* [pull request][net-next 00/16] mlx5 updates 2021-08-02
@ 2021-08-03 2:28 Saeed Mahameed
2021-08-03 2:28 ` [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table Saeed Mahameed
` (15 more replies)
0 siblings, 16 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Saeed Mahameed
From: Saeed Mahameed <saeedm@nvidia.com>
Hi Dave/Jakub,
This series adds misc cleanup and some refactroing to mlx5e driver,
needed for upcoming submissions.
For more information please see tag log below.
Please pull and let me know if there is any problem.
Thanks,
Saeed.
---
The following changes since commit 1187c8c4642d109037202b43a5054adaef78b760:
net: phy: mscc: make some arrays static const, makes object smaller (2021-08-02 09:15:07 -0700)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5-updates-2021-08-02
for you to fetch changes up to bcd68c04c7692416206414dc8971730aa140eba7:
net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set() (2021-08-02 19:26:29 -0700)
----------------------------------------------------------------
mlx5-updates-2021-08-02
This patch-set changes the TTC (Traffic Type Classification) logic
to be independent from the mlx5 ethernet driver by renaming the traffic
types enums and making the TTC API generic to the mlx5 core driver.
It allows to decouple TTC logic from mlx5e and reused by other parts
of mlx5 drivers, namely ADQ and lag TX steering hashing.
Patches overview:
1 - Rename traffic type enums to be mlx5 generic.
2 - Rename related TTC arguments and functions.
3 - Remove dependency in the mlx5e driver from the TTC implementation.
4 - Move TTC logic to fs_ttc.
5 - Embed struct mlx5_ttc_table in fs_ttc.
The refactoring series is followed by misc' cleanup patches.
----------------------------------------------------------------
Jiapeng Chong (1):
net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set()
Maor Gottlieb (5):
net/mlx5e: Rename traffic type enums
net/mlx5e: Rename some related TTC args and functions
net/mlx5e: Decouple TTC logic from mlx5e
net/mlx5: Move TTC logic to fs_ttc
net/mlx5: Embed mlx5_ttc_table
Maxim Mikityanskiy (4):
net/mlx5e: Use a new initializer to build uniform indir table
net/mlx5e: Introduce mlx5e_channels API to get RQNs
net/mlx5e: Hide all implementation details of mlx5e_rx_res
net/mlx5e: Allocate the array of channels according to the real max_nch
Roi Dayan (6):
net/mlx5e: Remove redundant tc act includes
net/mlx5e: Remove redundant filter_dev arg from parse_tc_fdb_actions()
net/mlx5e: Remove redundant cap check for flow counter
net/mlx5e: Remove redundant parse_attr arg
net/mlx5e: Remove redundant assignment of counter to null
net/mlx5e: Return -EOPNOTSUPP if more relevant when parsing tc actions
drivers/net/ethernet/mellanox/mlx5/core/Makefile | 4 +-
drivers/net/ethernet/mellanox/mlx5/core/en.h | 20 +-
.../net/ethernet/mellanox/mlx5/core/en/channels.c | 46 ++
.../net/ethernet/mellanox/mlx5/core/en/channels.h | 16 +
drivers/net/ethernet/mellanox/mlx5/core/en/fs.h | 84 +--
.../mellanox/mlx5/core/en/fs_tt_redirect.c | 30 +-
.../mellanox/mlx5/core/en/fs_tt_redirect.h | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c | 6 +-
drivers/net/ethernet/mellanox/mlx5/core/en/rqt.c | 9 +
drivers/net/ethernet/mellanox/mlx5/core/en/rqt.h | 3 +
.../net/ethernet/mellanox/mlx5/core/en/rx_res.c | 782 ++++++++++++++++++++-
.../net/ethernet/mellanox/mlx5/core/en/rx_res.h | 76 +-
.../net/ethernet/mellanox/mlx5/core/en/xsk/pool.c | 4 +-
.../net/ethernet/mellanox/mlx5/core/en/xsk/setup.c | 56 --
.../net/ethernet/mellanox/mlx5/core/en/xsk/setup.h | 4 -
.../ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c | 12 +-
.../mellanox/mlx5/core/en_accel/ipsec_fs.c | 13 +-
.../ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c | 24 +-
.../net/ethernet/mellanox/mlx5/core/en_ethtool.c | 69 +-
drivers/net/ethernet/mellanox/mlx5/core/en_fs.c | 677 +++---------------
.../ethernet/mellanox/mlx5/core/en_fs_ethtool.c | 55 +-
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 507 +------------
drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 66 +-
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 86 +--
.../ethernet/mellanox/mlx5/core/eswitch_offloads.c | 5 +-
.../net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c | 55 +-
.../net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c | 602 ++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h | 70 ++
include/linux/mlx5/fs.h | 2 +
30 files changed, 1852 insertions(+), 1535 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en/channels.c
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en/channels.h
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
^ permalink raw reply [flat|nested] 18+ messages in thread
* [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 10:30 ` patchwork-bot+netdevbpf
2021-08-03 2:28 ` [net-next 02/16] net/mlx5e: Introduce mlx5e_channels API to get RQNs Saeed Mahameed
` (14 subsequent siblings)
15 siblings, 1 reply; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maxim Mikityanskiy, Tariq Toukan, Saeed Mahameed
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Replace mlx5e_build_default_indir_rqt with a new initializer of struct
mlx5e_rss_params_indir that works directly with the struct, rather than
its internals.
The new initializer is called mlx5e_rss_params_indir_init_uniform, which
also reflects the purpose (uniform spreading) better.
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 3 ---
drivers/net/ethernet/mellanox/mlx5/core/en/rqt.c | 9 +++++++++
drivers/net/ethernet/mellanox/mlx5/core/en/rqt.h | 3 +++
.../net/ethernet/mellanox/mlx5/core/en_main.c | 16 +++-------------
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 4 ++--
5 files changed, 17 insertions(+), 18 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 35668986878a..87dabb8b8ac4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -984,9 +984,6 @@ void mlx5e_activate_priv_channels(struct mlx5e_priv *priv);
void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv);
int mlx5e_ptp_rx_manage_fs_ctx(struct mlx5e_priv *priv, void *ctx);
-void mlx5e_build_default_indir_rqt(u32 *indirection_rqt, int len,
- int num_channels);
-
int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state);
void mlx5e_activate_rq(struct mlx5e_rq *rq);
void mlx5e_deactivate_rq(struct mlx5e_rq *rq);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.c
index 38d0e9dbd6bd..b915fb29dd2c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.c
@@ -4,6 +4,15 @@
#include "rqt.h"
#include <linux/mlx5/transobj.h>
+void mlx5e_rss_params_indir_init_uniform(struct mlx5e_rss_params_indir *indir,
+ unsigned int num_channels)
+{
+ unsigned int i;
+
+ for (i = 0; i < MLX5E_INDIR_RQT_SIZE; i++)
+ indir->table[i] = i % num_channels;
+}
+
static int mlx5e_rqt_init(struct mlx5e_rqt *rqt, struct mlx5_core_dev *mdev,
u16 max_size, u32 *init_rqns, u16 init_size)
{
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.h b/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.h
index d2c76649efb0..60c985a12f24 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rqt.h
@@ -14,6 +14,9 @@ struct mlx5e_rss_params_indir {
u32 table[MLX5E_INDIR_RQT_SIZE];
};
+void mlx5e_rss_params_indir_init_uniform(struct mlx5e_rss_params_indir *indir,
+ unsigned int num_channels);
+
struct mlx5e_rqt {
struct mlx5_core_dev *mdev;
u32 rqtn;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 3e644d3955a8..68be4e0e77bf 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -2572,8 +2572,8 @@ int mlx5e_num_channels_changed(struct mlx5e_priv *priv)
/* This function may be called on attach, before priv->rx_res is created. */
if (!netif_is_rxfh_configured(priv->netdev) && priv->rx_res)
- mlx5e_build_default_indir_rqt(priv->rx_res->rss_params.indir.table,
- MLX5E_INDIR_RQT_SIZE, count);
+ mlx5e_rss_params_indir_init_uniform(&priv->rx_res->rss_params.indir,
+ count);
return 0;
}
@@ -4459,15 +4459,6 @@ const struct net_device_ops mlx5e_netdev_ops = {
.ndo_get_devlink_port = mlx5e_get_devlink_port,
};
-void mlx5e_build_default_indir_rqt(u32 *indirection_rqt, int len,
- int num_channels)
-{
- int i;
-
- for (i = 0; i < len; i++)
- indirection_rqt[i] = i % num_channels;
-}
-
static u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeout)
{
int i;
@@ -4488,8 +4479,7 @@ void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params,
rss_params->hash.hfunc = ETH_RSS_HASH_TOP;
netdev_rss_key_fill(rss_params->hash.toeplitz_hash_key,
sizeof(rss_params->hash.toeplitz_hash_key));
- mlx5e_build_default_indir_rqt(rss_params->indir.table,
- MLX5E_INDIR_RQT_SIZE, num_channels);
+ mlx5e_rss_params_indir_init_uniform(&rss_params->indir, num_channels);
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
rss_params->rx_hash_fields[tt] =
mlx5e_rss_get_default_tt_config(tt).rx_hash_fields;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 2ef02fea119a..4d7ed24ae13c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -525,9 +525,9 @@ static int mlx5e_hairpin_create_indirect_rqt(struct mlx5e_hairpin *hp)
if (!indir)
return -ENOMEM;
- mlx5e_build_default_indir_rqt(indir->table, MLX5E_INDIR_RQT_SIZE, hp->num_channels);
+ mlx5e_rss_params_indir_init_uniform(indir, hp->num_channels);
err = mlx5e_rqt_init_indir(&hp->indir_rqt, mdev, hp->pair->rqn, hp->num_channels,
- priv->rx_res->rss_params.hash.hfunc, indir);
+ priv->rx_res->rss_params.hash.hfunc, indir);
kvfree(indir);
return err;
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 02/16] net/mlx5e: Introduce mlx5e_channels API to get RQNs
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
2021-08-03 2:28 ` [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 03/16] net/mlx5e: Hide all implementation details of mlx5e_rx_res Saeed Mahameed
` (13 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maxim Mikityanskiy, Tariq Toukan, Saeed Mahameed
From: Maxim Mikityanskiy <maximmi@nvidia.com>
Currently, struct mlx5e_channels is defined in en.h, along with a lot of
other stuff. In the following commit mlx5e_rx_res will need to get RQNs
(RQ hardware IDs), given a pointer to mlx5e_channels and the channel
index. In order to make it possible without including the whole en.h,
this commit introduces functions that will hide the implementation
details of mlx5e_channels.
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
.../ethernet/mellanox/mlx5/core/en/channels.c | 46 +++++++++++++++++++
.../ethernet/mellanox/mlx5/core/en/channels.h | 16 +++++++
3 files changed, 63 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en/channels.c
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en/channels.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
index 6378dc815df7..e8522ccb3519 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
@@ -28,7 +28,7 @@ mlx5_core-$(CONFIG_MLX5_CORE_EN) += en_main.o en_common.o en_fs.o en_ethtool.o \
en/reporter_tx.o en/reporter_rx.o en/params.o en/xsk/pool.o \
en/xsk/setup.o en/xsk/rx.o en/xsk/tx.o en/devlink.o en/ptp.o \
en/qos.o en/trap.o en/fs_tt_redirect.o en/rqt.o en/tir.o \
- en/rx_res.o
+ en/rx_res.o en/channels.o
#
# Netdev extra
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c
new file mode 100644
index 000000000000..e7c14c0de0a7
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c
@@ -0,0 +1,46 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+/* Copyright (c) 2021, Mellanox Technologies inc. All rights reserved. */
+
+#include "channels.h"
+#include "en.h"
+#include "en/ptp.h"
+
+unsigned int mlx5e_channels_get_num(struct mlx5e_channels *chs)
+{
+ return chs->num;
+}
+
+void mlx5e_channels_get_regular_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn)
+{
+ struct mlx5e_channel *c;
+
+ WARN_ON(ix >= mlx5e_channels_get_num(chs));
+ c = chs->c[ix];
+
+ *rqn = c->rq.rqn;
+}
+
+bool mlx5e_channels_get_xsk_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn)
+{
+ struct mlx5e_channel *c;
+
+ WARN_ON(ix >= mlx5e_channels_get_num(chs));
+ c = chs->c[ix];
+
+ if (!test_bit(MLX5E_CHANNEL_STATE_XSK, c->state))
+ return false;
+
+ *rqn = c->xskrq.rqn;
+ return true;
+}
+
+bool mlx5e_channels_get_ptp_rqn(struct mlx5e_channels *chs, u32 *rqn)
+{
+ struct mlx5e_ptp *c = chs->ptp;
+
+ if (!c || !test_bit(MLX5E_PTP_STATE_RX, c->state))
+ return false;
+
+ *rqn = c->rq.rqn;
+ return true;
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h
new file mode 100644
index 000000000000..ca00cbc827cb
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/* Copyright (c) 2021, Mellanox Technologies inc. All rights reserved. */
+
+#ifndef __MLX5_EN_CHANNELS_H__
+#define __MLX5_EN_CHANNELS_H__
+
+#include <linux/kernel.h>
+
+struct mlx5e_channels;
+
+unsigned int mlx5e_channels_get_num(struct mlx5e_channels *chs);
+void mlx5e_channels_get_regular_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn);
+bool mlx5e_channels_get_xsk_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn);
+bool mlx5e_channels_get_ptp_rqn(struct mlx5e_channels *chs, u32 *rqn);
+
+#endif /* __MLX5_EN_CHANNELS_H__ */
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 03/16] net/mlx5e: Hide all implementation details of mlx5e_rx_res
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
2021-08-03 2:28 ` [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table Saeed Mahameed
2021-08-03 2:28 ` [net-next 02/16] net/mlx5e: Introduce mlx5e_channels API to get RQNs Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 04/16] net/mlx5e: Allocate the array of channels according to the real max_nch Saeed Mahameed
` (12 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maxim Mikityanskiy, Tariq Toukan, Saeed Mahameed
From: Maxim Mikityanskiy <maximmi@nvidia.com>
This commit moves all implementation details of struct mlx5e_rx_res
under en/rx_res.c. All access to RX resources is now done using methods.
