From 43c075959de3c45608636d9d80ff9e61d166fb21 Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Wed, 26 Jan 2022 10:57:07 -0800
Subject: mlx5: remove unused static inlines

mlx5 has some unused static inline helpers in include/
while at it also clean static inlines in the driver itself.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 include/linux/mlx5/driver.h | 10 ----------
 1 file changed, 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 78655d8d13a7..1b398c9e17b9 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -863,20 +863,10 @@ struct mlx5_hca_vport_context {
 	bool			grh_required;
 };
 
-static inline void *mlx5_buf_offset(struct mlx5_frag_buf *buf, int offset)
-{
-		return buf->frags->buf + offset;
-}
-
 #define STRUCT_FIELD(header, field) \
 	.struct_offset_bytes = offsetof(struct ib_unpacked_ ## header, field),      \
 	.struct_size_bytes   = sizeof((struct ib_unpacked_ ## header *)0)->field
 
-static inline struct mlx5_core_dev *pci2mlx5_core_dev(struct pci_dev *pdev)
-{
-	return pci_get_drvdata(pdev);
-}
-
 extern struct dentry *mlx5_debugfs_root;
 
 static inline u16 fw_rev_maj(struct mlx5_core_dev *dev)
-- 
cgit v1.2.3


From c2c922dae77f36e24d246c6e310cee0c61afc6fb Mon Sep 17 00:00:00 2001
From: Mark Bloch <mbloch@nvidia.com>
Date: Mon, 29 Nov 2021 16:24:28 +0200
Subject: net/mlx5: Add ability to insert to specific flow group

If the flow table isn't an autogroup the upper driver has to create the
flow groups explicitly. This information can't later be used when
creating rules to insert into a specific flow group. Allow such use case.

Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 9 ++++++++-
 include/linux/mlx5/fs.h                           | 1 +
 2 files changed, 9 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index b628917e38e4..ebb7960ec62b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -1696,6 +1696,7 @@ static void free_match_list(struct match_list *head, bool ft_locked)
 static int build_match_list(struct match_list *match_head,
 			    struct mlx5_flow_table *ft,
 			    const struct mlx5_flow_spec *spec,
+			    struct mlx5_flow_group *fg,
 			    bool ft_locked)
 {
 	struct rhlist_head *tmp, *list;
@@ -1710,6 +1711,9 @@ static int build_match_list(struct match_list *match_head,
 	rhl_for_each_entry_rcu(g, tmp, list, hash) {
 		struct match_list *curr_match;
 
+		if (fg && fg != g)
+			continue;
+
 		if (unlikely(!tree_get_node(&g->node)))
 			continue;
 
@@ -1889,6 +1893,9 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft,
 	if (!check_valid_spec(spec))
 		return ERR_PTR(-EINVAL);
 
+	if (flow_act->fg && ft->autogroup.active)
+		return ERR_PTR(-EINVAL);
+
 	for (i = 0; i < dest_num; i++) {
 		if (!dest_is_valid(&dest[i], flow_act, ft))
 			return ERR_PTR(-EINVAL);
@@ -1898,7 +1905,7 @@ search_again_locked:
 	version = atomic_read(&ft->node.version);
 
 	/* Collect all fgs which has a matching match_criteria */
-	err = build_match_list(&match_head, ft, spec, take_write);
+	err = build_match_list(&match_head, ft, spec, flow_act->fg, take_write);
 	if (err) {
 		if (take_write)
 			up_write_ref_node(&ft->node, false);
diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h
index b1aad14689e3..e3bfed68b08a 100644
--- a/include/linux/mlx5/fs.h
+++ b/include/linux/mlx5/fs.h
@@ -224,6 +224,7 @@ struct mlx5_flow_act {
 	u32 flags;
 	struct mlx5_fs_vlan vlan[MLX5_FS_VLAN_DEPTH];
 	struct ib_counters *counters;
+	struct mlx5_flow_group *fg;
 };
 
 #define MLX5_DECLARE_FLOW_ACT(name) \
-- 
cgit v1.2.3


From 605bef0015b163867127202b821dce79804d603d Mon Sep 17 00:00:00 2001
From: Saeed Mahameed <saeedm@mellanox.com>
Date: Sun, 5 Apr 2020 17:50:25 -0700
Subject: net/mlx5: cmdif, cmd_check refactoring

Do not mangle the command outbox in the internal low level cmd_exec and
cmd_invoke functions.

Instead return a proper unique error code and move the driver error
checking to be at a higher level in mlx5_cmd_exec().

Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/cmd.c  | 173 +++++++++++++------------
 drivers/net/ethernet/mellanox/mlx5/core/main.c |   5 +-
 include/linux/mlx5/driver.h                    |   1 -
 3 files changed, 95 insertions(+), 84 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 3c6a533ee0c9..7ff01b901f53 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -760,43 +760,61 @@ struct mlx5_ifc_mbox_in_bits {
 	u8         reserved_at_40[0x40];
 };
 
-void mlx5_cmd_mbox_status(void *out, u8 *status, u32 *syndrome)
-{
-	*status = MLX5_GET(mbox_out, out, status);
-	*syndrome = MLX5_GET(mbox_out, out, syndrome);
-}
-
-static int mlx5_cmd_check(struct mlx5_core_dev *dev, void *in, void *out)
+static void cmd_status_print(struct mlx5_core_dev *dev, void *in, void *out)
 {
+	u16 opcode, op_mod;
 	u32 syndrome;
 	u8  status;
-	u16 opcode;
-	u16 op_mod;
 	u16 uid;
+	int err;
 
-	mlx5_cmd_mbox_status(out, &status, &syndrome);
-	if (!status)
-		return 0;
+	syndrome = MLX5_GET(mbox_out, out, syndrome);
+	status = MLX5_GET(mbox_out, out, status);
 
 	opcode = MLX5_GET(mbox_in, in, opcode);
 	op_mod = MLX5_GET(mbox_in, in, op_mod);
 	uid    = MLX5_GET(mbox_in, in, uid);
 
