_direct_tirs(priv, priv->direct_tir); err_destroy_indirect_tirs: - mlx5e_destroy_indirect_tirs(priv, false); + mlx5e_destroy_indirect_tirs(priv); err_destroy_direct_rqts: mlx5e_destroy_direct_rqts(priv, priv->direct_tir); err_destroy_indirect_rqts: @@ -1761,7 +1761,7 @@ static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv) mlx5e_destroy_rep_root_ft(priv); mlx5e_destroy_ttc_table(priv, &priv->fs.ttc); mlx5e_destroy_direct_tirs(priv, priv->direct_tir); - mlx5e_destroy_indirect_tirs(priv, false); + mlx5e_destroy_indirect_tirs(priv); mlx5e_destroy_direct_rqts(priv, priv->direct_tir); mlx5e_destroy_rqt(priv, &priv->indir_rqt); mlx5e_close_drop_rq(&priv->drop_rq); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c index 673aaa815f57..505cf6eeae25 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c @@ -396,7 +396,7 @@ static int mlx5i_init_rx(struct mlx5e_priv *priv) err_destroy_direct_tirs: mlx5e_destroy_direct_tirs(priv, priv->direct_tir); err_destroy_indirect_tirs: - mlx5e_destroy_indirect_tirs(priv, true); + mlx5e_destroy_indirect_tirs(priv); err_destroy_direct_rqts: mlx5e_destroy_direct_rqts(priv, priv->direct_tir); err_destroy_indirect_rqts: @@ -412,7 +412,7 @@ static void mlx5i_cleanup_rx(struct mlx5e_priv *priv) { mlx5i_destroy_flow_steering(priv); mlx5e_destroy_direct_tirs(priv, priv->direct_tir); - mlx5e_destroy_indirect_tirs(priv, true); + mlx5e_destroy_indirect_tirs(priv); mlx5e_destroy_direct_rqts(priv, priv->direct_tir); mlx5e_destroy_rqt(priv, &priv->indir_rqt); mlx5e_close_drop_rq(&priv->drop_rq); From df14ad1eccb04a4a28c90389214dbacab085b244 Mon Sep 17 00:00:00 2001 From: Moshe Shemesh Date: Wed, 29 Apr 2020 23:56:58 +0300 Subject: [PATCH 07/13] net/mlx5: Fix memory leak in mlx5_events_init Fix memory leak in mlx5_events_init(), in case create_single_thread_workqueue() fails, events struct should be freed. Fixes: 5d3c537f9070 ("net/mlx5: Handle event of power detection in the PCIE slot") Signed-off-by: Moshe Shemesh Reviewed-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/events.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/events.c b/drivers/net/ethernet/mellanox/mlx5/core/events.c index 8bcf3426b9c6..3ce17c3d7a00 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/events.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/events.c @@ -346,8 +346,10 @@ int mlx5_events_init(struct mlx5_core_dev *dev) events->dev = dev; dev->priv.events = events; events->wq = create_singlethread_workqueue("mlx5_events"); - if (!events->wq) + if (!events->wq) { + kfree(events); return -ENOMEM; + } INIT_WORK(&events->pcie_core_work, mlx5_pcie_event); return 0; From aee37f3d940ca732df71c3df49347bccaafc0b24 Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Mon, 11 May 2020 16:32:09 +0300 Subject: [PATCH 08/13] net/mlx5: Fix cleaning unmanaged flow tables Unmanaged flow tables doesn't have a parent and tree_put_node() assume there is always a parent if cleaning is needed. fix that. Fixes: 5281a0c90919 ("net/mlx5: fs_core: Introduce unmanaged flow tables") Signed-off-by: Roi Dayan Reviewed-by: Mark Bloch Reviewed-by: Paul Blakey Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index d5defe09339a..8f62bfcf57af 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -344,14 +344,13 @@ static void tree_put_node(struct fs_node *node, bool locked) if (node->del_hw_func) node->del_hw_func(node); if (parent_node) { - /* Only root namespace doesn't have parent and we just - * need to free its node. - */ down_write_ref_node(parent_node, locked); list_del_init(&node->list); if (node->del_sw_func) node->del_sw_func(node); up_write_ref_node(parent_node, locked); + } else if (node->del_sw_func) { + node->del_sw_func(node); } else { kfree(node); } @@ -468,8 +467,10 @@ static void del_sw_flow_table(struct fs_node *node) fs_get_obj(ft, node); rhltable_destroy(&ft->fgs_hash); - fs_get_obj(prio, ft->node.parent); - prio->num_ft--; + if (ft->node.parent) { + fs_get_obj(prio, ft->node.parent); + prio->num_ft--; + } kfree(ft); } From 6eb7a268a99bad8346d4baa148a14456d061c1c3 Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Mon, 11 May 2020 16:37:11 +0300 Subject: [PATCH 09/13] net/mlx5: Don't maintain a case of del_sw_func being null Add del_sw_func cb for root ns. Now there is no need to maintain a case of del_sw_func being null when freeing the node. Fixes: 2cc43b494a6c ("net/mlx5_core: Managing root flow table") Signed-off-by: Roi Dayan Reviewed-by: Mark Bloch Reviewed-by: Paul Blakey Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/fs_core.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 8f62bfcf57af..02d0f94eaaad 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -346,14 +346,10 @@ static void tree_put_node(struct fs_node *node, bool locked) if (parent_node) { down_write_ref_node(parent_node, locked); list_del_init(&node->list); - if (node->del_sw_func) - node->del_sw_func(node); - up_write_ref_node(parent_node, locked); - } else if (node->del_sw_func) { - node->del_sw_func(node); - } else { - kfree(node); } + node->del_sw_func(node); + if (parent_node) + up_write_ref_node(parent_node, locked); node = NULL; } if (!node && parent_node) @@ -2352,6 +2348,11 @@ static int init_root_tree(struct mlx5_flow_steering *steering, return 0; } +static void del_sw_root_ns(struct fs_node *node) +{ + kfree(node); +} + static struct mlx5_flow_root_namespace *create_root_ns(struct mlx5_flow_steering *steering, enum fs_flow_table_type table_type) @@ -2378,7 +2379,7 @@ static struct mlx5_flow_root_namespace ns = &root_ns->ns; fs_init_namespace(ns); mutex_init(&root_ns->chain_lock); - tree_init_node(&ns->node, NULL, NULL); + tree_init_node(&ns->node, NULL, del_sw_root_ns); tree_add_node(&ns->node, NULL); return root_ns; From 9ca415399dae133b00273a4283ef31d003a6818d Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Thu, 14 May 2020 23:44:38 +0300 Subject: [PATCH 10/13] net/mlx5: Annotate mutex destroy for root ns Invoke mutex_destroy() to catch any errors. Fixes: 2cc43b494a6c ("net/mlx5_core: Managing root flow table") Signed-off-by: Roi Dayan Reviewed-by: Mark Bloch Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 02d0f94eaaad..9620c8650e13 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -2350,6 +2350,12 @@ static int init_root_tree(struct mlx5_flow_steering *steering, static void del_sw_root_ns(struct fs_node *node) { + struct mlx5_flow_root_namespace *root_ns; + struct mlx5_flow_namespace *ns; + + fs_get_obj(ns, node); + root_ns = container_of(ns, struct mlx5_flow_root_namespace, ns); + mutex_destroy(&root_ns->chain_lock); kfree(node); } From 5e911e2c06bd8c17df29147a5e2d4b17fafda024 Mon Sep 17 00:00:00 2001 From: Moshe Shemesh Date: Tue, 7 Apr 2020 17:38:28 +0300 Subject: [PATCH 11/13] net/mlx5e: Update netdev txq on completions during closure On sq closure when we free its descriptors, we should also update netdev txq on completions which would not arrive. Otherwise if we reopen sqs and attach them back, for example on fw fatal recovery flow, we may get tx timeout. Fixes: 29429f3300a3 ("net/mlx5e: Timeout if SQ doesn't flush during close") Signed-off-by: Moshe Shemesh Reviewed-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_tx.