Encapsulating RX resources into an object allows for better
manageability, because all the implementation details are now in a
single place, and external code can use only a limited set of API
methods to init/teardown the whole thing, reconfigure RSS and LRO
parameters, connect TIRs to flow steering and activate/deactivate TIRs.
mlx5e_rx_res is self-contained and doesn't depend on struct mlx5e_priv
or include en.h.
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 14 -
.../net/ethernet/mellanox/mlx5/core/en/ptp.c | 2 +-
.../ethernet/mellanox/mlx5/core/en/rx_res.c | 748 ++++++++++++++++++
.../ethernet/mellanox/mlx5/core/en/rx_res.h | 72 +-
.../ethernet/mellanox/mlx5/core/en/xsk/pool.c | 4 +-
.../mellanox/mlx5/core/en/xsk/setup.c | 56 --
.../mellanox/mlx5/core/en/xsk/setup.h | 4 -
.../mellanox/mlx5/core/en_accel/ktls_rx.c | 2 +-
.../net/ethernet/mellanox/mlx5/core/en_arfs.c | 6 +-
.../ethernet/mellanox/mlx5/core/en_ethtool.c | 69 +-
.../net/ethernet/mellanox/mlx5/core/en_fs.c | 7 +-
.../mellanox/mlx5/core/en_fs_ethtool.c | 19 +-
.../net/ethernet/mellanox/mlx5/core/en_main.c | 491 +-----------
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 47 +-
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 9 +-
.../ethernet/mellanox/mlx5/core/ipoib/ipoib.c | 45 +-
16 files changed, 875 insertions(+), 720 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 87dabb8b8ac4..968e6a473cec 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -921,8 +921,6 @@ int mlx5e_vlan_rx_kill_vid(struct net_device *dev, __always_unused __be16 proto,
u16 vid);
void mlx5e_timestamp_init(struct mlx5e_priv *priv);
-int mlx5e_modify_tirs_hash(struct mlx5e_priv *priv);
-
struct mlx5e_xsk_param;
struct mlx5e_rq_param;
@@ -1033,16 +1031,6 @@ void mlx5e_close_drop_rq(struct mlx5e_rq *drop_rq);
int mlx5e_init_di_list(struct mlx5e_rq *rq, int wq_sz, int node);
void mlx5e_free_di_list(struct mlx5e_rq *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);
-
-int mlx5e_create_direct_rqts(struct mlx5e_priv *priv);
-void mlx5e_destroy_direct_rqts(struct mlx5e_priv *priv);
-int mlx5e_create_direct_tirs(struct mlx5e_priv *priv);
-void mlx5e_destroy_direct_tirs(struct mlx5e_priv *priv);
-
int mlx5e_create_tis(struct mlx5_core_dev *mdev, void *in, u32 *tisn);
void mlx5e_destroy_tis(struct mlx5_core_dev *mdev, u32 tisn);
@@ -1130,8 +1118,6 @@ int mlx5e_netdev_change_profile(struct mlx5e_priv *priv,
void mlx5e_netdev_attach_nic_profile(struct mlx5e_priv *priv);
void mlx5e_set_netdev_mtu_boundaries(struct mlx5e_priv *priv);
void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, u16 mtu);
-void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params,
- u16 num_channels);
void mlx5e_rx_dim_work(struct work_struct *work);
void mlx5e_tx_dim_work(struct work_struct *work);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
index 8ff8b02c056f..f9c96e5a7f54 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
@@ -605,8 +605,8 @@ static void mlx5e_ptp_rx_unset_fs(struct mlx5e_priv *priv)
static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv)
{
+ u32 tirn = mlx5e_rx_res_get_tirn_ptp(priv->rx_res);
struct mlx5e_ptp_fs *ptp_fs = priv->fs.ptp_fs;
- u32 tirn = priv->rx_res->ptp.tir.tirn;
struct mlx5_flow_handle *rule;
int err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
index 8fc1dfc4e830..a6b3a9473405 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
@@ -2,6 +2,8 @@
/* Copyright (c) 2021, Mellanox Technologies inc. All rights reserved. */
#include "rx_res.h"
+#include "channels.h"
+#include "params.h"
static const struct mlx5e_rss_params_traffic_type rss_default_config[MLX5E_NUM_INDIR_TIRS] = {
[MLX5E_TT_IPV4_TCP] = {
@@ -62,6 +64,539 @@ mlx5e_rss_get_default_tt_config(enum mlx5e_traffic_types tt)
return rss_default_config[tt];
}
+struct mlx5e_rx_res {
+ struct mlx5_core_dev *mdev;
+ enum mlx5e_rx_res_features features;
+ unsigned int max_nch;
+ u32 drop_rqn;
+
+ struct {
+ struct mlx5e_rss_params_hash hash;
+ struct mlx5e_rss_params_indir indir;
+ u32 rx_hash_fields[MLX5E_NUM_INDIR_TIRS];
+ } rss_params;
+
+ struct mlx5e_rqt indir_rqt;
+ struct {
+ struct mlx5e_tir indir_tir;
+ struct mlx5e_tir inner_indir_tir;
+ } rss[MLX5E_NUM_INDIR_TIRS];
+
+ bool rss_active;
+ u32 rss_rqns[MLX5E_INDIR_RQT_SIZE];
+ unsigned int rss_nch;
+
+ struct {
+ struct mlx5e_rqt direct_rqt;
+ struct mlx5e_tir direct_tir;
+ struct mlx5e_rqt xsk_rqt;
+ struct mlx5e_tir xsk_tir;
+ } channels[MLX5E_MAX_NUM_CHANNELS];
+
+ struct {
+ struct mlx5e_rqt rqt;
+ struct mlx5e_tir tir;
+ } ptp;
+};
+
+struct mlx5e_rx_res *mlx5e_rx_res_alloc(void)
+{
+ return kvzalloc(sizeof(struct mlx5e_rx_res), GFP_KERNEL);
+}
+
+static void mlx5e_rx_res_rss_params_init(struct mlx5e_rx_res *res, unsigned int init_nch)
+{
+ enum mlx5e_traffic_types tt;
+
+ res->rss_params.hash.hfunc = ETH_RSS_HASH_TOP;
+ netdev_rss_key_fill(res->rss_params.hash.toeplitz_hash_key,
+ sizeof(res->rss_params.hash.toeplitz_hash_key));
+ mlx5e_rss_params_indir_init_uniform(&res->rss_params.indir, init_nch);
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
+ res->rss_params.rx_hash_fields[tt] =
+ mlx5e_rss_get_default_tt_config(tt).rx_hash_fields;
+}
+
+static int mlx5e_rx_res_rss_init(struct mlx5e_rx_res *res,
+ const struct mlx5e_lro_param *init_lro_param)
+{
+ bool inner_ft_support = res->features & MLX5E_RX_RES_FEATURE_INNER_FT;
+ enum mlx5e_traffic_types tt, max_tt;
+ struct mlx5e_tir_builder *builder;
+ u32 indir_rqtn;
+ int err;
+
+ builder = mlx5e_tir_builder_alloc(false);
+ if (!builder)
+ return -ENOMEM;
+
+ err = mlx5e_rqt_init_direct(&res->indir_rqt, res->mdev, true, res->drop_rqn);
+ if (err)
+ goto out;
+
+ indir_rqtn = mlx5e_rqt_get_rqtn(&res->indir_rqt);
+
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
+ struct mlx5e_rss_params_traffic_type rss_tt;
+
+ mlx5e_tir_builder_build_rqt(builder, res->mdev->mlx5e_res.hw_objs.td.tdn,
+ indir_rqtn, inner_ft_support);
+ mlx5e_tir_builder_build_lro(builder, init_lro_param);
+ rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
+ mlx5e_tir_builder_build_rss(builder, &res->rss_params.hash, &rss_tt, false);
+
+ err = mlx5e_tir_init(&res->rss[tt].indir_tir, builder, res->mdev, true);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to create an indirect TIR: err = %d, tt = %d\n",
+ err, tt);
+ goto err_destroy_tirs;
+ }
+
+ mlx5e_tir_builder_clear(builder);
+ }
+
+ if (!inner_ft_support)
+ goto out;
+
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
+ struct mlx5e_rss_params_traffic_type rss_tt;
+
+ mlx5e_tir_builder_build_rqt(builder, res->mdev->mlx5e_res.hw_objs.td.tdn,
+ indir_rqtn, inner_ft_support);
+ mlx5e_tir_builder_build_lro(builder, init_lro_param);
+ rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
+ mlx5e_tir_builder_build_rss(builder, &res->rss_params.hash, &rss_tt, true);
+
+ err = mlx5e_tir_init(&res->rss[tt].inner_indir_tir, builder, res->mdev, true);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to create an inner indirect TIR: err = %d, tt = %d\n",
+ err, tt);
+ goto err_destroy_inner_tirs;
+ }
+
+ mlx5e_tir_builder_clear(builder);
+ }
+
+ goto out;
+
+err_destroy_inner_tirs:
+ max_tt = tt;
+ for (tt = 0; tt < max_tt; tt++)
+ mlx5e_tir_destroy(&res->rss[tt].inner_indir_tir);
+
+ tt = MLX5E_NUM_INDIR_TIRS;
+err_destroy_tirs:
+ max_tt = tt;
+ for (tt = 0; tt < max_tt; tt++)
+ mlx5e_tir_destroy(&res->rss[tt].indir_tir);
+
+ mlx5e_rqt_destroy(&res->indir_rqt);
+
+out:
+ mlx5e_tir_builder_free(builder);
+
+ return err;
+}
+
+static int mlx5e_rx_res_channels_init(struct mlx5e_rx_res *res,
+ const struct mlx5e_lro_param *init_lro_param)
+{
+ bool inner_ft_support = res->features & MLX5E_RX_RES_FEATURE_INNER_FT;
+ struct mlx5e_tir_builder *builder;
+ int err = 0;
+ int ix;
+
+ builder = mlx5e_tir_builder_alloc(false);
+ if (!builder)
+ return -ENOMEM;
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ err = mlx5e_rqt_init_direct(&res->channels[ix].direct_rqt,
+ res->mdev, false, res->drop_rqn);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to create a direct RQT: err = %d, ix = %u\n",
+ err, ix);
+ goto err_destroy_direct_rqts;
+ }
+ }
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ mlx5e_tir_builder_build_rqt(builder, res->mdev->mlx5e_res.hw_objs.td.tdn,
+ mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt),
+ inner_ft_support);
+ mlx5e_tir_builder_build_lro(builder, init_lro_param);
+ mlx5e_tir_builder_build_direct(builder);
+
+ err = mlx5e_tir_init(&res->channels[ix].direct_tir, builder, res->mdev, true);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to create a direct TIR: err = %d, ix = %u\n",
+ err, ix);
+ goto err_destroy_direct_tirs;
+ }
+
+ mlx5e_tir_builder_clear(builder);
+ }
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_XSK))
+ goto out;
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ err = mlx5e_rqt_init_direct(&res->channels[ix].xsk_rqt,
+ res->mdev, false, res->drop_rqn);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to create an XSK RQT: err = %d, ix = %u\n",
+ err, ix);
+ goto err_destroy_xsk_rqts;
+ }
+ }
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ mlx5e_tir_builder_build_rqt(builder, res->mdev->mlx5e_res.hw_objs.td.tdn,
+ mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt),
+ inner_ft_support);
+ mlx5e_tir_builder_build_lro(builder, init_lro_param);
+ mlx5e_tir_builder_build_direct(builder);
+
+ err = mlx5e_tir_init(&res->channels[ix].xsk_tir, builder, res->mdev, true);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to create an XSK TIR: err = %d, ix = %u\n",
+ err, ix);
+ goto err_destroy_xsk_tirs;
+ }
+
+ mlx5e_tir_builder_clear(builder);
+ }
+
+ goto out;
+
+err_destroy_xsk_tirs:
+ while (--ix >= 0)
+ mlx5e_tir_destroy(&res->channels[ix].xsk_tir);
+
+ ix = res->max_nch;
+err_destroy_xsk_rqts:
+ while (--ix >= 0)
+ mlx5e_rqt_destroy(&res->channels[ix].xsk_rqt);
+
+ ix = res->max_nch;
+err_destroy_direct_tirs:
+ while (--ix >= 0)
+ mlx5e_tir_destroy(&res->channels[ix].direct_tir);
+
+ ix = res->max_nch;
+err_destroy_direct_rqts:
+ while (--ix >= 0)
+ mlx5e_rqt_destroy(&res->channels[ix].direct_rqt);
+
+out:
+ mlx5e_tir_builder_free(builder);
+
+ return err;
+}
+
+static int mlx5e_rx_res_ptp_init(struct mlx5e_rx_res *res)
+{
+ bool inner_ft_support = res->features & MLX5E_RX_RES_FEATURE_INNER_FT;
+ struct mlx5e_tir_builder *builder;
+ int err;
+
+ builder = mlx5e_tir_builder_alloc(false);
+ if (!builder)
+ return -ENOMEM;
+
+ err = mlx5e_rqt_init_direct(&res->ptp.rqt, res->mdev, false, res->drop_rqn);
+ if (err)
+ goto out;
+
+ mlx5e_tir_builder_build_rqt(builder, res->mdev->mlx5e_res.hw_objs.td.tdn,
+ mlx5e_rqt_get_rqtn(&res->ptp.rqt),
+ inner_ft_support);
+ mlx5e_tir_builder_build_direct(builder);
+
+ err = mlx5e_tir_init(&res->ptp.tir, builder, res->mdev, true);
+ if (err)
+ goto err_destroy_ptp_rqt;
+
+ goto out;
+
+err_destroy_ptp_rqt:
+ mlx5e_rqt_destroy(&res->ptp.rqt);
+
+out:
+ mlx5e_tir_builder_free(builder);
+ return err;
+}
+
+static void mlx5e_rx_res_rss_destroy(struct mlx5e_rx_res *res)
+{
+ enum mlx5e_traffic_types tt;
+
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
+ mlx5e_tir_destroy(&res->rss[tt].indir_tir);
+
+ if (res->features & MLX5E_RX_RES_FEATURE_INNER_FT)
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
+ mlx5e_tir_destroy(&res->rss[tt].inner_indir_tir);
+
+ mlx5e_rqt_destroy(&res->indir_rqt);
+}
+
+static void mlx5e_rx_res_channels_destroy(struct mlx5e_rx_res *res)
+{
+ unsigned int ix;
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ mlx5e_tir_destroy(&res->channels[ix].direct_tir);
+ mlx5e_rqt_destroy(&res->channels[ix].direct_rqt);
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_XSK))
+ continue;
+
+ mlx5e_tir_destroy(&res->channels[ix].xsk_tir);
+ mlx5e_rqt_destroy(&res->channels[ix].xsk_rqt);
+ }
+}
+
+static void mlx5e_rx_res_ptp_destroy(struct mlx5e_rx_res *res)
+{
+ mlx5e_tir_destroy(&res->ptp.tir);
+ mlx5e_rqt_destroy(&res->ptp.rqt);
+}
+
+int mlx5e_rx_res_init(struct mlx5e_rx_res *res, struct mlx5_core_dev *mdev,
+ enum mlx5e_rx_res_features features, unsigned int max_nch,
+ u32 drop_rqn, const struct mlx5e_lro_param *init_lro_param,
+ unsigned int init_nch)
+{
+ int err;
+
+ res->mdev = mdev;
+ res->features = features;
+ res->max_nch = max_nch;
+ res->drop_rqn = drop_rqn;
+
+ mlx5e_rx_res_rss_params_init(res, init_nch);
+
+ err = mlx5e_rx_res_rss_init(res, init_lro_param);
+ if (err)
+ return err;
+
+ err = mlx5e_rx_res_channels_init(res, init_lro_param);
+ if (err)
+ goto err_rss_destroy;
+
+ err = mlx5e_rx_res_ptp_init(res);
+ if (err)
+ goto err_channels_destroy;
+
+ return 0;
+
+err_channels_destroy:
+ mlx5e_rx_res_channels_destroy(res);
+err_rss_destroy:
+ mlx5e_rx_res_rss_destroy(res);
+ return err;
+}
+
+void mlx5e_rx_res_destroy(struct mlx5e_rx_res *res)
+{
+ mlx5e_rx_res_ptp_destroy(res);
+ mlx5e_rx_res_channels_destroy(res);
+ mlx5e_rx_res_rss_destroy(res);
+}
+
+void mlx5e_rx_res_free(struct mlx5e_rx_res *res)
+{
+ kvfree(res);
+}
+
+u32 mlx5e_rx_res_get_tirn_direct(struct mlx5e_rx_res *res, unsigned int ix)
+{
+ return mlx5e_tir_get_tirn(&res->channels[ix].direct_tir);
+}
+
+u32 mlx5e_rx_res_get_tirn_xsk(struct mlx5e_rx_res *res, unsigned int ix)
+{
+ WARN_ON(!(res->features & MLX5E_RX_RES_FEATURE_XSK));
+
+ return mlx5e_tir_get_tirn(&res->channels[ix].xsk_tir);
+}
+
+u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+{
+ return mlx5e_tir_get_tirn(&res->rss[tt].indir_tir);
+}
+
+u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+{
+ WARN_ON(!(res->features & MLX5E_RX_RES_FEATURE_INNER_FT));
+ return mlx5e_tir_get_tirn(&res->rss[tt].inner_indir_tir);
+}
+
+u32 mlx5e_rx_res_get_tirn_ptp(struct mlx5e_rx_res *res)
+{
+ WARN_ON(!(res->features & MLX5E_RX_RES_FEATURE_PTP));
+ return mlx5e_tir_get_tirn(&res->ptp.tir);
+}
+
+u32 mlx5e_rx_res_get_rqtn_direct(struct mlx5e_rx_res *res, unsigned int ix)
+{
+ return mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt);
+}
+
+static void mlx5e_rx_res_rss_enable(struct mlx5e_rx_res *res)
+{
+ int err;
+
+ res->rss_active = true;
+
+ err = mlx5e_rqt_redirect_indir(&res->indir_rqt, res->rss_rqns, res->rss_nch,
+ res->rss_params.hash.hfunc,
+ &res->rss_params.indir);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect indirect RQT %#x to channels: err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->indir_rqt), err);
+}
+
+static void mlx5e_rx_res_rss_disable(struct mlx5e_rx_res *res)
+{
+ int err;
+
+ res->rss_active = false;
+
+ err = mlx5e_rqt_redirect_direct(&res->indir_rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect indirect RQT %#x to drop RQ %#x: err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->indir_rqt), res->drop_rqn, err);
+}
+
+void mlx5e_rx_res_channels_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs)
+{
+ unsigned int nch, ix;
+ int err;
+
+ nch = mlx5e_channels_get_num(chs);
+
+ for (ix = 0; ix < chs->num; ix++)
+ mlx5e_channels_get_regular_rqn(chs, ix, &res->rss_rqns[ix]);
+ res->rss_nch = chs->num;
+
+ mlx5e_rx_res_rss_enable(res);
+
+ for (ix = 0; ix < nch; ix++) {
+ u32 rqn;
+
+ mlx5e_channels_get_regular_rqn(chs, ix, &rqn);
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt),
+ rqn, ix, err);
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_XSK))
+ continue;
+
+ if (!mlx5e_channels_get_xsk_rqn(chs, ix, &rqn))
+ rqn = res->drop_rqn;
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt),
+ rqn, ix, err);
+ }
+ for (ix = nch; ix < res->max_nch; ix++) {
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to drop RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt),
+ res->drop_rqn, ix, err);
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_XSK))
+ continue;
+
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to drop RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt),
+ res->drop_rqn, ix, err);
+ }
+
+ if (res->features & MLX5E_RX_RES_FEATURE_PTP) {
+ u32 rqn;
+
+ if (mlx5e_channels_get_ptp_rqn(chs, &rqn))
+ rqn = res->drop_rqn;
+
+ err = mlx5e_rqt_redirect_direct(&res->ptp.rqt, rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to RQ %#x (PTP): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->ptp.rqt),
+ rqn, err);
+ }
+}
+
+void mlx5e_rx_res_channels_deactivate(struct mlx5e_rx_res *res)
+{
+ unsigned int ix;
+ int err;
+
+ mlx5e_rx_res_rss_disable(res);
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to drop RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt),
+ res->drop_rqn, ix, err);
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_XSK))
+ continue;
+
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to drop RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt),
+ res->drop_rqn, ix, err);
+ }
+
+ if (res->features & MLX5E_RX_RES_FEATURE_PTP) {
+ err = mlx5e_rqt_redirect_direct(&res->ptp.rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to drop RQ %#x (PTP): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->ptp.rqt),
+ res->drop_rqn, err);
+ }
+}
+
+int mlx5e_rx_res_xsk_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs,
+ unsigned int ix)
+{
+ u32 rqn;
+ int err;
+
+ if (!mlx5e_channels_get_xsk_rqn(chs, ix, &rqn))
+ return -EINVAL;
+
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to XSK RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt),
+ rqn, ix, err);
+ return err;
+}
+
+int mlx5e_rx_res_xsk_deactivate(struct mlx5e_rx_res *res, unsigned int ix)
+{
+ int err;
+
+ err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, res->drop_rqn);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to drop RQ %#x (channel %u): err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt),
+ res->drop_rqn, ix, err);
+ return err;
+}
+
struct mlx5e_rss_params_traffic_type
mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
{
@@ -71,3 +606,216 @@ mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5e_traf
rss_tt.rx_hash_fields = res->rss_params.rx_hash_fields[tt];
return rss_tt;
}
+
+void mlx5e_rx_res_rss_set_indir_uniform(struct mlx5e_rx_res *res, unsigned int nch)
+{
+ mlx5e_rss_params_indir_init_uniform(&res->rss_params.indir, nch);
+
+ if (!res->rss_active)
+ return;
+
+ mlx5e_rx_res_rss_enable(res);
+}
+
+void mlx5e_rx_res_rss_get_rxfh(struct mlx5e_rx_res *res, u32 *indir, u8 *key, u8 *hfunc)
+{
+ unsigned int i;
+
+ if (indir)
+ for (i = 0; i < MLX5E_INDIR_RQT_SIZE; i++)
+ indir[i] = res->rss_params.indir.table[i];
+
+ if (key)
+ memcpy(key, res->rss_params.hash.toeplitz_hash_key,
+ sizeof(res->rss_params.hash.toeplitz_hash_key));
+
+ if (hfunc)
+ *hfunc = res->rss_params.hash.hfunc;
+}
+
+static int mlx5e_rx_res_rss_update_tir(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt,
+ bool inner)
+{
+ struct mlx5e_rss_params_traffic_type rss_tt;
+ struct mlx5e_tir_builder *builder;
+ struct mlx5e_tir *tir;
+ int err;
+
+ builder = mlx5e_tir_builder_alloc(true);
+ if (!builder)
+ return -ENOMEM;
+
+ rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
+
+ mlx5e_tir_builder_build_rss(builder, &res->rss_params.hash, &rss_tt, inner);
+ tir = inner ? &res->rss[tt].inner_indir_tir : &res->rss[tt].indir_tir;
+ err = mlx5e_tir_modify(tir, builder);
+
+ mlx5e_tir_builder_free(builder);
+ return err;
+}
+
+int mlx5e_rx_res_rss_set_rxfh(struct mlx5e_rx_res *res, const u32 *indir,
+ const u8 *key, const u8 *hfunc)
+{
+ enum mlx5e_traffic_types tt;
+ bool changed_indir = false;
+ bool changed_hash = false;
+ int err;
+
+ if (hfunc && *hfunc != res->rss_params.hash.hfunc) {
+ switch (*hfunc) {
+ case ETH_RSS_HASH_XOR:
+ case ETH_RSS_HASH_TOP:
+ break;
+ default:
+ return -EINVAL;
+ }
+ changed_hash = true;
+ changed_indir = true;
+ res->rss_params.hash.hfunc = *hfunc;
+ }
+
+ if (key) {
+ if (res->rss_params.hash.hfunc == ETH_RSS_HASH_TOP)
+ changed_hash = true;
+ memcpy(res->rss_params.hash.toeplitz_hash_key, key,
+ sizeof(res->rss_params.hash.toeplitz_hash_key));
+ }
+
+ if (indir) {
+ unsigned int i;
+
+ changed_indir = true;
+
+ for (i = 0; i < MLX5E_INDIR_RQT_SIZE; i++)
+ res->rss_params.indir.table[i] = indir[i];
+ }
+
+ if (changed_indir && res->rss_active) {
+ err = mlx5e_rqt_redirect_indir(&res->indir_rqt, res->rss_rqns, res->rss_nch,
+ res->rss_params.hash.hfunc,
+ &res->rss_params.indir);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to redirect indirect RQT %#x to channels: err = %d\n",
+ mlx5e_rqt_get_rqtn(&res->indir_rqt), err);
+ }
+
+ if (changed_hash)
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
+ err = mlx5e_rx_res_rss_update_tir(res, tt, false);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to update RSS hash of indirect TIR for traffic type %d: err = %d\n",
+ tt, err);
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_INNER_FT))
+ continue;
+
+ err = mlx5e_rx_res_rss_update_tir(res, tt, true);
+ if (err)
+ mlx5_core_warn(res->mdev, "Failed to update RSS hash of inner indirect TIR for traffic type %d: err = %d\n",
+ tt, err);
+ }
+
+ return 0;
+}
+
+u8 mlx5e_rx_res_rss_get_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+{
+ return res->rss_params.rx_hash_fields[tt];
+}
+
+int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt,
+ u8 rx_hash_fields)
+{
+ u8 old_rx_hash_fields;
+ int err;
+
+ old_rx_hash_fields = res->rss_params.rx_hash_fields[tt];
+
+ if (old_rx_hash_fields == rx_hash_fields)
+ return 0;
+
+ res->rss_params.rx_hash_fields[tt] = rx_hash_fields;
+
+ err = mlx5e_rx_res_rss_update_tir(res, tt, false);
+ if (err) {
+ res->rss_params.rx_hash_fields[tt] = old_rx_hash_fields;
+ mlx5_core_warn(res->mdev, "Failed to update RSS hash fields of indirect TIR for traffic type %d: err = %d\n",
+ tt, err);
+ return err;
+ }
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_INNER_FT))
+ return 0;
+
+ err = mlx5e_rx_res_rss_update_tir(res, tt, true);
+ if (err) {
+ /* Partial update happened. Try to revert - it may fail too, but
+ * there is nothing more we can do.