+	err = cmd_status_to_err(status);
+
 	if (!uid && opcode != MLX5_CMD_OP_DESTROY_MKEY)
 		mlx5_core_err_rl(dev,
-			"%s(0x%x) op_mod(0x%x) failed, status %s(0x%x), syndrome (0x%x)\n",
+			"%s(0x%x) op_mod(0x%x) failed, status %s(0x%x), syndrome (0x%x), err(%d)\n",
 			mlx5_command_str(opcode), opcode, op_mod,
-			cmd_status_str(status), status, syndrome);
+			cmd_status_str(status), status, syndrome, err);
 	else
 		mlx5_core_dbg(dev,
-		      "%s(0x%x) op_mod(0x%x) failed, status %s(0x%x), syndrome (0x%x)\n",
-		      mlx5_command_str(opcode),
-		      opcode, op_mod,
-		      cmd_status_str(status),
-		      status,
-		      syndrome);
+			"%s(0x%x) op_mod(0x%x) uid(%d) failed, status %s(0x%x), syndrome (0x%x), err(%d)\n",
+			mlx5_command_str(opcode), opcode, op_mod, uid,
+			cmd_status_str(status), status, syndrome, err);
+}
+
+static int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *out)
+{
+	/* aborted due to PCI error or via reset flow mlx5_cmd_trigger_completions() */
+	if (err == -ENXIO) {
+		u16 opcode = MLX5_GET(mbox_in, in, opcode);
+		u32 syndrome;
+		u8 status;
+
+		/* PCI Error, emulate command return status, for smooth reset */
+		err = mlx5_internal_err_ret_value(dev, opcode, &syndrome, &status);
+		MLX5_SET(mbox_out, out, status, status);
+		MLX5_SET(mbox_out, out, syndrome, syndrome);
+		if (!err)
+			return 0;
+	}
+
+	/* driver or FW delivery error */
+	if (err)
+		return err;
 
-	return cmd_status_to_err(status);
+	/* check outbox status */
+	err = cmd_status_to_err(MLX5_GET(mbox_out, out, status));
+	if (err)
+		cmd_status_print(dev, in, out);
+
+	return err;
 }
 
 static void dump_command(struct mlx5_core_dev *dev,
@@ -980,13 +998,7 @@ static void cmd_work_handler(struct work_struct *work)
 
 	/* Skip sending command to fw if internal error */
 	if (mlx5_cmd_is_down(dev) || !opcode_allowed(&dev->cmd, ent->op)) {
-		u8 status = 0;
-		u32 drv_synd;
-
-		ent->ret = mlx5_internal_err_ret_value(dev, msg_to_opcode(ent->in), &drv_synd, &status);
-		MLX5_SET(mbox_out, ent->out, status, status);
-		MLX5_SET(mbox_out, ent->out, syndrome, drv_synd);
-
+		ent->ret = -ENXIO;
 		mlx5_cmd_comp_handler(dev, 1ULL << ent->idx, true);
 		return;
 	}
@@ -1005,6 +1017,31 @@ static void cmd_work_handler(struct work_struct *work)
 	}
 }
 
+static int deliv_status_to_err(u8 status)
+{
+	switch (status) {
+	case MLX5_CMD_DELIVERY_STAT_OK:
+	case MLX5_DRIVER_STATUS_ABORTED:
+		return 0;
+	case MLX5_CMD_DELIVERY_STAT_SIGNAT_ERR:
+	case MLX5_CMD_DELIVERY_STAT_TOK_ERR:
+		return -EBADR;
+	case MLX5_CMD_DELIVERY_STAT_BAD_BLK_NUM_ERR:
+	case MLX5_CMD_DELIVERY_STAT_OUT_PTR_ALIGN_ERR:
+	case MLX5_CMD_DELIVERY_STAT_IN_PTR_ALIGN_ERR:
+		return -EFAULT; /* Bad address */
+	case MLX5_CMD_DELIVERY_STAT_IN_LENGTH_ERR:
+	case MLX5_CMD_DELIVERY_STAT_OUT_LENGTH_ERR:
+	case MLX5_CMD_DELIVERY_STAT_CMD_DESCR_ERR:
+	case MLX5_CMD_DELIVERY_STAT_RES_FLD_NOT_CLR_ERR:
+		return -ENOMSG;
+	case MLX5_CMD_DELIVERY_STAT_FW_ERR:
+		return -EIO;
+	default:
+		return -EINVAL;
+	}
+}
+
 static const char *deliv_status_to_str(u8 status)
 {
 	switch (status) {
@@ -1622,15 +1659,15 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force
 			ent->ts2 = ktime_get_ns();
 			memcpy(ent->out->first.data, ent->lay->out, sizeof(ent->lay->out));
 			dump_command(dev, ent, 0);
-			if (!ent->ret) {
+
+			if (vec & MLX5_TRIGGERED_CMD_COMP)
+				ent->ret = -ENXIO;
+
+			if (!ent->ret) { /* Command completed by FW */
 				if (!cmd->checksum_disabled)
 					ent->ret = verify_signature(ent);
-				else
-					ent->ret = 0;
-				if (vec & MLX5_TRIGGERED_CMD_COMP)
-					ent->status = MLX5_DRIVER_STATUS_ABORTED;
-				else
-					ent->status = ent->lay->status_own >> 1;
+
+				ent->status = ent->lay->status_own >> 1;
 
 				mlx5_core_dbg(dev, "command completed. ret 0x%x, delivery status %s(0x%x)\n",
 					      ent->ret, deliv_status_to_str(ent->status), ent->status);
@@ -1649,14 +1686,13 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force
 				callback = ent->callback;
 				context = ent->context;
 				err = ent->ret;
-				if (!err) {
+				if (!err && !ent->status) {
 					err = mlx5_copy_from_msg(ent->uout,
 								 ent->out,
 								 ent->uout_size);
 
-					err = err ? err : mlx5_cmd_check(dev,
-									ent->in->first.data,
-									ent->uout);
+					err = mlx5_cmd_check(dev, err, ent->in->first.data,
+							     ent->uout);
 				}
 
 				mlx5_free_cmd_msg(dev, ent->out);
@@ -1729,31 +1765,6 @@ void mlx5_cmd_flush(struct mlx5_core_dev *dev)
 		up(&cmd->sem);
 }
 
