From: Saeed Mahameed <saeed@kernel.org>
To: "David S. Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
Eric Dumazet <edumazet@google.com>
Cc: Saeed Mahameed <saeedm@nvidia.com>,
netdev@vger.kernel.org, Tariq Toukan <tariqt@nvidia.com>,
Shay Drory <shayd@nvidia.com>, Moshe Shemesh <moshe@nvidia.com>
Subject: [net-next 12/15] net/mlx5: Split function_setup() to enable and open functions
Date: Fri, 9 Jun 2023 18:42:51 -0700 [thread overview]
Message-ID: <20230610014254.343576-13-saeed@kernel.org> (raw)
In-Reply-To: <20230610014254.343576-1-saeed@kernel.org>
From: Shay Drory <shayd@nvidia.com>
mlx5_cmd_init_hca() is taking ~0.2 seconds. In case of a user who
desire to disable some of the SF aux devices, and with large scale-1K
SFs for example, this user will waste more than 3 minutes on
mlx5_cmd_init_hca() which isn't needed at this stage.
Downstream patch will change SFs which are probe over the E-switch,
local SFs, to be probed without any aux dev. In order to support this,
split function_setup() to avoid executing mlx5_cmd_init_hca().
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/main.c | 83 +++++++++++++------
1 file changed, 58 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index fed8b48a5b20..0faae77d84e6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -1118,7 +1118,7 @@ static void mlx5_cleanup_once(struct mlx5_core_dev *dev)
mlx5_devcom_unregister_device(dev->priv.devcom);
}
-static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot, u64 timeout)
+static int mlx5_function_enable(struct mlx5_core_dev *dev, bool boot, u64 timeout)
{
int err;
@@ -1183,28 +1183,56 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot, u64 timeout
goto reclaim_boot_pages;
}
+ return 0;
+
+reclaim_boot_pages:
+ mlx5_reclaim_startup_pages(dev);
+err_disable_hca:
+ mlx5_core_disable_hca(dev, 0);
+stop_health_poll:
+ mlx5_stop_health_poll(dev, boot);
+err_cmd_cleanup:
+ mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN);
+ mlx5_cmd_cleanup(dev);
+
+ return err;
+}
+
+static void mlx5_function_disable(struct mlx5_core_dev *dev, bool boot)
+{
+ mlx5_reclaim_startup_pages(dev);
+ mlx5_core_disable_hca(dev, 0);
+ mlx5_stop_health_poll(dev, boot);
+ mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN);
+ mlx5_cmd_cleanup(dev);
+}
+
+static int mlx5_function_open(struct mlx5_core_dev *dev)
+{
+ int err;
+
err = set_hca_ctrl(dev);
if (err) {
mlx5_core_err(dev, "set_hca_ctrl failed\n");
- goto reclaim_boot_pages;
+ return err;
}
err = set_hca_cap(dev);
if (err) {
mlx5_core_err(dev, "set_hca_cap failed\n");
- goto reclaim_boot_pages;
+ return err;
}
err = mlx5_satisfy_startup_pages(dev, 0);
if (err) {
mlx5_core_err(dev, "failed to allocate init pages\n");
- goto reclaim_boot_pages;
+ return err;
}
err = mlx5_cmd_init_hca(dev, sw_owner_id);
if (err) {
mlx5_core_err(dev, "init hca failed\n");
- goto reclaim_boot_pages;
+ return err;
}
mlx5_set_driver_version(dev);
@@ -1212,26 +1240,13 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot, u64 timeout
err = mlx5_query_hca_caps(dev);
if (err) {
mlx5_core_err(dev, "query hca failed\n");
- goto reclaim_boot_pages;
+ return err;
}
mlx5_start_health_fw_log_up(dev);
-
return 0;
-
-reclaim_boot_pages:
- mlx5_reclaim_startup_pages(dev);
-err_disable_hca:
- mlx5_core_disable_hca(dev, 0);
-stop_health_poll:
- mlx5_stop_health_poll(dev, boot);
-err_cmd_cleanup:
- mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN);
- mlx5_cmd_cleanup(dev);
-
- return err;
}
-static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot)
+static int mlx5_function_close(struct mlx5_core_dev *dev)
{
int err;