+ */
+ res->rss_params.rx_hash_fields[tt] = old_rx_hash_fields;
+ mlx5_core_warn(res->mdev, "Failed to update RSS hash fields of inner indirect TIR for traffic type %d: err = %d\n",
+ tt, err);
+ if (mlx5e_rx_res_rss_update_tir(res, tt, false))
+ mlx5_core_warn(res->mdev, "Partial update of RSS hash fields happened: failed to revert indirect TIR for traffic type %d to the old values\n",
+ tt);
+ }
+
+ return err;
+}
+
+int mlx5e_rx_res_lro_set_param(struct mlx5e_rx_res *res, struct mlx5e_lro_param *lro_param)
+{
+ struct mlx5e_tir_builder *builder;
+ enum mlx5e_traffic_types tt;
+ int err, final_err;
+ unsigned int ix;
+
+ builder = mlx5e_tir_builder_alloc(true);
+ if (!builder)
+ return -ENOMEM;
+
+ mlx5e_tir_builder_build_lro(builder, lro_param);
+
+ final_err = 0;
+
+ for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
+ err = mlx5e_tir_modify(&res->rss[tt].indir_tir, builder);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to update LRO state of indirect TIR %#x for traffic type %d: err = %d\n",
+ mlx5e_tir_get_tirn(&res->rss[tt].indir_tir), tt, err);
+ if (!final_err)
+ final_err = err;
+ }
+
+ if (!(res->features & MLX5E_RX_RES_FEATURE_INNER_FT))
+ continue;
+
+ err = mlx5e_tir_modify(&res->rss[tt].inner_indir_tir, builder);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to update LRO state of inner indirect TIR %#x for traffic type %d: err = %d\n",
+ mlx5e_tir_get_tirn(&res->rss[tt].inner_indir_tir), tt, err);
+ if (!final_err)
+ final_err = err;
+ }
+ }
+
+ for (ix = 0; ix < res->max_nch; ix++) {
+ err = mlx5e_tir_modify(&res->channels[ix].direct_tir, builder);
+ if (err) {
+ mlx5_core_warn(res->mdev, "Failed to update LRO state of direct TIR %#x for channel %u: err = %d\n",
+ mlx5e_tir_get_tirn(&res->channels[ix].direct_tir), ix, err);
+ if (!final_err)
+ final_err = err;
+ }
+ }
+
+ mlx5e_tir_builder_free(builder);
+ return final_err;
+}
+
+struct mlx5e_rss_params_hash mlx5e_rx_res_get_current_hash(struct mlx5e_rx_res *res)
+{
+ return res->rss_params.hash;
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
index 068e48140a6f..0092ee80a2cf 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
@@ -11,37 +11,59 @@
#define MLX5E_MAX_NUM_CHANNELS (MLX5E_INDIR_RQT_SIZE / 2)
-struct mlx5e_rss_params {
- struct mlx5e_rss_params_hash hash;
- struct mlx5e_rss_params_indir indir;
- u32 rx_hash_fields[MLX5E_NUM_INDIR_TIRS];
-};
+struct mlx5e_rx_res;
+
+struct mlx5e_channels;
+struct mlx5e_rss_params_hash;
-struct mlx5e_rx_res {
- struct mlx5e_rss_params rss_params;
-
- struct mlx5e_rqt indir_rqt;
- struct {
- struct mlx5e_tir indir_tir;
- struct mlx5e_tir inner_indir_tir;
- } rss[MLX5E_NUM_INDIR_TIRS];
-
- struct {
- struct mlx5e_rqt direct_rqt;
- struct mlx5e_tir direct_tir;
- struct mlx5e_rqt xsk_rqt;
- struct mlx5e_tir xsk_tir;
- } channels[MLX5E_MAX_NUM_CHANNELS];
-
- struct {
- struct mlx5e_rqt rqt;
- struct mlx5e_tir tir;
- } ptp;
+enum mlx5e_rx_res_features {
+ MLX5E_RX_RES_FEATURE_INNER_FT = BIT(0),
+ MLX5E_RX_RES_FEATURE_XSK = BIT(1),
+ MLX5E_RX_RES_FEATURE_PTP = BIT(2),
};
struct mlx5e_rss_params_traffic_type
mlx5e_rss_get_default_tt_config(enum mlx5e_traffic_types tt);
+
+/* Setup */
+struct mlx5e_rx_res *mlx5e_rx_res_alloc(void);
+int mlx5e_rx_res_init(struct mlx5e_rx_res *res, struct mlx5_core_dev *mdev,
+ enum mlx5e_rx_res_features features, unsigned int max_nch,
+ u32 drop_rqn, const struct mlx5e_lro_param *init_lro_param,
+ unsigned int init_nch);
+void mlx5e_rx_res_destroy(struct mlx5e_rx_res *res);
+void mlx5e_rx_res_free(struct mlx5e_rx_res *res);
+
+/* TIRN getters for flow steering */
+u32 mlx5e_rx_res_get_tirn_direct(struct mlx5e_rx_res *res, unsigned int ix);
+u32 mlx5e_rx_res_get_tirn_xsk(struct mlx5e_rx_res *res, unsigned int ix);
+u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
+u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
+u32 mlx5e_rx_res_get_tirn_ptp(struct mlx5e_rx_res *res);
+
+/* RQTN getters for modules that create their own TIRs */
+u32 mlx5e_rx_res_get_rqtn_direct(struct mlx5e_rx_res *res, unsigned int ix);
+
+/* Activate/deactivate API */
+void mlx5e_rx_res_channels_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs);
+void mlx5e_rx_res_channels_deactivate(struct mlx5e_rx_res *res);
+int mlx5e_rx_res_xsk_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs,
+ unsigned int ix);
+int mlx5e_rx_res_xsk_deactivate(struct mlx5e_rx_res *res, unsigned int ix);
+
+/* Configuration API */
struct mlx5e_rss_params_traffic_type
mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
+void mlx5e_rx_res_rss_set_indir_uniform(struct mlx5e_rx_res *res, unsigned int nch);
+void mlx5e_rx_res_rss_get_rxfh(struct mlx5e_rx_res *res, u32 *indir, u8 *key, u8 *hfunc);
+int mlx5e_rx_res_rss_set_rxfh(struct mlx5e_rx_res *res, const u32 *indir,
+ const u8 *key, const u8 *hfunc);
+u8 mlx5e_rx_res_rss_get_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
+int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt,
+ u8 rx_hash_fields);
+int mlx5e_rx_res_lro_set_param(struct mlx5e_rx_res *res, struct mlx5e_lro_param *lro_param);
+
+/* Workaround for hairpin */
+struct mlx5e_rss_params_hash mlx5e_rx_res_get_current_hash(struct mlx5e_rx_res *res);
#endif /* __MLX5_EN_RX_RES_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c
index 71e8d66fa150..7b562d2c8a19 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c
@@ -122,7 +122,7 @@ static int mlx5e_xsk_enable_locked(struct mlx5e_priv *priv,
* any Fill Ring entries at the setup stage.
*/
- err = mlx5e_xsk_redirect_rqt_to_channel(priv, priv->channels.c[ix]);
+ err = mlx5e_rx_res_xsk_activate(priv->rx_res, &priv->channels, ix);
if (unlikely(err))
goto err_deactivate;
@@ -169,7 +169,7 @@ static int mlx5e_xsk_disable_locked(struct mlx5e_priv *priv, u16 ix)
goto remove_pool;
c = priv->channels.c[ix];
- mlx5e_xsk_redirect_rqt_to_drop(priv, ix);
+ mlx5e_rx_res_xsk_deactivate(priv->rx_res, ix);
mlx5e_deactivate_xsk(c);
mlx5e_close_xsk(c);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
index ab485d082729..c06267477b27 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
@@ -183,59 +183,3 @@ void mlx5e_deactivate_xsk(struct mlx5e_channel *c)
mlx5e_deactivate_rq(&c->xskrq);
/* TX queue is disabled on close. */
}
-
-int mlx5e_xsk_redirect_rqt_to_channel(struct mlx5e_priv *priv, struct mlx5e_channel *c)
-{
- return mlx5e_rqt_redirect_direct(&priv->rx_res->channels[c->ix].xsk_rqt, c->xskrq.rqn);
-}
-
-int mlx5e_xsk_redirect_rqt_to_drop(struct mlx5e_priv *priv, u16 ix)
-{
- return mlx5e_rqt_redirect_direct(&priv->rx_res->channels[ix].xsk_rqt, priv->drop_rq.rqn);
-}
-
-int mlx5e_xsk_redirect_rqts_to_channels(struct mlx5e_priv *priv, struct mlx5e_channels *chs)
-{
- int err, i;
-
- if (!priv->xsk.refcnt)
- return 0;
-
- for (i = 0; i < chs->num; i++) {
- struct mlx5e_channel *c = chs->c[i];
-
- if (!test_bit(MLX5E_CHANNEL_STATE_XSK, c->state))
- continue;
-
- err = mlx5e_xsk_redirect_rqt_to_channel(priv, c);
- if (unlikely(err))
- goto err_stop;
- }
-
- return 0;
-
-err_stop:
- for (i--; i >= 0; i--) {
- if (!test_bit(MLX5E_CHANNEL_STATE_XSK, chs->c[i]->state))
- continue;
-
- mlx5e_xsk_redirect_rqt_to_drop(priv, i);
- }
-
- return err;
-}
-
-void mlx5e_xsk_redirect_rqts_to_drop(struct mlx5e_priv *priv, struct mlx5e_channels *chs)
-{
- int i;
-
- if (!priv->xsk.refcnt)
- return;
-
- for (i = 0; i < chs->num; i++) {
- if (!test_bit(MLX5E_CHANNEL_STATE_XSK, chs->c[i]->state))
- continue;
-
- mlx5e_xsk_redirect_rqt_to_drop(priv, i);
- }
-}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.h
index ca20f1ff5e39..50e111b85efd 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.h
@@ -17,9 +17,5 @@ int mlx5e_open_xsk(struct mlx5e_priv *priv, struct mlx5e_params *params,
void mlx5e_close_xsk(struct mlx5e_channel *c);
void mlx5e_activate_xsk(struct mlx5e_channel *c);
void mlx5e_deactivate_xsk(struct mlx5e_channel *c);
-int mlx5e_xsk_redirect_rqt_to_channel(struct mlx5e_priv *priv, struct mlx5e_channel *c);
-int mlx5e_xsk_redirect_rqt_to_drop(struct mlx5e_priv *priv, u16 ix);
-int mlx5e_xsk_redirect_rqts_to_channels(struct mlx5e_priv *priv, struct mlx5e_channels *chs);
-void mlx5e_xsk_redirect_rqts_to_drop(struct mlx5e_priv *priv, struct mlx5e_channels *chs);
#endif /* __MLX5_EN_XSK_SETUP_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c
index bfdbc3060755..62abce008c7b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c
@@ -628,7 +628,7 @@ int mlx5e_ktls_add_rx(struct net_device *netdev, struct sock *sk,
priv_rx->sw_stats = &priv->tls->sw_stats;
mlx5e_set_ktls_rx_priv_ctx(tls_ctx, priv_rx);
- rqtn = mlx5e_rqt_get_rqtn(&priv->rx_res->channels[rxq].direct_rqt);
+ rqtn = mlx5e_rx_res_get_rqtn_direct(priv->rx_res, rxq);
err = mlx5e_ktls_create_tir(mdev, &priv_rx->tir, rqtn);
if (err)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
index db6c6a96a6c9..5077367f3ea0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
@@ -208,7 +208,7 @@ static int arfs_add_default_rule(struct mlx5e_priv *priv,
/* FIXME: Must use mlx5e_ttc_get_default_dest(),
* but can't since TTC default is not setup yet !