-static int deliv_status_to_err(u8 status)
-{
-	switch (status) {
-	case MLX5_CMD_DELIVERY_STAT_OK:
-	case MLX5_DRIVER_STATUS_ABORTED:
-		return 0;
-	case MLX5_CMD_DELIVERY_STAT_SIGNAT_ERR:
-	case MLX5_CMD_DELIVERY_STAT_TOK_ERR:
-		return -EBADR;
-	case MLX5_CMD_DELIVERY_STAT_BAD_BLK_NUM_ERR:
-	case MLX5_CMD_DELIVERY_STAT_OUT_PTR_ALIGN_ERR:
-	case MLX5_CMD_DELIVERY_STAT_IN_PTR_ALIGN_ERR:
-		return -EFAULT; /* Bad address */
-	case MLX5_CMD_DELIVERY_STAT_IN_LENGTH_ERR:
-	case MLX5_CMD_DELIVERY_STAT_OUT_LENGTH_ERR:
-	case MLX5_CMD_DELIVERY_STAT_CMD_DESCR_ERR:
-	case MLX5_CMD_DELIVERY_STAT_RES_FLD_NOT_CLR_ERR:
-		return -ENOMSG;
-	case MLX5_CMD_DELIVERY_STAT_FW_ERR:
-		return -EIO;
-	default:
-		return -EINVAL;
-	}
-}
-
 static struct mlx5_cmd_msg *alloc_msg(struct mlx5_core_dev *dev, int in_size,
 				      gfp_t gfp)
 {
@@ -1812,15 +1823,8 @@ static int cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 	u8 token;
 	int err;
 
-	if (mlx5_cmd_is_down(dev) || !opcode_allowed(&dev->cmd, opcode)) {
-		u32 drv_synd;
-		u8 status;
-
-		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);
-		return err;
-	}
+	if (mlx5_cmd_is_down(dev) || !opcode_allowed(&dev->cmd, opcode))
+		return -ENXIO;
 
 	pages_queue = is_manage_pages(in);
 	gfp = callback ? GFP_ATOMIC : GFP_KERNEL;
@@ -1865,13 +1869,24 @@ out_in:
 	return err;
 }
 
+/**
+ * mlx5_cmd_exec - Executes a fw command, wait for completion
+ *
+ * @dev: mlx5 core device
+ * @in: inbox mlx5_ifc command buffer
+ * @in_size: inbox buffer size
+ * @out: outbox mlx5_ifc buffer
+ * @out_size: outbox size
+ *
+ * @return: 0 if no error, FW command execution was successful,
+ *          and outbox status is ok.
+ */
 int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 		  int out_size)
 {
-	int err;
+	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, false);
 
-	err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, false);
-	return err ? : mlx5_cmd_check(dev, in, out);
+	return mlx5_cmd_check(dev, err, in, out);
 }
 EXPORT_SYMBOL(mlx5_cmd_exec);
 
@@ -1932,11 +1947,9 @@ EXPORT_SYMBOL(mlx5_cmd_exec_cb);
 int mlx5_cmd_exec_polling(struct mlx5_core_dev *dev, void *in, int in_size,
 			  void *out, int out_size)
 {
-	int err;
-
-	err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, true);
+	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, true);
 
-	return err ? : mlx5_cmd_check(dev, in, out);
+	return mlx5_cmd_check(dev, err, in, out);
 }
 EXPORT_SYMBOL(mlx5_cmd_exec_polling);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index 2c774f367199..cea1a8ac196e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -736,10 +736,9 @@ static int mlx5_core_set_issi(struct mlx5_core_dev *dev)
 	MLX5_SET(query_issi_in, query_in, opcode, MLX5_CMD_OP_QUERY_ISSI);
 	err = mlx5_cmd_exec_inout(dev, query_issi, query_in, query_out);
 	if (err) {
-		u32 syndrome;
-		u8 status;
+		u32 syndrome = MLX5_GET(query_issi_out, query_out, syndrome);
+		u8 status = MLX5_GET(query_issi_out, query_out, status);
 
-		mlx5_cmd_mbox_status(query_out, &status, &syndrome);
 		if (!status || syndrome == MLX5_DRIVER_SYND) {
 			mlx5_core_err(dev, "Failed to query ISSI err(%d) status(%d) synd(%d)\n",
 				      err, status, syndrome);
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 1b398c9e17b9..8a8408708e6c 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -981,7 +981,6 @@ int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 
 int mlx5_cmd_exec_polling(struct mlx5_core_dev *dev, void *in, int in_size,
 			  void *out, int out_size);
-void mlx5_cmd_mbox_status(void *out, u8 *status, u32 *syndrome);
 bool mlx5_cmd_is_down(struct mlx5_core_dev *dev);
 
 int mlx5_core_get_caps(struct mlx5_core_dev *dev, enum mlx5_cap_type cap_type);
-- 
cgit v1.2.3


From f23519e542e51c19ab3081deb089bb3f8fec7bb9 Mon Sep 17 00:00:00 2001
From: Saeed Mahameed <saeedm@mellanox.com>
Date: Sat, 17 Aug 2019 03:05:10 -0700
Subject: net/mlx5: cmdif, Add new api for command execution

Add mlx5_cmd_do. Unlike mlx5_cmd_exec, this function will not modify
or translate outbox.status.

The function will return:

return = 0: Command was executed, outbox.status == MLX5_CMD_STAT_OK.

return = -EREMOTEIO: Executed, outbox.status != MLX5_CMD_STAT_OK.

return < 0: Command execution couldn't be performed by FW or driver.

And document other mlx5_cmd_exec functions.

Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 79 ++++++++++++++++++++++-----
 include/linux/mlx5/driver.h                   |  2 +
 2 files changed, 68 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 7ff01b901f53..a2f87a686a18 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -789,7 +789,7 @@ static void cmd_status_print(struct mlx5_core_dev *dev, void *in, void *out)
 			cmd_status_str(status), status, syndrome, err);
 }
 
-static int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *out)
+int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *out)
 {
 	/* aborted due to PCI error or via reset flow mlx5_cmd_trigger_completions() */
 	if (err == -ENXIO) {
@@ -806,7 +806,7 @@ static int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *ou
 	}
 
 	/* driver or FW delivery error */
-	if (err)
+	if (err != -EREMOTEIO && err)
 		return err;
 
 	/* check outbox status */
@@ -816,6 +816,7 @@ static int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *ou
 
 	return err;
 }
+EXPORT_SYMBOL(mlx5_cmd_check);
 
 static void dump_command(struct mlx5_core_dev *dev,
 			 struct mlx5_cmd_work_ent *ent, int input)
@@ -1869,6 +1870,38 @@ out_in:
 	return err;
 }
 