@@ -1240,15 +1255,33 @@ static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot)
mlx5_core_err(dev, "tear_down_hca failed, skip cleanup\n");
return err;
}
- mlx5_reclaim_startup_pages(dev);
- mlx5_core_disable_hca(dev, 0);
- mlx5_stop_health_poll(dev, boot);
- mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN);
- mlx5_cmd_cleanup(dev);
return 0;
}
+static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot, u64 timeout)
+{
+ int err;
+
+ err = mlx5_function_enable(dev, boot, timeout);
+ if (err)
+ return err;
+
+ err = mlx5_function_open(dev);
+ if (err)
+ mlx5_function_disable(dev, boot);
+ return err;
+}
+
+static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot)
+{
+ int err = mlx5_function_close(dev);
+
+ if (!err)
+ mlx5_function_disable(dev, boot);
+ return err;
+}
+
static int mlx5_load(struct mlx5_core_dev *dev)
{
int err;
--
2.40.1
next prev parent reply other threads:[~2023-06-10 1:43 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-06-10 1:42 [pull request][net-next 00/15] mlx5 updates 2023-06-09 Saeed Mahameed
2023-06-10 1:42 ` [net-next 01/15] net/mlx5: Simplify unload all rep code Saeed Mahameed
2023-06-12 11:00 ` patchwork-bot+netdevbpf
2023-06-10 1:42 ` [net-next 02/15] net/mlx5: mlx5_ifc updates for embedded CPU SRIOV Saeed Mahameed
2023-06-10 1:42 ` [net-next 03/15] net/mlx5: Enable devlink port for embedded cpu VF vports Saeed Mahameed
2023-06-10 1:42 ` [net-next 04/15] net/mlx5: Update vport caps query/set for EC VFs Saeed Mahameed
2023-06-10 1:42 ` [net-next 05/15] net/mlx5: Add management of EC VF vports Saeed Mahameed
2023-06-10 1:42 ` [net-next 06/15] net/mlx5: Add/remove peer miss rules for EC VFs Saeed Mahameed
2023-06-10 1:42 ` [net-next 07/15] net/mlx5: Add new page type for EC VF pages Saeed Mahameed
2023-06-10 1:42 ` [net-next 08/15] net/mlx5: Use correct vport when restoring GUIDs Saeed Mahameed
2023-06-10 1:42 ` [net-next 09/15] net/mlx5: Query correct caps for min msix vectors Saeed Mahameed
2023-06-10 1:42 ` [net-next 10/15] net/mlx5: Update SRIOV enable/disable to handle EC/VFs Saeed Mahameed
2023-06-10 1:42 ` [net-next 11/15] net/mlx5: Set max number of embedded CPU VFs Saeed Mahameed
2023-06-10 1:42 ` Saeed Mahameed [this message]
2023-06-10 1:42 ` [net-next 13/15] net/mlx5: Move esw multiport devlink param to eswitch code Saeed Mahameed
2023-06-10 1:42 ` [net-next 14/15] net/mlx5: Light probe local SFs Saeed Mahameed
2023-06-10 7:01 ` Jakub Kicinski
2023-06-11 4:15 ` Saeed Mahameed
2023-06-11 5:10 ` Samudrala, Sridhar
2023-06-13 23:41 ` Saeed Mahameed
2023-06-12 17:51 ` Jakub Kicinski
2023-06-13 23:32 ` Saeed Mahameed
2023-06-14 2:05 ` Jakub Kicinski
2023-06-15 10:51 ` Jiri Pirko
2023-06-15 16:37 ` Jakub Kicinski
2023-06-15 17:37 ` Jiri Pirko
2023-06-15 19:33 ` Jakub Kicinski
2023-06-21 13:14 ` Jiri Pirko
2023-06-21 18:23 ` Jakub Kicinski
2023-06-22 6:42 ` Jiri Pirko
2023-06-22 6:38 ` Jiri Pirko
2023-06-22 16:35 ` Jakub Kicinski
2023-06-23 9:27 ` Jiri Pirko
2023-06-23 15:21 ` Jakub Kicinski
2023-06-24 9:33 ` Jiri Pirko
2023-06-24 20:47 ` Jakub Kicinski
2023-06-27 10:12 ` Jiri Pirko
2023-06-27 15:24 ` Jakub Kicinski
2023-06-27 17:16 ` Jiri Pirko
2023-06-27 17:35 ` Jakub Kicinski
2023-06-10 1:42 ` [net-next 15/15] net/mlx5e: Remove a useless function call 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=20230610014254.343576-13-saeed@kernel.org \
--to=saeed@kernel.org \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=kuba@kernel.org \
--cc=moshe@nvidia.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=saeedm@nvidia.com \
--cc=shayd@nvidia.com \
--cc=tariqt@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).