*/
- dest.tir_num = priv->rx_res->rss[tt].indir_tir.tirn;
+ dest.tir_num = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
arfs_t->default_rule = mlx5_add_flow_rules(arfs_t->ft.t, NULL,
&flow_act,
&dest, 1);
@@ -552,7 +552,7 @@ static struct mlx5_flow_handle *arfs_add_rule(struct mlx5e_priv *priv,
16);
}
dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
- dest.tir_num = priv->rx_res->channels[arfs_rule->rxq].direct_tir.tirn;
+ dest.tir_num = mlx5e_rx_res_get_tirn_direct(priv->rx_res, arfs_rule->rxq);
rule = mlx5_add_flow_rules(ft, spec, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -575,7 +575,7 @@ static void arfs_modify_rule_rq(struct mlx5e_priv *priv,
int err = 0;
dst.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
- dst.tir_num = priv->rx_res->channels[rxq].direct_tir.tirn;
+ dst.tir_num = mlx5e_rx_res_get_tirn_direct(priv->rx_res, rxq);
err = mlx5_modify_rule_destination(rule, &dst, NULL);
if (err)
netdev_warn(priv->netdev,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index 9264d18b0964..2cf59bb5f898 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -1172,7 +1172,7 @@ static int mlx5e_set_link_ksettings(struct net_device *netdev,
u32 mlx5e_ethtool_get_rxfh_key_size(struct mlx5e_priv *priv)
{
- return sizeof(priv->rx_res->rss_params.hash.toeplitz_hash_key);
+ return sizeof_field(struct mlx5e_rss_params_hash, toeplitz_hash_key);
}
static u32 mlx5e_get_rxfh_key_size(struct net_device *netdev)
@@ -1198,18 +1198,10 @@ int mlx5e_get_rxfh(struct net_device *netdev, u32 *indir, u8 *key,
u8 *hfunc)
{
struct mlx5e_priv *priv = netdev_priv(netdev);
- struct mlx5e_rss_params *rss;
- rss = &priv->rx_res->rss_params;
-
- if (indir)
- memcpy(indir, rss->indir.table, sizeof(rss->indir.table));
-
- if (key)
- memcpy(key, rss->hash.toeplitz_hash_key, sizeof(rss->hash.toeplitz_hash_key));
-
- if (hfunc)
- *hfunc = rss->hash.hfunc;
+ mutex_lock(&priv->state_lock);
+ mlx5e_rx_res_rss_get_rxfh(priv->rx_res, indir, key, hfunc);
+ mutex_unlock(&priv->state_lock);
return 0;
}
@@ -1218,58 +1210,13 @@ int mlx5e_set_rxfh(struct net_device *dev, const u32 *indir,
const u8 *key, const u8 hfunc)
{
struct mlx5e_priv *priv = netdev_priv(dev);
- struct mlx5e_rss_params *rss;
- bool refresh_tirs = false;
- bool refresh_rqt = false;
-
- if ((hfunc != ETH_RSS_HASH_NO_CHANGE) &&
- (hfunc != ETH_RSS_HASH_XOR) &&
- (hfunc != ETH_RSS_HASH_TOP))
- return -EINVAL;
+ int err;
mutex_lock(&priv->state_lock);
-
- rss = &priv->rx_res->rss_params;
-
- if (hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != rss->hash.hfunc) {
- rss->hash.hfunc = hfunc;
- refresh_rqt = true;
- refresh_tirs = true;
- }
-
- if (indir) {
- memcpy(rss->indir.table, indir, sizeof(rss->indir.table));
- refresh_rqt = true;
- }
-
- if (key) {
- memcpy(rss->hash.toeplitz_hash_key, key, sizeof(rss->hash.toeplitz_hash_key));
- refresh_tirs = refresh_tirs || rss->hash.hfunc == ETH_RSS_HASH_TOP;
- }
-
- if (refresh_rqt && test_bit(MLX5E_STATE_OPENED, &priv->state)) {
- u32 *rqns;
-
- rqns = kvmalloc_array(priv->channels.num, sizeof(*rqns), GFP_KERNEL);
- if (rqns) {
- unsigned int ix;
-
- for (ix = 0; ix < priv->channels.num; ix++)
- rqns[ix] = priv->channels.c[ix]->rq.rqn;
-
- mlx5e_rqt_redirect_indir(&priv->rx_res->indir_rqt, rqns,
- priv->channels.num,
- rss->hash.hfunc, &rss->indir);
- kvfree(rqns);
- }
- }
-
- if (refresh_tirs)
- mlx5e_modify_tirs_hash(priv);
-
+ err = mlx5e_rx_res_rss_set_rxfh(priv->rx_res, indir, key,
+ hfunc == ETH_RSS_HASH_NO_CHANGE ? NULL : &hfunc);
mutex_unlock(&priv->state_lock);
-
- return 0;
+ return err;
}
#define MLX5E_PFC_PREVEN_AUTO_TOUT_MSEC 100
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index e79815763edf..776f73cb592b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -1320,7 +1320,7 @@ static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
void mlx5e_set_ttc_basic_params(struct mlx5e_priv *priv,
struct ttc_params *ttc_params)
{
- ttc_params->any_tt_tirn = priv->rx_res->channels[0].direct_tir.tirn;
+ ttc_params->any_tt_tirn = mlx5e_rx_res_get_tirn_direct(priv->rx_res, 0);
ttc_params->inner_ttc = &priv->fs.inner_ttc;
}
@@ -1786,7 +1786,8 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
if (mlx5e_tunnel_inner_ft_supported(priv->mdev)) {
mlx5e_set_inner_ttc_ft_params(&ttc_params);
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = priv->rx_res->rss[tt].inner_indir_tir.tirn;
+ ttc_params.indir_tirn[tt] =
+ mlx5e_rx_res_get_tirn_rss_inner(priv->rx_res, tt);
err = mlx5e_create_inner_ttc_table(priv, &ttc_params, &priv->fs.inner_ttc);
if (err) {
@@ -1798,7 +1799,7 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
mlx5e_set_ttc_ft_params(&ttc_params);
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = priv->rx_res->rss[tt].indir_tir.tirn;
+ ttc_params.indir_tirn[tt] = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
if (err) {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
index 494f6f832407..c057f830a15d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
@@ -433,9 +433,9 @@ add_ethtool_flow_rule(struct mlx5e_priv *priv,
dst->type = MLX5_FLOW_DESTINATION_TYPE_TIR;
if (group == MLX5E_RQ_GROUP_XSK)
- dst->tir_num = priv->rx_res->channels[ix].xsk_tir.tirn;
+ dst->tir_num = mlx5e_rx_res_get_tirn_xsk(priv->rx_res, ix);
else
- dst->tir_num = priv->rx_res->channels[ix].direct_tir.tirn;
+ dst->tir_num = mlx5e_rx_res_get_tirn_direct(priv->rx_res, ix);
flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
}
@@ -819,6 +819,7 @@ static int mlx5e_set_rss_hash_opt(struct mlx5e_priv *priv,
{
enum mlx5e_traffic_types tt;
u8 rx_hash_field = 0;
+ int err;
tt = flow_type_to_traffic_type(nfc->flow_type);
if (tt == MLX5E_NUM_INDIR_TIRS)
@@ -848,16 +849,10 @@ static int mlx5e_set_rss_hash_opt(struct mlx5e_priv *priv,
rx_hash_field |= MLX5_HASH_FIELD_SEL_L4_DPORT;
mutex_lock(&priv->state_lock);
-
- if (rx_hash_field == priv->rx_res->rss_params.rx_hash_fields[tt])
- goto out;
-
- priv->rx_res->rss_params.rx_hash_fields[tt] = rx_hash_field;
- mlx5e_modify_tirs_hash(priv);
-
-out:
+ err = mlx5e_rx_res_rss_set_hash_fields(priv->rx_res, tt, rx_hash_field);
mutex_unlock(&priv->state_lock);
- return 0;
+
+ return err;
}
static int mlx5e_get_rss_hash_opt(struct mlx5e_priv *priv,
@@ -870,7 +865,7 @@ static int mlx5e_get_rss_hash_opt(struct mlx5e_priv *priv,
if (tt == MLX5E_NUM_INDIR_TIRS)
return -EINVAL;
- hash_field = priv->rx_res->rss_params.rx_hash_fields[tt];
+ hash_field = mlx5e_rx_res_rss_get_hash_fields(priv->rx_res, tt);
nfc->data = 0;
if (hash_field & MLX5_HASH_FIELD_SEL_SRC_IP)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 68be4e0e77bf..6797328e0afd 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -2194,202 +2194,14 @@ void mlx5e_close_channels(struct mlx5e_channels *chs)
chs->num = 0;
}
-int mlx5e_create_indirect_rqt(struct mlx5e_priv *priv)
-{
- int err;
-
- err = mlx5e_rqt_init_direct(&priv->rx_res->indir_rqt, priv->mdev, true,
- priv->drop_rq.rqn);
- if (err)
- mlx5_core_warn(priv->mdev, "create indirect rqts failed, %d\n", err);
- return err;
-}
-
-int mlx5e_create_direct_rqts(struct mlx5e_priv *priv)
-{
- int err;
- int ix;
-
- for (ix = 0; ix < priv->max_nch; ix++) {
- err = mlx5e_rqt_init_direct(&priv->rx_res->channels[ix].direct_rqt,
- priv->mdev, false, priv->drop_rq.rqn);
- if (unlikely(err))
- goto err_destroy_rqts;
- }
-
- return 0;
-
-err_destroy_rqts:
- mlx5_core_warn(priv->mdev, "create direct rqts failed, %d\n", err);
- while (--ix >= 0)
- mlx5e_rqt_destroy(&priv->rx_res->channels[ix].direct_rqt);
-
- return err;
-}
-
-static int mlx5e_create_xsk_rqts(struct mlx5e_priv *priv)
-{
- int err;
- int ix;
-
- for (ix = 0; ix < priv->max_nch; ix++) {
- err = mlx5e_rqt_init_direct(&priv->rx_res->channels[ix].xsk_rqt,
- priv->mdev, false, priv->drop_rq.rqn);
- if (unlikely(err))
- goto err_destroy_rqts;
- }
-
- return 0;
-
-err_destroy_rqts:
- mlx5_core_warn(priv->mdev, "create xsk rqts failed, %d\n", err);
- while (--ix >= 0)
- mlx5e_rqt_destroy(&priv->rx_res->channels[ix].xsk_rqt);
-
- return err;
-}
-
-void mlx5e_destroy_direct_rqts(struct mlx5e_priv *priv)
-{
- unsigned int ix;
-
- for (ix = 0; ix < priv->max_nch; ix++)
- mlx5e_rqt_destroy(&priv->rx_res->channels[ix].direct_rqt);
-}
-
-static void mlx5e_destroy_xsk_rqts(struct mlx5e_priv *priv)
-{
- unsigned int ix;
-
- for (ix = 0; ix < priv->max_nch; ix++)
- mlx5e_rqt_destroy(&priv->rx_res->channels[ix].xsk_rqt);
-}
-
-static void mlx5e_redirect_rqts_to_channels(struct mlx5e_priv *priv,
- struct mlx5e_channels *chs)
-{
- struct mlx5e_rx_res *res = priv->rx_res;
- unsigned int ix;
- u32 *rqns;
-
- rqns = kvmalloc_array(chs->num, sizeof(*rqns), GFP_KERNEL);
- if (rqns) {
- for (ix = 0; ix < chs->num; ix++)
- rqns[ix] = chs->c[ix]->rq.rqn;
-
- mlx5e_rqt_redirect_indir(&res->indir_rqt, rqns, chs->num,
- res->rss_params.hash.hfunc,
- &res->rss_params.indir);
- kvfree(rqns);
- }
-
- for (ix = 0; ix < priv->max_nch; ix++) {
- u32 rqn = priv->drop_rq.rqn;
-
- if (ix < chs->num)
- rqn = chs->c[ix]->rq.rqn;
-
- mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, rqn);
- }
-
- if (priv->profile->rx_ptp_support) {
- u32 rqn;
-
- if (mlx5e_ptp_get_rqn(priv->channels.ptp, &rqn))
- rqn = priv->drop_rq.rqn;
-
- mlx5e_rqt_redirect_direct(&res->ptp.rqt, rqn);
- }
-}
-
-static void mlx5e_redirect_rqts_to_drop(struct mlx5e_priv *priv)
-{
- struct mlx5e_rx_res *res = priv->rx_res;
- unsigned int ix;
-
- mlx5e_rqt_redirect_direct(&res->indir_rqt, priv->drop_rq.rqn);
-
- for (ix = 0; ix < priv->max_nch; ix++)
- mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, priv->drop_rq.rqn);
-
- if (priv->profile->rx_ptp_support)
- mlx5e_rqt_redirect_direct(&res->ptp.rqt, priv->drop_rq.rqn);
-}
-
-int mlx5e_modify_tirs_hash(struct mlx5e_priv *priv)
-{
- struct mlx5e_rss_params_hash *rss_hash = &priv->rx_res->rss_params.hash;
- struct mlx5e_rss_params_traffic_type rss_tt;
- struct mlx5e_rx_res *res = priv->rx_res;
- struct mlx5e_tir_builder *builder;
- enum mlx5e_traffic_types tt;
-
- builder = mlx5e_tir_builder_alloc(true);
- if (!builder)
- return -ENOMEM;
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
- rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
- mlx5e_tir_builder_build_rss(builder, rss_hash, &rss_tt, false);
- mlx5e_tir_modify(&res->rss[tt].indir_tir, builder);
- mlx5e_tir_builder_clear(builder);
- }
-
- /* Verify inner tirs resources allocated */
- if (!res->rss[0].inner_indir_tir.tirn)
- goto out;
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
- rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
- mlx5e_tir_builder_build_rss(builder, rss_hash, &rss_tt, true);
- mlx5e_tir_modify(&res->rss[tt].indir_tir, builder);
- mlx5e_tir_builder_clear(builder);
- }
-
-out:
- mlx5e_tir_builder_free(builder);
- return 0;
-}
-
static int mlx5e_modify_tirs_lro(struct mlx5e_priv *priv)
{
struct mlx5e_rx_res *res = priv->rx_res;
- struct mlx5e_tir_builder *builder;
struct mlx5e_lro_param lro_param;
- enum mlx5e_traffic_types tt;
- int err;
- int ix;
-
- builder = mlx5e_tir_builder_alloc(true);
- if (!builder)
- return -ENOMEM;
lro_param = mlx5e_get_lro_param(&priv->channels.params);
- mlx5e_tir_builder_build_lro(builder, &lro_param);
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
- err = mlx5e_tir_modify(&res->rss[tt].indir_tir, builder);
- if (err)
- goto err_free_builder;
-
- /* Verify inner tirs resources allocated */
- if (!res->rss[0].inner_indir_tir.tirn)