+/**
+ * mlx5_cmd_do - Executes a fw command, wait for completion.
+ * Unlike mlx5_cmd_exec, this function will not translate or intercept
+ * outbox.status and will return -EREMOTEIO when
+ * outbox.status != MLX5_CMD_STAT_OK
+ *
+ * @dev: mlx5 core device
+ * @in: inbox mlx5_ifc command buffer
+ * @in_size: inbox buffer size
+ * @out: outbox mlx5_ifc buffer
+ * @out_size: outbox size
+ *
+ * @return:
+ * -EREMOTEIO : Command executed by FW, outbox.status != MLX5_CMD_STAT_OK.
+ *              Caller must check FW outbox status.
+ *   0 : Command execution successful, outbox.status == MLX5_CMD_STAT_OK.
+ * < 0 : Command execution couldn't be performed by firmware or driver
+ */
+int mlx5_cmd_do(struct mlx5_core_dev *dev, void *in, int in_size, void *out, int out_size)
+{
+	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, false);
+
+	if (err) /* -EREMOTEIO is preserved */
+		return err == -EREMOTEIO ? -EIO : err;
+
+	if (MLX5_GET(mbox_out, out, status) != MLX5_CMD_STAT_OK)
+		return -EREMOTEIO;
+
+	return 0;
+}
+EXPORT_SYMBOL(mlx5_cmd_do);
+
 /**
  * mlx5_cmd_exec - Executes a fw command, wait for completion
  *
@@ -1878,18 +1911,47 @@ out_in:
  * @out: outbox mlx5_ifc buffer
  * @out_size: outbox size
  *
- * @return: 0 if no error, FW command execution was successful,
+ * @return: 0 if no error, FW command execution was successful
  *          and outbox status is ok.
  */
 int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 		  int out_size)
 {
-	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, false);
+	int err = mlx5_cmd_do(dev, in, in_size, out, out_size);
 
 	return mlx5_cmd_check(dev, err, in, out);
 }
 EXPORT_SYMBOL(mlx5_cmd_exec);
 
+/**
+ * mlx5_cmd_exec_polling - Executes a fw command, poll for completion
+ *	Needed for driver force teardown, when command completion EQ
+ *	will not be available to complete the command
+ *
+ * @dev: mlx5 core device
+ * @in: inbox mlx5_ifc command buffer
+ * @in_size: inbox buffer size
+ * @out: outbox mlx5_ifc buffer
+ * @out_size: outbox size
+ *
+ * @return: 0 if no error, FW command execution was successful
+ *          and outbox status is ok.
+ */
+int mlx5_cmd_exec_polling(struct mlx5_core_dev *dev, void *in, int in_size,
+			  void *out, int out_size)
+{
+	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, true);
+
+	if (err) /* -EREMOTEIO is preserved */
+		return err == -EREMOTEIO ? -EIO : err;
+
+	if (MLX5_GET(mbox_out, out, status) != MLX5_CMD_STAT_OK)
+		err = -EREMOTEIO;
+
+	return mlx5_cmd_check(dev, err, in, out);
+}
+EXPORT_SYMBOL(mlx5_cmd_exec_polling);
+
 void mlx5_cmd_init_async_ctx(struct mlx5_core_dev *dev,
 			     struct mlx5_async_ctx *ctx)
 {
@@ -1944,15 +2006,6 @@ int mlx5_cmd_exec_cb(struct mlx5_async_ctx *ctx, void *in, int in_size,
 }
 EXPORT_SYMBOL(mlx5_cmd_exec_cb);
 
-int mlx5_cmd_exec_polling(struct mlx5_core_dev *dev, void *in, int in_size,
-			  void *out, int out_size)
-{
-	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, true);
-
-	return mlx5_cmd_check(dev, err, in, out);
-}
-EXPORT_SYMBOL(mlx5_cmd_exec_polling);
-
 static void destroy_msg_cache(struct mlx5_core_dev *dev)
 {
 	struct cmd_msg_cache *ch;
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 8a8408708e6c..1b9bec8fa870 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -964,6 +964,8 @@ int mlx5_cmd_exec_cb(struct mlx5_async_ctx *ctx, void *in, int in_size,
 		     void *out, int out_size, mlx5_async_cbk_t callback,
 		     struct mlx5_async_work *work);
 
+int mlx5_cmd_do(struct mlx5_core_dev *dev, void *in, int in_size, void *out, int out_size);
+int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *out);
 int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 		  int out_size);
 
-- 
cgit v1.2.3


From 31803e59233efc838b9dcb26edea28a4b2389e97 Mon Sep 17 00:00:00 2001
From: Saeed Mahameed <saeedm@mellanox.com>
Date: Mon, 30 Mar 2020 21:12:58 -0700
Subject: net/mlx5: Use mlx5_cmd_do() in core create_{cq,dct}

mlx5_core_create_{cq/dct} functions are non-trivial mlx5 commands
functions. They check command execution status themselves and hide
valuable FW failure information.

For mlx5_core/eth kernel user this is what we actually want, but for a
devx/rdma user the hidden information is essential and should be propagated
up to the caller, thus we convert these commands to use mlx5_cmd_do
to return the FW/driver and command outbox status as is, and let the caller
decide what to do with it.

For kernel callers of mlx5_core_create_{cq/dct} or those who only care about
the binary status (FAIL/SUCCESS) they must check status themselves via
mlx5_cmd_check() to restore the current behavior.

err = mlx5_create_cq(in, out)
err = mlx5_cmd_check(err, in, out)
if (err)
    // handle err

For DEVX users and those who care about full visibility, They will just
propagate the error to user space, and app can check if err == -EREMOTEIO,
then outbox.{status,syndrome} are valid.

API Note:
mlx5_cmd_check() must be used by kernel users since it allows the driver
to intercept the command execution status and return a driver simulated
status in case of driver induced error handling or reset/recovery flows.

Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/infiniband/hw/mlx5/devx.c            |  6 +++---
 drivers/infiniband/hw/mlx5/qp.c              |  1 +
 drivers/infiniband/hw/mlx5/qpc.c             |  2 +-
 drivers/net/ethernet/mellanox/mlx5/core/cq.c | 17 ++++++++++++++---
 include/linux/mlx5/cq.h                      |  2 ++
 5 files changed, 21 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx5/devx.c b/drivers/infiniband/hw/mlx5/devx.c
index 08b7f6bc56c3..1f62c0ede048 100644
--- a/drivers/infiniband/hw/mlx5/devx.c
+++ b/drivers/infiniband/hw/mlx5/devx.c
@@ -1497,9 +1497,9 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_DEVX_OBJ_CREATE)(
 		   !is_apu_cq(dev, cmd_in)) {
 		obj->flags |= DEVX_OBJ_FLAGS_CQ;
 		obj->core_cq.comp = devx_cq_comp;
-		err = mlx5_core_create_cq(dev->mdev, &obj->core_cq,
-					  cmd_in, cmd_in_len, cmd_out,
-					  cmd_out_len);
+		err = mlx5_create_cq(dev->mdev, &obj->core_cq,
+				     cmd_in, cmd_in_len, cmd_out,
+				     cmd_out_len);
 	} else {
 		err = mlx5_cmd_exec(dev->mdev, cmd_in,
 				    cmd_in_len,
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 29475cf8c7c3..b7fe47107d76 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -4465,6 +4465,7 @@ static int mlx5_ib_modify_dct(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		err = mlx5_core_create_dct(dev, &qp->dct.mdct, qp->dct.in,
 					   MLX5_ST_SZ_BYTES(create_dct_in), out,
 					   sizeof(out));
+		err = mlx5_cmd_check(dev->mdev, err, qp->dct.in, out);
 		if (err)
 			return err;
 		resp.dctn = qp->dct.mdct.mqp.qpn;
diff --git a/drivers/infiniband/hw/mlx5/qpc.c b/drivers/infiniband/hw/mlx5/qpc.c
index 8844eacf2380..542e4c63a8de 100644
--- a/drivers/infiniband/hw/mlx5/qpc.c
+++ b/drivers/infiniband/hw/mlx5/qpc.c
@@ -220,7 +220,7 @@ int mlx5_core_create_dct(struct mlx5_ib_dev *dev, struct mlx5_core_dct *dct,
 	init_completion(&dct->drained);
 	MLX5_SET(create_dct_in, in, opcode, MLX5_CMD_OP_CREATE_DCT);
 
-	err = mlx5_cmd_exec(dev->mdev, in, inlen, out, outlen);
+	err = mlx5_cmd_do(dev->mdev, in, inlen, out, outlen);
 	if (err)
 		return err;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cq.c b/drivers/net/ethernet/mellanox/mlx5/core/cq.c
index 5371ad0a12eb..15a74966be7d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cq.c
@@ -86,8 +86,9 @@ static void mlx5_add_cq_to_tasklet(struct mlx5_core_cq *cq,
 	spin_unlock_irqrestore(&tasklet_ctx->lock, flags);
 }
 
-int mlx5_core_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
-			u32 *in, int inlen, u32 *out, int outlen)
+/* Callers must verify outbox status in case of err */
+int mlx5_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
+		   u32 *in, int inlen, u32 *out, int outlen)
 {
 	int eqn = MLX5_GET(cqc, MLX5_ADDR_OF(create_cq_in, in, cq_context),
 			   c_eqn_or_apu_element);
@@ -101,7 +102,7 @@ int mlx5_core_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
 
 	memset(out, 0, outlen);
 	MLX5_SET(create_cq_in, in, opcode, MLX5_CMD_OP_CREATE_CQ);
-	err = mlx5_cmd_exec(dev, in, inlen, out, outlen);
+	err = mlx5_cmd_do(dev, in, inlen, out, outlen);
 	if (err)
 		return err;
 
@@ -148,6 +149,16 @@ err_cmd:
 	mlx5_cmd_exec_in(dev, destroy_cq, din);
 	return err;
 }
+EXPORT_SYMBOL(mlx5_create_cq);
+
+/* oubox is checked and err val is normalized */
+int mlx5_core_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
+			u32 *in, int inlen, u32 *out, int outlen)
+{
+	int err = mlx5_create_cq(dev, cq, in, inlen, out, outlen);
+
+	return mlx5_cmd_check(dev, err, in, out);
+}
 EXPORT_SYMBOL(mlx5_core_create_cq);
 
 int mlx5_core_destroy_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq)
diff --git a/include/linux/mlx5/cq.h b/include/linux/mlx5/cq.h
index 7bfb67363434..cb15308b5cb0 100644
--- a/include/linux/mlx5/cq.h
+++ b/include/linux/mlx5/cq.h
@@ -183,6 +183,8 @@ static inline void mlx5_cq_put(struct mlx5_core_cq *cq)
 		complete(&cq->free);
 }
 
+int mlx5_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
+		   u32 *in, int inlen, u32 *out, int outlen);
 int mlx5_core_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
 			u32 *in, int inlen, u32 *out, int outlen);
 int mlx5_core_destroy_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq);
-- 
cgit v1.2.3


From 0a41527608e7f3da61e76564f5a8749a1fddc7f1 Mon Sep 17 00:00:00 2001
From: Saeed Mahameed <saeedm@mellanox.com>
Date: Tue, 31 Mar 2020 22:03:00 -0700
Subject: net/mlx5: cmdif, Refactor error handling and reporting of async
 commands

Same as the new mlx5_cmd_do API, report all information to callers and
let them handle the error values and outbox parsing.

The user callback status "work->user_callback(status)" is now similar to
the error rc code returned from the blocking mlx5_cmd_do() version,
and now is defined as follows:

 -EREMOTEIO : Command executed by FW, outbox.status != MLX5_CMD_STAT_OK.
              Caller must check FW outbox status.
 0 : Command execution successful,  outbox.status == MLX5_CMD_STAT_OK.
 < 0 : Command couldn't execute, FW or driver induced error.

Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/infiniband/hw/mlx5/mr.c               | 15 ++++++-
 drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 61 ++++++++++++++++-----------
 include/linux/mlx5/driver.h                   |  3 +-
 3 files changed, 52 insertions(+), 27 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 157d862fb864..06e4b8cea6bd 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -140,6 +140,19 @@ static int destroy_mkey(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
 	return mlx5_core_destroy_mkey(dev->mdev, mr->mmkey.key);
 }
 
+static void create_mkey_warn(struct mlx5_ib_dev *dev, int status, void *out)
+{
+	if (status == -ENXIO) /* core driver is not available */
+		return;
+
+	mlx5_ib_warn(dev, "async reg mr failed. status %d\n", status);
+	if (status != -EREMOTEIO) /* driver specific failure */
+		return;
+
+	/* Failed in FW, print cmd out failure details */
+	mlx5_cmd_out_err(dev->mdev, MLX5_CMD_OP_CREATE_MKEY, 0, out);
+}
+
 static void create_mkey_callback(int status, struct mlx5_async_work *context)
 {
 	struct mlx5_ib_mr *mr =
@@ -149,7 +162,7 @@ static void create_mkey_callback(int status, struct mlx5_async_work *context)
 	unsigned long flags;
 
 	if (status) {
-		mlx5_ib_warn(dev, "async reg mr failed. status %d\n", status);
+		create_mkey_warn(dev, status, mr->out);
 		kfree(mr);
 		spin_lock_irqsave(&ent->lock, flags);
 		ent->pending--;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index a2f87a686a18..823d5808d5a0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -760,6 +760,18 @@ struct mlx5_ifc_mbox_in_bits {
 	u8         reserved_at_40[0x40];
 };
 
+void mlx5_cmd_out_err(struct mlx5_core_dev *dev, u16 opcode, u16 op_mod, void *out)
+{
+	u32 syndrome = MLX5_GET(mbox_out, out, syndrome);
+	u8 status = MLX5_GET(mbox_out, out, status);
+
+	mlx5_core_err_rl(dev,
+			 "%s(0x%x) op_mod(0x%x) failed, status %s(0x%x), syndrome (0x%x), err(%d)\n",
+			 mlx5_command_str(opcode), opcode, op_mod,
+			 cmd_status_str(status), status, syndrome, cmd_status_to_err(status));
+}
+EXPORT_SYMBOL(mlx5_cmd_out_err);
+
 static void cmd_status_print(struct mlx5_core_dev *dev, void *in, void *out)
 {
 	u16 opcode, op_mod;
@@ -778,10 +790,7 @@ static void cmd_status_print(struct mlx5_core_dev *dev, void *in, void *out)
 	err = cmd_status_to_err(status);
 
 	if (!uid && opcode != MLX5_CMD_OP_DESTROY_MKEY)
-		mlx5_core_err_rl(dev,
-			"%s(0x%x) op_mod(0x%x) failed, status %s(0x%x), syndrome (0x%x), err(%d)\n",
-			mlx5_command_str(opcode), opcode, op_mod,
-			cmd_status_str(status), status, syndrome, err);
+		mlx5_cmd_out_err(dev, opcode, op_mod, out);
 	else
 		mlx5_core_dbg(dev,
 			"%s(0x%x) op_mod(0x%x) uid(%d) failed, status %s(0x%x), syndrome (0x%x), err(%d)\n",
@@ -1686,20 +1695,18 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force
 
 				callback = ent->callback;
 				context = ent->context;
-				err = ent->ret;
-				if (!err && !ent->status) {
+				err = ent->ret ? : ent->status;
+				if (err > 0) /* Failed in FW, command didn't execute */
+					err = deliv_status_to_err(err);
+
+				if (!err)
 					err = mlx5_copy_from_msg(ent->uout,
 								 ent->out,
 								 ent->uout_size);
 
-					err = mlx5_cmd_check(dev, err, ent->in->first.data,
-							     ent->uout);
-				}
-
 				mlx5_free_cmd_msg(dev, ent->out);
 				free_msg(dev, ent->in);
 
-				err = err ? err : ent->status;
 				/* final consumer is done, release ent */
 				cmd_ent_put(ent);
 				callback(err, context);
@@ -1870,6 +1877,18 @@ out_in:
 	return err;
 }
 
+/* preserve -EREMOTEIO for outbox.status != OK, otherwise return err as is */
+static int cmd_status_err(int err, void *out)
+{
+	if (err) /* -EREMOTEIO is preserved */
+		return err == -EREMOTEIO ? -EIO : err;
+
+	if (MLX5_GET(mbox_out, out, status) != MLX5_CMD_STAT_OK)
+		return -EREMOTEIO;
+
+	return 0;
+}
+
 /**
  * mlx5_cmd_do - Executes a fw command, wait for completion.
  * Unlike mlx5_cmd_exec, this function will not translate or intercept
@@ -1892,13 +1911,7 @@ int mlx5_cmd_do(struct mlx5_core_dev *dev, void *in, int in_size, void *out, int
 {
 	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, false);
 
-	if (err) /* -EREMOTEIO is preserved */
-		return err == -EREMOTEIO ? -EIO : err;
-
-	if (MLX5_GET(mbox_out, out, status) != MLX5_CMD_STAT_OK)
-		return -EREMOTEIO;
-
-	return 0;
+	return cmd_status_err(err, out);
 }
 EXPORT_SYMBOL(mlx5_cmd_do);
 
@@ -1942,12 +1955,7 @@ int mlx5_cmd_exec_polling(struct mlx5_core_dev *dev, void *in, int in_size,
 {
 	int err = cmd_exec(dev, in, in_size, out, out_size, NULL, NULL, true);
 
-	if (err) /* -EREMOTEIO is preserved */
-		return err == -EREMOTEIO ? -EIO : err;
-
-	if (MLX5_GET(mbox_out, out, status) != MLX5_CMD_STAT_OK)
-		err = -EREMOTEIO;
-
+	err = cmd_status_err(err, out);
 	return mlx5_cmd_check(dev, err, in, out);
 }
 EXPORT_SYMBOL(mlx5_cmd_exec_polling);
@@ -1980,8 +1988,10 @@ EXPORT_SYMBOL(mlx5_cmd_cleanup_async_ctx);
 static void mlx5_cmd_exec_cb_handler(int status, void *_work)
 {
 	struct mlx5_async_work *work = _work;
-	struct mlx5_async_ctx *ctx = work->ctx;
+	struct mlx5_async_ctx *ctx;
 
