All of lore.kernel.org
 help / color / mirror / Atom feed
From: Saeed Mahameed <saeed@kernel.org>
To: "David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	Jason Gunthorpe <jgg@nvidia.com>
Cc: Leon Romanovsky <leonro@nvidia.com>,
	linux-rdma@vger.kernel.org, netdev@vger.kernel.org,
	Saeed Mahameed <saeedm@mellanox.com>,
	Saeed Mahameed <saeedm@nvidia.com>
Subject: [mlx5-next 14/17] net/mlx5: cmdif, Refactor error handling and reporting of async commands
Date: Tue, 22 Feb 2022 21:09:29 -0800	[thread overview]
Message-ID: <20220223050932.244668-15-saeed@kernel.org> (raw)
In-Reply-To: <20220223050932.244668-1-saeed@kernel.org>

From: Saeed Mahameed <saeedm@mellanox.com>

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(-)

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 @@ static int cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 	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,
-- 
2.35.1


  parent reply	other threads:[~2022-02-23  5:10 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-23  5:09 [pull request][net-next/rdma-next 00/17] mlx5-next updates 2022-02-22 Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 01/17] mlx5: remove usused static inlines Saeed Mahameed
2022-02-23 23:17   ` Jakub Kicinski
2022-02-23 23:20     ` Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 02/17] net/mlx5: Add ability to insert to specific flow group Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 03/17] net/mlx5: E-Switch, reserve and use same uplink metadata across ports Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 04/17] net/mlx5: E-switch, remove special uplink ingress ACL handling Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 05/17] net/mlx5: E-switch, add drop rule support to ingress ACL Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 06/17] net/mlx5: Lag, use local variable already defined to access E-Switch Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 07/17] net/mlx5: Lag, don't use magic numbers for ports Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 08/17] net/mlx5: Lag, record inactive state of bond device Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 09/17] net/mlx5: Lag, offload active-backup drops to hardware Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 10/17] net/mlx5: cmdif, Return value improvements Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 11/17] net/mlx5: cmdif, cmd_check refactoring Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 12/17] net/mlx5: cmdif, Add new api for command execution Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 13/17] net/mlx5: Use mlx5_cmd_do() in core create_{cq,dct} Saeed Mahameed
2022-02-23 23:20   ` Jakub Kicinski
2022-02-23 23:57     ` Saeed Mahameed
2022-02-24  2:00       ` Jakub Kicinski
2022-02-23  5:09 ` Saeed Mahameed [this message]
2022-02-23  5:09 ` [mlx5-next 15/17] RDMA/mlx5: Use new command interface API Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 16/17] net/mlx5: Add reset_state field to MFRL register Saeed Mahameed
2022-02-23  5:09 ` [mlx5-next 17/17] net/mlx5: Add clarification on sync reset failure Saeed Mahameed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220223050932.244668-15-saeed@kernel.org \
    --to=saeed@kernel.org \
    --cc=davem@davemloft.net \
    --cc=jgg@nvidia.com \
    --cc=kuba@kernel.org \
    --cc=leonro@nvidia.com \
    --cc=linux-rdma@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=saeedm@mellanox.com \
    --cc=saeedm@nvidia.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.