- continue;
-
- err = mlx5e_tir_modify(&res->rss[tt].inner_indir_tir, builder);
- if (err)
- goto err_free_builder;
- }
-
- for (ix = 0; ix < priv->max_nch; ix++) {
- err = mlx5e_tir_modify(&res->channels[ix].direct_tir, builder);
- if (err)
- goto err_free_builder;
- }
-
-err_free_builder:
- mlx5e_tir_builder_free(builder);
- return err;
+ return mlx5e_rx_res_lro_set_param(res, &lro_param);
}
static MLX5E_DEFINE_PREACTIVATE_WRAPPER_CTX(mlx5e_modify_tirs_lro);
@@ -2572,8 +2384,7 @@ int mlx5e_num_channels_changed(struct mlx5e_priv *priv)
/* This function may be called on attach, before priv->rx_res is created. */
if (!netif_is_rxfh_configured(priv->netdev) && priv->rx_res)
- mlx5e_rss_params_indir_init_uniform(&priv->rx_res->rss_params.indir,
- count);
+ mlx5e_rx_res_rss_set_indir_uniform(priv->rx_res, count);
return 0;
}
@@ -2633,18 +2444,14 @@ void mlx5e_activate_priv_channels(struct mlx5e_priv *priv)
mlx5e_wait_channels_min_rx_wqes(&priv->channels);
- if (priv->rx_res) {
- mlx5e_redirect_rqts_to_channels(priv, &priv->channels);
- mlx5e_xsk_redirect_rqts_to_channels(priv, &priv->channels);
- }
+ if (priv->rx_res)
+ mlx5e_rx_res_channels_activate(priv->rx_res, &priv->channels);
}
void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv)
{
- if (priv->rx_res) {
- mlx5e_xsk_redirect_rqts_to_drop(priv, &priv->channels);
- mlx5e_redirect_rqts_to_drop(priv);
- }
+ if (priv->rx_res)
+ mlx5e_rx_res_channels_deactivate(priv->rx_res);
if (mlx5e_is_vport_rep(priv))
mlx5e_remove_sqs_fwd_rules(priv);
@@ -3019,194 +2826,6 @@ static void mlx5e_cleanup_nic_tx(struct mlx5e_priv *priv)
mlx5e_destroy_tises(priv);
}
-int mlx5e_create_indirect_tirs(struct mlx5e_priv *priv, bool inner_ttc)
-{
- struct mlx5e_rss_params_hash *rss_hash = &priv->rx_res->rss_params.hash;
- bool inner_ft_support = priv->channels.params.tunneled_offload_en;
- struct mlx5e_rss_params_traffic_type rss_tt;
- struct mlx5e_rx_res *res = priv->rx_res;
- enum mlx5e_traffic_types tt, max_tt;
- struct mlx5e_tir_builder *builder;
- struct mlx5e_lro_param lro_param;
- u32 indir_rqtn;
- int err = 0;
-
- builder = mlx5e_tir_builder_alloc(false);
- if (!builder)
- return -ENOMEM;
-
- lro_param = mlx5e_get_lro_param(&priv->channels.params);
- indir_rqtn = mlx5e_rqt_get_rqtn(&res->indir_rqt);
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
- mlx5e_tir_builder_build_rqt(builder, priv->mdev->mlx5e_res.hw_objs.td.tdn,
- indir_rqtn, inner_ft_support);
- mlx5e_tir_builder_build_lro(builder, &lro_param);
- rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
- mlx5e_tir_builder_build_rss(builder, rss_hash, &rss_tt, false);
-
- err = mlx5e_tir_init(&res->rss[tt].indir_tir, builder, priv->mdev, true);
- if (err) {
- mlx5_core_warn(priv->mdev, "create indirect tirs failed, %d\n", err);
- goto err_destroy_tirs;
- }
-
- mlx5e_tir_builder_clear(builder);
- }
-
- if (!inner_ttc || !mlx5e_tunnel_inner_ft_supported(priv->mdev))
- goto out;
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
- mlx5e_tir_builder_build_rqt(builder, priv->mdev->mlx5e_res.hw_objs.td.tdn,
- indir_rqtn, inner_ft_support);
- mlx5e_tir_builder_build_lro(builder, &lro_param);
- rss_tt = mlx5e_rx_res_rss_get_current_tt_config(res, tt);
- mlx5e_tir_builder_build_rss(builder, rss_hash, &rss_tt, true);
-
- err = mlx5e_tir_init(&res->rss[tt].inner_indir_tir, builder, priv->mdev, true);
- if (err) {
- mlx5_core_warn(priv->mdev, "create inner indirect tirs failed, %d\n", err);
- goto err_destroy_inner_tirs;
- }
-
- mlx5e_tir_builder_clear(builder);
- }
-
- goto out;
-
-err_destroy_inner_tirs:
- max_tt = tt;
- for (tt = 0; tt < max_tt; tt++)
- mlx5e_tir_destroy(&res->rss[tt].inner_indir_tir);
-
- tt = MLX5E_NUM_INDIR_TIRS;
-err_destroy_tirs:
- max_tt = tt;
- for (tt = 0; tt < max_tt; tt++)
- mlx5e_tir_destroy(&res->rss[tt].indir_tir);
-
-out:
- mlx5e_tir_builder_free(builder);
-
- return err;
-}
-
-static int mlx5e_create_direct_tir(struct mlx5e_priv *priv, struct mlx5e_tir *tir,
- struct mlx5e_tir_builder *builder, struct mlx5e_rqt *rqt)
-{
- bool inner_ft_support = priv->channels.params.tunneled_offload_en;
- struct mlx5e_lro_param lro_param;
- int err = 0;
-
- lro_param = mlx5e_get_lro_param(&priv->channels.params);
-
- mlx5e_tir_builder_build_rqt(builder, priv->mdev->mlx5e_res.hw_objs.td.tdn,
- mlx5e_rqt_get_rqtn(rqt), inner_ft_support);
- mlx5e_tir_builder_build_lro(builder, &lro_param);
- mlx5e_tir_builder_build_direct(builder);
-
- err = mlx5e_tir_init(tir, builder, priv->mdev, true);
- if (unlikely(err))
- mlx5_core_warn(priv->mdev, "create tirs failed, %d\n", err);
-
- mlx5e_tir_builder_clear(builder);
-
- return err;
-}
-
-int mlx5e_create_direct_tirs(struct mlx5e_priv *priv)
-{
- struct mlx5e_rx_res *res = priv->rx_res;
- struct mlx5e_tir_builder *builder;
- int err = 0;
- int ix;
-
- builder = mlx5e_tir_builder_alloc(false);
- if (!builder)
- return -ENOMEM;
-
- for (ix = 0; ix < priv->max_nch; ix++) {
- err = mlx5e_create_direct_tir(priv, &res->channels[ix].direct_tir,
- builder, &res->channels[ix].direct_rqt);
- if (err)
- goto err_destroy_tirs;
- }
-
- goto out;
-
-err_destroy_tirs:
- while (--ix >= 0)
- mlx5e_tir_destroy(&res->channels[ix].direct_tir);
-
-out:
- mlx5e_tir_builder_free(builder);
-
- return err;
-}
-
-static int mlx5e_create_xsk_tirs(struct mlx5e_priv *priv)
-{
- struct mlx5e_rx_res *res = priv->rx_res;
- struct mlx5e_tir_builder *builder;
- int err;
- int ix;
-
- builder = mlx5e_tir_builder_alloc(false);
- if (!builder)
- return -ENOMEM;
-
- for (ix = 0; ix < priv->max_nch; ix++) {
- err = mlx5e_create_direct_tir(priv, &res->channels[ix].xsk_tir,
- builder, &res->channels[ix].xsk_rqt);
- if (err)
- goto err_destroy_tirs;
- }
-
- goto out;
-
-err_destroy_tirs:
- while (--ix >= 0)
- mlx5e_tir_destroy(&res->channels[ix].xsk_tir);
-
-out:
- mlx5e_tir_builder_free(builder);
-
- return err;
-}
-
-void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv)
-{
- struct mlx5e_rx_res *res = priv->rx_res;
- enum mlx5e_traffic_types tt;
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- mlx5e_tir_destroy(&res->rss[tt].indir_tir);
-
- /* Verify inner tirs resources allocated */
- if (!res->rss[0].inner_indir_tir.tirn)
- return;
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- mlx5e_tir_destroy(&res->rss[tt].inner_indir_tir);
-}
-
-void mlx5e_destroy_direct_tirs(struct mlx5e_priv *priv)
-{
- unsigned int ix;
-
- for (ix = 0; ix < priv->max_nch; ix++)
- mlx5e_tir_destroy(&priv->rx_res->channels[ix].direct_tir);
-}
-
-static void mlx5e_destroy_xsk_tirs(struct mlx5e_priv *priv)
-{
- unsigned int ix;
-
- for (ix = 0; ix < priv->max_nch; ix++)
- mlx5e_tir_destroy(&priv->rx_res->channels[ix].xsk_tir);
-}
-
static int mlx5e_modify_channels_scatter_fcs(struct mlx5e_channels *chs, bool enable)
{
int err = 0;
@@ -4471,20 +4090,6 @@ static u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeo
return MLX5_CAP_ETH(mdev, lro_timer_supported_periods[i]);
}
-void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params,
- u16 num_channels)
-{
- enum mlx5e_traffic_types tt;
-
- rss_params->hash.hfunc = ETH_RSS_HASH_TOP;
- netdev_rss_key_fill(rss_params->hash.toeplitz_hash_key,
- sizeof(rss_params->hash.toeplitz_hash_key));
- mlx5e_rss_params_indir_init_uniform(&rss_params->indir, num_channels);
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- rss_params->rx_hash_fields[tt] =
- mlx5e_rss_get_default_tt_config(tt).rx_hash_fields;
-}
-
void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, u16 mtu)
{
struct mlx5e_params *params = &priv->channels.params;
@@ -4809,15 +4414,14 @@ static void mlx5e_nic_cleanup(struct mlx5e_priv *priv)
static int mlx5e_init_nic_rx(struct mlx5e_priv *priv)
{
struct mlx5_core_dev *mdev = priv->mdev;
- struct mlx5e_tir_builder *tir_builder;
+ enum mlx5e_rx_res_features features;
+ struct mlx5e_lro_param lro_param;
int err;
- priv->rx_res = kvzalloc(sizeof(*priv->rx_res), GFP_KERNEL);
+ priv->rx_res = mlx5e_rx_res_alloc();
if (!priv->rx_res)
return -ENOMEM;
- mlx5e_build_rss_params(&priv->rx_res->rss_params, priv->channels.params.num_channels);
-
mlx5e_create_q_counters(priv);
err = mlx5e_open_drop_rq(priv, &priv->drop_rq);
@@ -4826,50 +4430,20 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv)
goto err_destroy_q_counters;
}
- err = mlx5e_create_indirect_rqt(priv);
+ features = MLX5E_RX_RES_FEATURE_XSK | MLX5E_RX_RES_FEATURE_PTP;
+ if (priv->channels.params.tunneled_offload_en)
+ features |= MLX5E_RX_RES_FEATURE_INNER_FT;
+ lro_param = mlx5e_get_lro_param(&priv->channels.params);
+ err = mlx5e_rx_res_init(priv->rx_res, priv->mdev, features,
+ priv->max_nch, priv->drop_rq.rqn, &lro_param,
+ priv->channels.params.num_channels);
if (err)
goto err_close_drop_rq;
- err = mlx5e_create_direct_rqts(priv);
- if (err)
- goto err_destroy_indirect_rqts;
-
- err = mlx5e_create_indirect_tirs(priv, true);
- if (err)
- goto err_destroy_direct_rqts;
-
- err = mlx5e_create_direct_tirs(priv);
- if (err)
- goto err_destroy_indirect_tirs;
-
- err = mlx5e_create_xsk_rqts(priv);
- if (unlikely(err))
- goto err_destroy_direct_tirs;
-
- err = mlx5e_create_xsk_tirs(priv);
- if (unlikely(err))
- goto err_destroy_xsk_rqts;
-
- err = mlx5e_rqt_init_direct(&priv->rx_res->ptp.rqt, priv->mdev, false,
- priv->drop_rq.rqn);
- if (err)
- goto err_destroy_xsk_tirs;
-
- tir_builder = mlx5e_tir_builder_alloc(false);
- if (!tir_builder) {
- err = -ENOMEM;
- goto err_destroy_ptp_rqt;
- }
- err = mlx5e_create_direct_tir(priv, &priv->rx_res->ptp.tir, tir_builder,
- &priv->rx_res->ptp.rqt);
- mlx5e_tir_builder_free(tir_builder);
- if (err)
- goto err_destroy_ptp_rqt;
-
err = mlx5e_create_flow_steering(priv);
if (err) {
mlx5_core_warn(mdev, "create flow steering failed, %d\n", err);
- goto err_destroy_ptp_direct_tir;
+ goto err_destroy_rx_res;
}
err = mlx5e_tc_nic_init(priv);
@@ -4890,27 +4464,13 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv)
mlx5e_tc_nic_cleanup(priv);
err_destroy_flow_steering:
mlx5e_destroy_flow_steering(priv);
-err_destroy_ptp_direct_tir:
- mlx5e_tir_destroy(&priv->rx_res->ptp.tir);
-err_destroy_ptp_rqt:
- mlx5e_rqt_destroy(&priv->rx_res->ptp.rqt);
-err_destroy_xsk_tirs:
- mlx5e_destroy_xsk_tirs(priv);
-err_destroy_xsk_rqts:
- mlx5e_destroy_xsk_rqts(priv);
-err_destroy_direct_tirs:
- mlx5e_destroy_direct_tirs(priv);
-err_destroy_indirect_tirs:
- mlx5e_destroy_indirect_tirs(priv);
-err_destroy_direct_rqts:
- mlx5e_destroy_direct_rqts(priv);
-err_destroy_indirect_rqts:
- mlx5e_rqt_destroy(&priv->rx_res->indir_rqt);
+err_destroy_rx_res:
+ mlx5e_rx_res_destroy(priv->rx_res);
err_close_drop_rq:
mlx5e_close_drop_rq(&priv->drop_rq);
err_destroy_q_counters:
mlx5e_destroy_q_counters(priv);
- kvfree(priv->rx_res);
+ mlx5e_rx_res_free(priv->rx_res);
priv->rx_res = NULL;
return err;
}
@@ -4920,17 +4480,10 @@ static void mlx5e_cleanup_nic_rx(struct mlx5e_priv *priv)
mlx5e_accel_cleanup_rx(priv);
mlx5e_tc_nic_cleanup(priv);
mlx5e_destroy_flow_steering(priv);
- mlx5e_tir_destroy(&priv->rx_res->ptp.tir);
- mlx5e_rqt_destroy(&priv->rx_res->ptp.rqt);
- mlx5e_destroy_xsk_tirs(priv);
- mlx5e_destroy_xsk_rqts(priv);
- mlx5e_destroy_direct_tirs(priv);
- mlx5e_destroy_indirect_tirs(priv);
- mlx5e_destroy_direct_rqts(priv);
- mlx5e_rqt_destroy(&priv->rx_res->indir_rqt);
+ mlx5e_rx_res_destroy(priv->rx_res);
mlx5e_close_drop_rq(&priv->drop_rq);
mlx5e_destroy_q_counters(priv);
- kvfree(priv->rx_res);