+	ctx = work->ctx;
+	status = cmd_status_err(status, work->out);
 	work->user_callback(status, work);
 	if (atomic_dec_and_test(&ctx->num_inflight))
 		wake_up(&ctx->wait);
@@ -1995,6 +2005,7 @@ int mlx5_cmd_exec_cb(struct mlx5_async_ctx *ctx, void *in, int in_size,
 
 	work->ctx = ctx;
 	work->user_callback = callback;
+	work->out = out;
 	if (WARN_ON(!atomic_inc_not_zero(&ctx->num_inflight)))
 		return -EIO;
 	ret = cmd_exec(ctx->dev, in, in_size, out, out_size,
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 1b9bec8fa870..432151d7d0bf 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -955,6 +955,7 @@ typedef void (*mlx5_async_cbk_t)(int status, struct mlx5_async_work *context);
 struct mlx5_async_work {
 	struct mlx5_async_ctx *ctx;
 	mlx5_async_cbk_t user_callback;
+	void *out; /* pointer to the cmd output buffer */
 };
 
 void mlx5_cmd_init_async_ctx(struct mlx5_core_dev *dev,
@@ -963,7 +964,7 @@ void mlx5_cmd_cleanup_async_ctx(struct mlx5_async_ctx *ctx);
 int mlx5_cmd_exec_cb(struct mlx5_async_ctx *ctx, void *in, int in_size,
 		     void *out, int out_size, mlx5_async_cbk_t callback,
 		     struct mlx5_async_work *work);
-
+void mlx5_cmd_out_err(struct mlx5_core_dev *dev, u16 opcode, u16 op_mod, void *out);
 int mlx5_cmd_do(struct mlx5_core_dev *dev, void *in, int in_size, void *out, int out_size);
 int mlx5_cmd_check(struct mlx5_core_dev *dev, int err, void *in, void *out);
 int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
-- 
cgit v1.2.3


From 72fb3b60a3114a1154a8ae5629ea3b43a88a7a4d Mon Sep 17 00:00:00 2001
From: Moshe Shemesh <moshe@nvidia.com>
Date: Sat, 14 Aug 2021 17:57:51 +0300
Subject: net/mlx5: Add reset_state field to MFRL register

Add new field reset_state to MFRL register. This field expose current
state of sync reset for fw update. This field enables sharing with the
user more details on why fw activate failed in case it failed the sync
reset stage.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 include/linux/mlx5/mlx5_ifc.h | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 598ac3bcc901..8ca2d65ff789 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -9694,7 +9694,8 @@ struct mlx5_ifc_pcam_reg_bits {
 };
 
 struct mlx5_ifc_mcam_enhanced_features_bits {
-	u8         reserved_at_0[0x6b];
+	u8         reserved_at_0[0x6a];
+	u8         reset_state[0x1];
 	u8         ptpcyc2realtime_modify[0x1];
 	u8         reserved_at_6c[0x2];
 	u8         pci_status_and_power[0x1];
@@ -10375,6 +10376,14 @@ struct mlx5_ifc_mcda_reg_bits {
 	u8         data[][0x20];
 };
 
+enum {
+	MLX5_MFRL_REG_RESET_STATE_IDLE = 0,
+	MLX5_MFRL_REG_RESET_STATE_IN_NEGOTIATION = 1,
+	MLX5_MFRL_REG_RESET_STATE_RESET_IN_PROGRESS = 2,
+	MLX5_MFRL_REG_RESET_STATE_TIMEOUT = 3,
+	MLX5_MFRL_REG_RESET_STATE_NACK = 4,
+};
+
 enum {
 	MLX5_MFRL_REG_RESET_TYPE_FULL_CHIP = BIT(0),
 	MLX5_MFRL_REG_RESET_TYPE_NET_PORT_ALIVE = BIT(1),
@@ -10393,7 +10402,8 @@ struct mlx5_ifc_mfrl_reg_bits {
 	u8         pci_sync_for_fw_update_start[0x1];
 	u8         pci_sync_for_fw_update_resp[0x2];
 	u8         rst_type_sel[0x3];
-	u8         reserved_at_28[0x8];
+	u8         reserved_at_28[0x4];
+	u8         reset_state[0x4];
 	u8         reset_type[0x8];
 	u8         reset_level[0x8];
 };
-- 
cgit v1.2.3


From 45fee8edb4b333af79efad7a99de51718ebda94b Mon Sep 17 00:00:00 2001
From: Moshe Shemesh <moshe@nvidia.com>
Date: Mon, 6 Sep 2021 11:02:44 +0300
Subject: net/mlx5: Add clarification on sync reset failure

In case devlink reload action fw_activate failed in sync reset stage,
use the new MFRL field reset_state to find why it failed and share this
clarification with the user.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/devlink.c  | 10 ++--
 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c | 57 +++++++++++++++++++---
 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h |  3 +-
 drivers/net/ethernet/mellanox/mlx5/core/port.c     | 20 ++++++--
 include/linux/mlx5/driver.h                        |  3 ++
 5 files changed, 74 insertions(+), 19 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index d1093bb2d436..057dde6f4417 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -100,15 +100,11 @@ static int mlx5_devlink_reload_fw_activate(struct devlink *devlink, struct netli
 	}
 
 	net_port_alive = !!(reset_type & MLX5_MFRL_REG_RESET_TYPE_NET_PORT_ALIVE);
-	err = mlx5_fw_reset_set_reset_sync(dev, net_port_alive);
+	err = mlx5_fw_reset_set_reset_sync(dev, net_port_alive, extack);
 	if (err)
-		goto out;
+		return err;
 
-	err = mlx5_fw_reset_wait_reset_done(dev);
-out:
-	if (err)
-		NL_SET_ERR_MSG_MOD(extack, "FW activate command failed");
-	return err;
+	return mlx5_fw_reset_wait_reset_done(dev);
 }
 
 static int mlx5_devlink_trigger_fw_live_patch(struct devlink *devlink,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index 0b0234f9d694..d438d7a61500 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -57,7 +57,8 @@ static int mlx5_reg_mfrl_set(struct mlx5_core_dev *dev, u8 reset_level,
 	return mlx5_core_access_reg(dev, in, sizeof(in), out, sizeof(out), MLX5_REG_MFRL, 0, 1);
 }
 
-static int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_type)
+static int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level,
+			       u8 *reset_type, u8 *reset_state)
 {
 	u32 out[MLX5_ST_SZ_DW(mfrl_reg)] = {};
 	u32 in[MLX5_ST_SZ_DW(mfrl_reg)] = {};
@@ -71,25 +72,67 @@ static int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *r
 		*reset_level = MLX5_GET(mfrl_reg, out, reset_level);
 	if (reset_type)
 		*reset_type = MLX5_GET(mfrl_reg, out, reset_type);
+	if (reset_state)
+		*reset_state = MLX5_GET(mfrl_reg, out, reset_state);
 
 	return 0;
 }
 
 int mlx5_fw_reset_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_type)
 {
-	return mlx5_reg_mfrl_query(dev, reset_level, reset_type);
+	return mlx5_reg_mfrl_query(dev, reset_level, reset_type, NULL);
 }
 