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c index fd6b2a1898c5..119a5c6cc167 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c @@ -537,10 +537,9 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq, int napi_budget) void mlx5e_free_txqsq_descs(struct mlx5e_txqsq *sq) { struct mlx5e_tx_wqe_info *wi; + u32 dma_fifo_cc, nbytes = 0; + u16 ci, sqcc, npkts = 0; struct sk_buff *skb; - u32 dma_fifo_cc; - u16 sqcc; - u16 ci; int i; sqcc = sq->cc; @@ -565,11 +564,15 @@ void mlx5e_free_txqsq_descs(struct mlx5e_txqsq *sq) } dev_kfree_skb_any(skb); + npkts++; + nbytes += wi->num_bytes; sqcc += wi->num_wqebbs; } sq->dma_fifo_cc = dma_fifo_cc; sq->cc = sqcc; + + netdev_tx_completed_queue(sq->txq, npkts, nbytes); } #ifdef CONFIG_MLX5_CORE_IPOIB From d37bd5e81ed0d58f0ebe2e01658c26722e0c033e Mon Sep 17 00:00:00 2001 From: Roi Dayan Date: Mon, 18 May 2020 20:21:11 +0300 Subject: [PATCH 12/13] net/mlx5e: CT: Correctly get flow rule The correct way is to us the flow_cls_offload_flow_rule() wrapper instead of f->rule directly. Fixes: 4c3844d9e97e ("net/mlx5e: CT: Introduce connection tracking") Signed-off-by: Roi Dayan Reviewed-by: Oz Shlomo Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 5 +++-- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.h | 4 +++- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index a172c5e39710..4eb305af0106 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -699,6 +699,7 @@ mlx5_tc_ct_parse_match(struct mlx5e_priv *priv, struct netlink_ext_ack *extack) { struct mlx5_tc_ct_priv *ct_priv = mlx5_tc_ct_get_ct_priv(priv); + struct flow_rule *rule = flow_cls_offload_flow_rule(f); struct flow_dissector_key_ct *mask, *key; bool trk, est, untrk, unest, new; u32 ctstate = 0, ctstate_mask = 0; @@ -706,7 +707,7 @@ mlx5_tc_ct_parse_match(struct mlx5e_priv *priv, u16 ct_state, ct_state_mask; struct flow_match_ct match; - if (!flow_rule_match_key(f->rule, FLOW_DISSECTOR_KEY_CT)) + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_CT)) return 0; if (!ct_priv) { @@ -715,7 +716,7 @@ mlx5_tc_ct_parse_match(struct mlx5e_priv *priv, return -EOPNOTSUPP; } - flow_rule_match_ct(f->rule, &match); + flow_rule_match_ct(rule, &match); key = match.key; mask = match.mask; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.h index 091d305b633e..626f6c04882e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.h @@ -130,7 +130,9 @@ mlx5_tc_ct_parse_match(struct mlx5e_priv *priv, struct flow_cls_offload *f, struct netlink_ext_ack *extack) { - if (!flow_rule_match_key(f->rule, FLOW_DISSECTOR_KEY_CT)) + struct flow_rule *rule = flow_cls_offload_flow_rule(f); + + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_CT)) return 0; NL_SET_ERR_MSG_MOD(extack, "mlx5 tc ct offload isn't enabled."); From 4f7400d5cbaef676e00cdffb0565bf731c6bb09e Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Wed, 6 May 2020 14:52:04 +0300 Subject: [PATCH 13/13] net/mlx5: Fix error flow in case of function_setup failure Currently, if an error occurred during mlx5_function_setup(), we keep dev->state as DEVICE_STATE_UP. Fixing it by adding a goto label. Fixes: e161105e58da ("net/mlx5: Function setup/teardown procedures") Signed-off-by: Shay Drory Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/main.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index a61e473db7e1..c1618b818f3a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -1195,7 +1195,7 @@ int mlx5_load_one(struct mlx5_core_dev *dev, bool boot) err = mlx5_function_setup(dev, boot); if (err) - goto out; + goto err_function; if (boot) { err = mlx5_init_once(dev); @@ -1233,6 +1233,7 @@ err_load: mlx5_cleanup_once(dev); function_teardown: mlx5_function_teardown(dev, boot); +err_function: dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR; mutex_unlock(&dev->intf_state_mutex);