+ mlx5e_rx_res_free(priv->rx_res);
priv->rx_res = NULL;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 2c54951c240d..f6e96b7d4698 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -655,7 +655,7 @@ static int mlx5e_create_rep_ttc_table(struct mlx5e_priv *priv)
MLX5_FLOW_NAMESPACE_KERNEL);
/* The inner_ttc in the ttc params is intentionally not set */
- ttc_params.any_tt_tirn = res->channels[0].direct_tir.tirn;
+ ttc_params.any_tt_tirn = mlx5e_rx_res_get_tirn_direct(res, 0);
mlx5e_set_ttc_ft_params(&ttc_params);
if (rep->vport != MLX5_VPORT_UPLINK)
@@ -663,7 +663,7 @@ static int mlx5e_create_rep_ttc_table(struct mlx5e_priv *priv)
ttc_params.ft_attr.level = MLX5E_TTC_FT_LEVEL + 1;
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = res->rss[tt].indir_tir.tirn;
+ ttc_params.indir_tirn[tt] = mlx5e_rx_res_get_tirn_rss(res, tt);
err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
if (err) {
@@ -758,14 +758,13 @@ int mlx5e_rep_bond_update(struct mlx5e_priv *priv, bool cleanup)
static int mlx5e_init_rep_rx(struct mlx5e_priv *priv)
{
struct mlx5_core_dev *mdev = priv->mdev;
+ struct mlx5e_lro_param lro_param;
int err;
- priv->rx_res = kvzalloc(sizeof(*priv->rx_res), GFP_KERNEL);
+ priv->rx_res = mlx5e_rx_res_alloc();
if (!priv->rx_res)
return -ENOMEM;
- mlx5e_build_rss_params(&priv->rx_res->rss_params, priv->channels.params.num_channels);
-
mlx5e_init_l2_addr(priv);
err = mlx5e_open_drop_rq(priv, &priv->drop_rq);
@@ -774,25 +773,16 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv)
return err;
}
- err = mlx5e_create_indirect_rqt(priv);
+ lro_param = mlx5e_get_lro_param(&priv->channels.params);
+ err = mlx5e_rx_res_init(priv->rx_res, priv->mdev, 0,
+ priv->max_nch, priv->drop_rq.rqn, &lro_param,
+ priv->channels.params.num_channels);
if (err)
goto err_close_drop_rq;
- err = mlx5e_create_direct_rqts(priv);
- if (err)
- goto err_destroy_indirect_rqts;
-
- err = mlx5e_create_indirect_tirs(priv, false);
- if (err)
- goto err_destroy_direct_rqts;
-
- err = mlx5e_create_direct_tirs(priv);
- if (err)
- goto err_destroy_indirect_tirs;
-
err = mlx5e_create_rep_ttc_table(priv);
if (err)
- goto err_destroy_direct_tirs;
+ goto err_destroy_rx_res;
err = mlx5e_create_rep_root_ft(priv);
if (err)
@@ -810,17 +800,11 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv)
mlx5e_destroy_rep_root_ft(priv);
err_destroy_ttc_table:
mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
-err_destroy_direct_tirs:
- mlx5e_destroy_direct_tirs(priv);
-err_destroy_indirect_tirs:
- mlx5e_destroy_indirect_tirs(priv);
-err_destroy_direct_rqts:
- mlx5e_destroy_direct_rqts(priv);
-err_destroy_indirect_rqts:
- mlx5e_rqt_destroy(&priv->rx_res->indir_rqt);
+err_destroy_rx_res:
+ mlx5e_rx_res_destroy(priv->rx_res);
err_close_drop_rq:
mlx5e_close_drop_rq(&priv->drop_rq);
- kvfree(priv->rx_res);
+ mlx5e_rx_res_free(priv->rx_res);
priv->rx_res = NULL;
return err;
}
@@ -831,12 +815,9 @@ static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv)
rep_vport_rx_rule_destroy(priv);
mlx5e_destroy_rep_root_ft(priv);
mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
- mlx5e_destroy_direct_tirs(priv);
- mlx5e_destroy_indirect_tirs(priv);
- mlx5e_destroy_direct_rqts(priv);
- mlx5e_rqt_destroy(&priv->rx_res->indir_rqt);
+ mlx5e_rx_res_destroy(priv->rx_res);
mlx5e_close_drop_rq(&priv->drop_rq);
- kvfree(priv->rx_res);
+ mlx5e_rx_res_free(priv->rx_res);
priv->rx_res = NULL;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 4d7ed24ae13c..04687ffaeffa 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -527,7 +527,8 @@ static int mlx5e_hairpin_create_indirect_rqt(struct mlx5e_hairpin *hp)
mlx5e_rss_params_indir_init_uniform(indir, hp->num_channels);
err = mlx5e_rqt_init_indir(&hp->indir_rqt, mdev, hp->pair->rqn, hp->num_channels,
- priv->rx_res->rss_params.hash.hfunc, indir);
+ mlx5e_rx_res_get_current_hash(priv->rx_res).hfunc,
+ indir);
kvfree(indir);
return err;
@@ -536,7 +537,7 @@ static int mlx5e_hairpin_create_indirect_rqt(struct mlx5e_hairpin *hp)
static int mlx5e_hairpin_create_indirect_tirs(struct mlx5e_hairpin *hp)
{
struct mlx5e_priv *priv = hp->func_priv;
- struct mlx5e_rss_params_hash *rss_hash;
+ struct mlx5e_rss_params_hash rss_hash;
enum mlx5e_traffic_types tt, max_tt;
struct mlx5e_tir_builder *builder;
int err = 0;
@@ -545,7 +546,7 @@ static int mlx5e_hairpin_create_indirect_tirs(struct mlx5e_hairpin *hp)
if (!builder)
return -ENOMEM;
- rss_hash = &priv->rx_res->rss_params.hash;
+ rss_hash = mlx5e_rx_res_get_current_hash(priv->rx_res);
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) {
struct mlx5e_rss_params_traffic_type rss_tt;
@@ -555,7 +556,7 @@ static int mlx5e_hairpin_create_indirect_tirs(struct mlx5e_hairpin *hp)
mlx5e_tir_builder_build_rqt(builder, hp->tdn,
mlx5e_rqt_get_rqtn(&hp->indir_rqt),
false);
- mlx5e_tir_builder_build_rss(builder, rss_hash, &rss_tt, false);
+ mlx5e_tir_builder_build_rss(builder, &rss_hash, &rss_tt, false);
err = mlx5e_tir_init(&hp->indir_tir[tt], builder, hp->func_mdev, false);
if (err) {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
index a126cbc6f0d6..1f118678ea9d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
@@ -333,7 +333,7 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
mlx5e_set_ttc_basic_params(priv, &ttc_params);
mlx5e_set_ttc_ft_params(&ttc_params);
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = priv->rx_res->rss[tt].indir_tir.tirn;
+ ttc_params.indir_tirn[tt] = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
if (err) {
@@ -359,14 +359,13 @@ static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv)
static int mlx5i_init_rx(struct mlx5e_priv *priv)
{
struct mlx5_core_dev *mdev = priv->mdev;
+ struct mlx5e_lro_param lro_param;
int err;
- priv->rx_res = kvzalloc(sizeof(*priv->rx_res), GFP_KERNEL);
+ priv->rx_res = mlx5e_rx_res_alloc();
if (!priv->rx_res)
return -ENOMEM;
- mlx5e_build_rss_params(&priv->rx_res->rss_params, priv->channels.params.num_channels);
-
mlx5e_create_q_counters(priv);
err = mlx5e_open_drop_rq(priv, &priv->drop_rq);
@@ -375,41 +374,26 @@ static int mlx5i_init_rx(struct mlx5e_priv *priv)
goto err_destroy_q_counters;
}
- err = mlx5e_create_indirect_rqt(priv);
+ lro_param = mlx5e_get_lro_param(&priv->channels.params);
+ err = mlx5e_rx_res_init(priv->rx_res, priv->mdev, 0,
+ priv->max_nch, priv->drop_rq.rqn, &lro_param,
+ priv->channels.params.num_channels);
if (err)
goto err_close_drop_rq;
- err = mlx5e_create_direct_rqts(priv);
- if (err)
- goto err_destroy_indirect_rqts;
-
- err = mlx5e_create_indirect_tirs(priv, false);
- if (err)
- goto err_destroy_direct_rqts;
-
- err = mlx5e_create_direct_tirs(priv);
- if (err)
- goto err_destroy_indirect_tirs;
-
err = mlx5i_create_flow_steering(priv);
if (err)
- goto err_destroy_direct_tirs;
+ goto err_destroy_rx_res;
return 0;
-err_destroy_direct_tirs:
- mlx5e_destroy_direct_tirs(priv);
-err_destroy_indirect_tirs:
- mlx5e_destroy_indirect_tirs(priv);
-err_destroy_direct_rqts:
- mlx5e_destroy_direct_rqts(priv);
-err_destroy_indirect_rqts:
- mlx5e_rqt_destroy(&priv->rx_res->indir_rqt);
+err_destroy_rx_res:
+ mlx5e_rx_res_destroy(priv->rx_res);
err_close_drop_rq:
mlx5e_close_drop_rq(&priv->drop_rq);
err_destroy_q_counters:
mlx5e_destroy_q_counters(priv);
- kvfree(priv->rx_res);
+ mlx5e_rx_res_free(priv->rx_res);
priv->rx_res = NULL;
return err;
}
@@ -417,13 +401,10 @@ static int mlx5i_init_rx(struct mlx5e_priv *priv)
static void mlx5i_cleanup_rx(struct mlx5e_priv *priv)
{
mlx5i_destroy_flow_steering(priv);
- mlx5e_destroy_direct_tirs(priv);
- mlx5e_destroy_indirect_tirs(priv);
- mlx5e_destroy_direct_rqts(priv);
- mlx5e_rqt_destroy(&priv->rx_res->indir_rqt);
+ mlx5e_rx_res_destroy(priv->rx_res);
mlx5e_close_drop_rq(&priv->drop_rq);
mlx5e_destroy_q_counters(priv);
- kvfree(priv->rx_res);
+ mlx5e_rx_res_free(priv->rx_res);
priv->rx_res = NULL;
}
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 04/16] net/mlx5e: Allocate the array of channels according to the real max_nch
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (2 preceding siblings ...)
2021-08-03 2:28 ` [net-next 03/16] net/mlx5e: Hide all implementation details of mlx5e_rx_res Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 05/16] net/mlx5e: Rename traffic type enums Saeed Mahameed
` (11 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maxim Mikityanskiy, Tariq Toukan, Saeed Mahameed
From: Maxim Mikityanskiy <maximmi@nvidia.com>
The channels array in struct mlx5e_rx_res is converted to a dynamic one,
which will use the dynamic value of max_nch instead of
implementation-defined maximum of MLX5E_MAX_NUM_CHANNELS.
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c | 12 +++++++++++-
drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h | 2 --
3 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 968e6a473cec..594b7971caf9 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -140,6 +140,7 @@ struct page_pool;
#define MLX5E_PARAMS_DEFAULT_MIN_RX_WQES_MPW 0x2
#define MLX5E_MIN_NUM_CHANNELS 0x1
+#define MLX5E_MAX_NUM_CHANNELS (MLX5E_INDIR_RQT_SIZE / 2)
#define MLX5E_MAX_NUM_SQS (MLX5E_MAX_NUM_CHANNELS * MLX5E_MAX_NUM_TC)
#define MLX5E_TX_CQ_POLL_BUDGET 128
#define MLX5E_TX_XSK_POLL_BUDGET 64
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
index a6b3a9473405..751b2cdc3ec1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
@@ -91,7 +91,7 @@ struct mlx5e_rx_res {
struct mlx5e_tir direct_tir;
struct mlx5e_rqt xsk_rqt;
struct mlx5e_tir xsk_tir;
- } channels[MLX5E_MAX_NUM_CHANNELS];
+ } *channels;
struct {
struct mlx5e_rqt rqt;
@@ -210,6 +210,12 @@ static int mlx5e_rx_res_channels_init(struct mlx5e_rx_res *res,
if (!builder)
return -ENOMEM;
+ res->channels = kvcalloc(res->max_nch, sizeof(*res->channels), GFP_KERNEL);
+ if (!res->channels) {
+ err = -ENOMEM;
+ goto out;
+ }
+
for (ix = 0; ix < res->max_nch; ix++) {
err = mlx5e_rqt_init_direct(&res->channels[ix].direct_rqt,
res->mdev, false, res->drop_rqn);
@@ -288,6 +294,8 @@ static int mlx5e_rx_res_channels_init(struct mlx5e_rx_res *res,
while (--ix >= 0)
mlx5e_rqt_destroy(&res->channels[ix].direct_rqt);
+ kvfree(res->channels);
+
out:
mlx5e_tir_builder_free(builder);
@@ -355,6 +363,8 @@ static void mlx5e_rx_res_channels_destroy(struct mlx5e_rx_res *res)
mlx5e_tir_destroy(&res->channels[ix].xsk_tir);
mlx5e_rqt_destroy(&res->channels[ix].xsk_rqt);
}
+
+ kvfree(res->channels);
}
static void mlx5e_rx_res_ptp_destroy(struct mlx5e_rx_res *res)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
index 0092ee80a2cf..934e41a0761f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
@@ -9,8 +9,6 @@
#include "tir.h"
#include "fs.h"
-#define MLX5E_MAX_NUM_CHANNELS (MLX5E_INDIR_RQT_SIZE / 2)
-
struct mlx5e_rx_res;
struct mlx5e_channels;
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 05/16] net/mlx5e: Rename traffic type enums
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (3 preceding siblings ...)
2021-08-03 2:28 ` [net-next 04/16] net/mlx5e: Allocate the array of channels according to the real max_nch Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 06/16] net/mlx5e: Rename some related TTC args and functions Saeed Mahameed
` (10 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maor Gottlieb, Tariq Toukan, Mark Bloch, Saeed Mahameed
From: Maor Gottlieb <maorg@nvidia.com>
Rename traffic type enums as part of the preparation for moving
the traffic type logic to a separate file.