-int mlx5_fw_reset_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel)
+static int mlx5_fw_reset_get_reset_state_err(struct mlx5_core_dev *dev,
+					     struct netlink_ext_ack *extack)
+{
+	u8 reset_state;
+
+	if (mlx5_reg_mfrl_query(dev, NULL, NULL, &reset_state))
+		goto out;
+
+	switch (reset_state) {
+	case MLX5_MFRL_REG_RESET_STATE_IN_NEGOTIATION:
+	case MLX5_MFRL_REG_RESET_STATE_RESET_IN_PROGRESS:
+		NL_SET_ERR_MSG_MOD(extack, "Sync reset was already triggered");
+		return -EBUSY;
+	case MLX5_MFRL_REG_RESET_STATE_TIMEOUT:
+		NL_SET_ERR_MSG_MOD(extack, "Sync reset got timeout");
+		return -ETIMEDOUT;
+	case MLX5_MFRL_REG_RESET_STATE_NACK:
+		NL_SET_ERR_MSG_MOD(extack, "One of the hosts disabled reset");
+		return -EPERM;
+	}
+
+out:
+	NL_SET_ERR_MSG_MOD(extack, "Sync reset failed");
+	return -EIO;
+}
+
+int mlx5_fw_reset_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel,
+				 struct netlink_ext_ack *extack)
 {
 	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+	u32 out[MLX5_ST_SZ_DW(mfrl_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(mfrl_reg)] = {};
 	int err;
 
 	set_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags);
-	err = mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, reset_type_sel, 0, true);
-	if (err)
-		clear_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags);
-	return err;
+
+	MLX5_SET(mfrl_reg, in, reset_level, MLX5_MFRL_REG_RESET_LEVEL3);
+	MLX5_SET(mfrl_reg, in, rst_type_sel, reset_type_sel);
+	MLX5_SET(mfrl_reg, in, pci_sync_for_fw_update_start, 1);
+	err = mlx5_access_reg(dev, in, sizeof(in), out, sizeof(out),
+			      MLX5_REG_MFRL, 0, 1, false);
+	if (!err)
+		return 0;
+
+	clear_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags);
+	if (err == -EREMOTEIO && MLX5_CAP_MCAM_FEATURE(dev, reset_state))
+		return mlx5_fw_reset_get_reset_state_err(dev, extack);
+
+	NL_SET_ERR_MSG_MOD(extack, "Sync reset command failed");
+	return mlx5_cmd_check(dev, err, in, out);
 }
 
 int mlx5_fw_reset_set_live_patch(struct mlx5_core_dev *dev)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
index 7761ee5fc7d0..694fc7cb2684 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
@@ -9,7 +9,8 @@
 void mlx5_fw_reset_enable_remote_dev_reset_set(struct mlx5_core_dev *dev, bool enable);
 bool mlx5_fw_reset_enable_remote_dev_reset_get(struct mlx5_core_dev *dev);
 int mlx5_fw_reset_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_type);
-int mlx5_fw_reset_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel);
+int mlx5_fw_reset_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel,
+				 struct netlink_ext_ack *extack);
 int mlx5_fw_reset_set_live_patch(struct mlx5_core_dev *dev);
 
 int mlx5_fw_reset_wait_reset_done(struct mlx5_core_dev *dev);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c
index 1ef2b6a848c1..d15b417d3e07 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/port.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c
@@ -33,9 +33,10 @@
 #include <linux/mlx5/port.h>
 #include "mlx5_core.h"
 
-int mlx5_core_access_reg(struct mlx5_core_dev *dev, void *data_in,
-			 int size_in, void *data_out, int size_out,
-			 u16 reg_id, int arg, int write)
+/* calling with verbose false will not print error to log */
+int mlx5_access_reg(struct mlx5_core_dev *dev, void *data_in, int size_in,
+		    void *data_out, int size_out, u16 reg_id, int arg,
+		    int write, bool verbose)
 {
 	int outlen = MLX5_ST_SZ_BYTES(access_register_out) + size_out;
 	int inlen = MLX5_ST_SZ_BYTES(access_register_in) + size_in;
@@ -57,7 +58,9 @@ int mlx5_core_access_reg(struct mlx5_core_dev *dev, void *data_in,
 	MLX5_SET(access_register_in, in, argument, arg);
 	MLX5_SET(access_register_in, in, register_id, reg_id);
 
-	err = mlx5_cmd_exec(dev, in, inlen, out, outlen);
+	err = mlx5_cmd_do(dev, in, inlen, out, outlen);
+	if (verbose)
+		err = mlx5_cmd_check(dev, err, in, out);
 	if (err)
 		goto out;
 
@@ -69,6 +72,15 @@ out:
 	kvfree(in);
 	return err;
 }
+EXPORT_SYMBOL_GPL(mlx5_access_reg);
+
+int mlx5_core_access_reg(struct mlx5_core_dev *dev, void *data_in,
+			 int size_in, void *data_out, int size_out,
+			 u16 reg_id, int arg, int write)
+{
+	return mlx5_access_reg(dev, data_in, size_in, data_out, size_out,
+			       reg_id, arg, write, true);
+}
 EXPORT_SYMBOL_GPL(mlx5_core_access_reg);
 
 int mlx5_query_pcam_reg(struct mlx5_core_dev *dev, u32 *pcam, u8 feature_group,
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 432151d7d0bf..d3b1a6a1f8d2 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -1031,6 +1031,9 @@ int mlx5_core_detach_mcg(struct mlx5_core_dev *dev, union ib_gid *mgid, u32 qpn)
 
 void mlx5_qp_debugfs_init(struct mlx5_core_dev *dev);
 void mlx5_qp_debugfs_cleanup(struct mlx5_core_dev *dev);
+int mlx5_access_reg(struct mlx5_core_dev *dev, void *data_in, int size_in,
+		    void *data_out, int size_out, u16 reg_id, int arg,
+		    int write, bool verbose);
 int mlx5_core_access_reg(struct mlx5_core_dev *dev, void *data_in,
 			 int size_in, void *data_out, int size_out,
 			 u16 reg_num, int arg, int write);
-- 
cgit v1.2.3