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/en/fs.h | 89 ++++++++++---------
.../mellanox/mlx5/core/en/fs_tt_redirect.c | 24 ++---
.../mellanox/mlx5/core/en/fs_tt_redirect.h | 2 +-
.../net/ethernet/mellanox/mlx5/core/en/ptp.c | 4 +-
.../ethernet/mellanox/mlx5/core/en/rx_res.c | 44 ++++-----
.../ethernet/mellanox/mlx5/core/en/rx_res.h | 12 +--
.../mellanox/mlx5/core/en_accel/fs_tcp.c | 6 +-
.../mellanox/mlx5/core/en_accel/ipsec_fs.c | 6 +-
.../net/ethernet/mellanox/mlx5/core/en_arfs.c | 12 +--
.../net/ethernet/mellanox/mlx5/core/en_fs.c | 80 ++++++++---------
.../mellanox/mlx5/core/en_fs_ethtool.c | 36 ++++----
.../net/ethernet/mellanox/mlx5/core/en_main.c | 2 +-
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 4 +-
13 files changed, 162 insertions(+), 159 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
index 0e053aab12b5..77fe98c42ec4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
@@ -67,22 +67,23 @@ struct mlx5e_l2_table {
bool promisc_enabled;
};
-enum mlx5e_traffic_types {
- MLX5E_TT_IPV4_TCP,
- MLX5E_TT_IPV6_TCP,
- MLX5E_TT_IPV4_UDP,
- MLX5E_TT_IPV6_UDP,
- MLX5E_TT_IPV4_IPSEC_AH,
- MLX5E_TT_IPV6_IPSEC_AH,
- MLX5E_TT_IPV4_IPSEC_ESP,
- MLX5E_TT_IPV6_IPSEC_ESP,
- MLX5E_TT_IPV4,
- MLX5E_TT_IPV6,
- MLX5E_TT_ANY,
- MLX5E_NUM_TT,
- MLX5E_NUM_INDIR_TIRS = MLX5E_TT_ANY,
+enum mlx5_traffic_types {
+ MLX5_TT_IPV4_TCP,
+ MLX5_TT_IPV6_TCP,
+ MLX5_TT_IPV4_UDP,
+ MLX5_TT_IPV6_UDP,
+ MLX5_TT_IPV4_IPSEC_AH,
+ MLX5_TT_IPV6_IPSEC_AH,
+ MLX5_TT_IPV4_IPSEC_ESP,
+ MLX5_TT_IPV6_IPSEC_ESP,
+ MLX5_TT_IPV4,
+ MLX5_TT_IPV6,
+ MLX5_TT_ANY,
+ MLX5_NUM_TT,
};
+#define MLX5E_NUM_INDIR_TIRS (MLX5_NUM_TT - 1)
+
#define MLX5_HASH_IP (MLX5_HASH_FIELD_SEL_SRC_IP |\
MLX5_HASH_FIELD_SEL_DST_IP)
#define MLX5_HASH_IP_L4PORTS (MLX5_HASH_FIELD_SEL_SRC_IP |\
@@ -93,14 +94,14 @@ enum mlx5e_traffic_types {
MLX5_HASH_FIELD_SEL_DST_IP |\
MLX5_HASH_FIELD_SEL_IPSEC_SPI)
-enum mlx5e_tunnel_types {
- MLX5E_TT_IPV4_GRE,
- MLX5E_TT_IPV6_GRE,
- MLX5E_TT_IPV4_IPIP,
- MLX5E_TT_IPV6_IPIP,
- MLX5E_TT_IPV4_IPV6,
- MLX5E_TT_IPV6_IPV6,
- MLX5E_NUM_TUNNEL_TT,
+enum mlx5_tunnel_types {
+ MLX5_TT_IPV4_GRE,
+ MLX5_TT_IPV6_GRE,
+ MLX5_TT_IPV4_IPIP,
+ MLX5_TT_IPV6_IPIP,
+ MLX5_TT_IPV4_IPV6,
+ MLX5_TT_IPV6_IPV6,
+ MLX5_NUM_TUNNEL_TT,
};
bool mlx5e_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev);
@@ -113,8 +114,8 @@ struct mlx5e_ttc_rule {
/* L3/L4 traffic type classifier */
struct mlx5e_ttc_table {
struct mlx5e_flow_table ft;
- struct mlx5e_ttc_rule rules[MLX5E_NUM_TT];
- struct mlx5_flow_handle *tunnel_rules[MLX5E_NUM_TUNNEL_TT];
+ struct mlx5e_ttc_rule rules[MLX5_NUM_TT];
+ struct mlx5_flow_handle *tunnel_rules[MLX5_NUM_TUNNEL_TT];
};
/* NIC prio FTS */
@@ -138,21 +139,21 @@ enum {
#endif
};
-#define MLX5E_TTC_NUM_GROUPS 3
-#define MLX5E_TTC_GROUP1_SIZE (BIT(3) + MLX5E_NUM_TUNNEL_TT)
-#define MLX5E_TTC_GROUP2_SIZE BIT(1)
-#define MLX5E_TTC_GROUP3_SIZE BIT(0)
-#define MLX5E_TTC_TABLE_SIZE (MLX5E_TTC_GROUP1_SIZE +\
- MLX5E_TTC_GROUP2_SIZE +\
- MLX5E_TTC_GROUP3_SIZE)
-
-#define MLX5E_INNER_TTC_NUM_GROUPS 3
-#define MLX5E_INNER_TTC_GROUP1_SIZE BIT(3)
-#define MLX5E_INNER_TTC_GROUP2_SIZE BIT(1)
-#define MLX5E_INNER_TTC_GROUP3_SIZE BIT(0)
-#define MLX5E_INNER_TTC_TABLE_SIZE (MLX5E_INNER_TTC_GROUP1_SIZE +\
- MLX5E_INNER_TTC_GROUP2_SIZE +\
- MLX5E_INNER_TTC_GROUP3_SIZE)
+#define MLX5_TTC_NUM_GROUPS 3
+#define MLX5_TTC_GROUP1_SIZE (BIT(3) + MLX5_NUM_TUNNEL_TT)
+#define MLX5_TTC_GROUP2_SIZE BIT(1)
+#define MLX5_TTC_GROUP3_SIZE BIT(0)
+#define MLX5_TTC_TABLE_SIZE (MLX5_TTC_GROUP1_SIZE +\
+ MLX5_TTC_GROUP2_SIZE +\
+ MLX5_TTC_GROUP3_SIZE)
+
+#define MLX5_INNER_TTC_NUM_GROUPS 3
+#define MLX5_INNER_TTC_GROUP1_SIZE BIT(3)
+#define MLX5_INNER_TTC_GROUP2_SIZE BIT(1)
+#define MLX5_INNER_TTC_GROUP3_SIZE BIT(0)
+#define MLX5_INNER_TTC_TABLE_SIZE (MLX5_INNER_TTC_GROUP1_SIZE +\
+ MLX5_INNER_TTC_GROUP2_SIZE +\
+ MLX5_INNER_TTC_GROUP3_SIZE)
struct mlx5e_priv;
@@ -251,11 +252,13 @@ void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv,
struct mlx5e_ttc_table *ttc);
void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft);
-int mlx5e_ttc_fwd_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types type,
+int mlx5e_ttc_fwd_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type,
struct mlx5_flow_destination *new_dest);
struct mlx5_flow_destination
-mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types type);
-int mlx5e_ttc_fwd_default_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types type);
+mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv,
+ enum mlx5_traffic_types type);
+int mlx5e_ttc_fwd_default_dest(struct mlx5e_priv *priv,
+ enum mlx5_traffic_types type);
void mlx5e_enable_cvlan_filter(struct mlx5e_priv *priv);
void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv);
@@ -263,7 +266,7 @@ void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv);
int mlx5e_create_flow_steering(struct mlx5e_priv *priv);
void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv);
-u8 mlx5e_get_proto_by_tunnel_type(enum mlx5e_tunnel_types tt);
+u8 mlx5e_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt);
int mlx5e_add_vlan_trap(struct mlx5e_priv *priv, int trap_id, int tir_num);
void mlx5e_remove_vlan_trap(struct mlx5e_priv *priv);
int mlx5e_add_mac_trap(struct mlx5e_priv *priv, int trap_id, int tir_num);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
index 909faa6c89d7..5645e8032218 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
@@ -33,22 +33,22 @@ static char *fs_udp_type2str(enum fs_udp_type i)
}
}
-static enum mlx5e_traffic_types fs_udp2tt(enum fs_udp_type i)
+static enum mlx5_traffic_types fs_udp2tt(enum fs_udp_type i)
{
switch (i) {
case FS_IPV4_UDP:
- return MLX5E_TT_IPV4_UDP;
+ return MLX5_TT_IPV4_UDP;
default: /* FS_IPV6_UDP */
- return MLX5E_TT_IPV6_UDP;
+ return MLX5_TT_IPV6_UDP;
}
}
-static enum fs_udp_type tt2fs_udp(enum mlx5e_traffic_types i)
+static enum fs_udp_type tt2fs_udp(enum mlx5_traffic_types i)
{
switch (i) {
- case MLX5E_TT_IPV4_UDP:
+ case MLX5_TT_IPV4_UDP:
return FS_IPV4_UDP;
- case MLX5E_TT_IPV6_UDP:
+ case MLX5_TT_IPV6_UDP:
return FS_IPV6_UDP;
default:
return FS_UDP_NUM_TYPES;
@@ -75,7 +75,7 @@ static void fs_udp_set_dport_flow(struct mlx5_flow_spec *spec, enum fs_udp_type
struct mlx5_flow_handle *
mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_priv *priv,
- enum mlx5e_traffic_types ttc_type,
+ enum mlx5_traffic_types ttc_type,
u32 tir_num, u16 d_port)
{
enum fs_udp_type type = tt2fs_udp(ttc_type);
@@ -401,7 +401,7 @@ static int fs_any_add_default_rule(struct mlx5e_priv *priv)
fs_any = priv->fs.any;
fs_any_t = &fs_any->table;
- dest = mlx5e_ttc_get_default_dest(priv, MLX5E_TT_ANY);
+ dest = mlx5e_ttc_get_default_dest(priv, MLX5_TT_ANY);
rule = mlx5_add_flow_rules(fs_any_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -514,11 +514,11 @@ static int fs_any_disable(struct mlx5e_priv *priv)
int err;
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5e_ttc_fwd_default_dest(priv, MLX5E_TT_ANY);
+ err = mlx5e_ttc_fwd_default_dest(priv, MLX5_TT_ANY);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
- __func__, MLX5E_TT_ANY, err);
+ __func__, MLX5_TT_ANY, err);
return err;
}
return 0;
@@ -533,11 +533,11 @@ static int fs_any_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.any->table.t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5e_ttc_fwd_dest(priv, MLX5E_TT_ANY, &dest);
+ err = mlx5e_ttc_fwd_dest(priv, MLX5_TT_ANY, &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
- __func__, MLX5E_TT_ANY, err);
+ __func__, MLX5_TT_ANY, err);
return err;
}
return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h
index 8385df24eb99..7a70c4f38fda 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h
@@ -12,7 +12,7 @@ void mlx5e_fs_tt_redirect_del_rule(struct mlx5_flow_handle *rule);
/* UDP traffic type redirect */
struct mlx5_flow_handle *
mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_priv *priv,
- enum mlx5e_traffic_types ttc_type,
+ enum mlx5_traffic_types ttc_type,
u32 tir_num, u16 d_port);
void mlx5e_fs_tt_redirect_udp_destroy(struct mlx5e_priv *priv);
int mlx5e_fs_tt_redirect_udp_create(struct mlx5e_priv *priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
index f9c96e5a7f54..f479ef31ca40 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
@@ -617,7 +617,7 @@ static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv)
if (err)
goto out_free;
- rule = mlx5e_fs_tt_redirect_udp_add_rule(priv, MLX5E_TT_IPV4_UDP,
+ rule = mlx5e_fs_tt_redirect_udp_add_rule(priv, MLX5_TT_IPV4_UDP,
tirn, PTP_EV_PORT);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -625,7 +625,7 @@ static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv)
}
ptp_fs->udp_v4_rule = rule;
- rule = mlx5e_fs_tt_redirect_udp_add_rule(priv, MLX5E_TT_IPV6_UDP,
+ rule = mlx5e_fs_tt_redirect_udp_add_rule(priv, MLX5_TT_IPV6_UDP,
tirn, PTP_EV_PORT);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
index 751b2cdc3ec1..e2a8fe13f29d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c
@@ -6,52 +6,52 @@
#include "params.h"
static const struct mlx5e_rss_params_traffic_type rss_default_config[MLX5E_NUM_INDIR_TIRS] = {
- [MLX5E_TT_IPV4_TCP] = {
+ [MLX5_TT_IPV4_TCP] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV4,
.l4_prot_type = MLX5_L4_PROT_TYPE_TCP,
.rx_hash_fields = MLX5_HASH_IP_L4PORTS,
},
- [MLX5E_TT_IPV6_TCP] = {
+ [MLX5_TT_IPV6_TCP] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV6,
.l4_prot_type = MLX5_L4_PROT_TYPE_TCP,
.rx_hash_fields = MLX5_HASH_IP_L4PORTS,
},
- [MLX5E_TT_IPV4_UDP] = {
+ [MLX5_TT_IPV4_UDP] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV4,
.l4_prot_type = MLX5_L4_PROT_TYPE_UDP,
.rx_hash_fields = MLX5_HASH_IP_L4PORTS,
},
- [MLX5E_TT_IPV6_UDP] = {
+ [MLX5_TT_IPV6_UDP] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV6,
.l4_prot_type = MLX5_L4_PROT_TYPE_UDP,
.rx_hash_fields = MLX5_HASH_IP_L4PORTS,
},
- [MLX5E_TT_IPV4_IPSEC_AH] = {
+ [MLX5_TT_IPV4_IPSEC_AH] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV4,
.l4_prot_type = 0,
.rx_hash_fields = MLX5_HASH_IP_IPSEC_SPI,
},
- [MLX5E_TT_IPV6_IPSEC_AH] = {
+ [MLX5_TT_IPV6_IPSEC_AH] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV6,
.l4_prot_type = 0,
.rx_hash_fields = MLX5_HASH_IP_IPSEC_SPI,
},
- [MLX5E_TT_IPV4_IPSEC_ESP] = {
+ [MLX5_TT_IPV4_IPSEC_ESP] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV4,
.l4_prot_type = 0,
.rx_hash_fields = MLX5_HASH_IP_IPSEC_SPI,
},
- [MLX5E_TT_IPV6_IPSEC_ESP] = {
+ [MLX5_TT_IPV6_IPSEC_ESP] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV6,
.l4_prot_type = 0,
.rx_hash_fields = MLX5_HASH_IP_IPSEC_SPI,
},
- [MLX5E_TT_IPV4] = {
+ [MLX5_TT_IPV4] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV4,
.l4_prot_type = 0,
.rx_hash_fields = MLX5_HASH_IP,
},
- [MLX5E_TT_IPV6] = {
+ [MLX5_TT_IPV6] = {
.l3_prot_type = MLX5_L3_PROT_TYPE_IPV6,
.l4_prot_type = 0,
.rx_hash_fields = MLX5_HASH_IP,
@@ -59,7 +59,7 @@ static const struct mlx5e_rss_params_traffic_type rss_default_config[MLX5E_NUM_I
};
struct mlx5e_rss_params_traffic_type
-mlx5e_rss_get_default_tt_config(enum mlx5e_traffic_types tt)
+mlx5e_rss_get_default_tt_config(enum mlx5_traffic_types tt)
{
return rss_default_config[tt];
}
@@ -106,7 +106,7 @@ struct mlx5e_rx_res *mlx5e_rx_res_alloc(void)
static void mlx5e_rx_res_rss_params_init(struct mlx5e_rx_res *res, unsigned int init_nch)
{
- enum mlx5e_traffic_types tt;
+ enum mlx5_traffic_types tt;
res->rss_params.hash.hfunc = ETH_RSS_HASH_TOP;
netdev_rss_key_fill(res->rss_params.hash.toeplitz_hash_key,
@@ -121,7 +121,7 @@ static int mlx5e_rx_res_rss_init(struct mlx5e_rx_res *res,
const struct mlx5e_lro_param *init_lro_param)
{
bool inner_ft_support = res->features & MLX5E_RX_RES_FEATURE_INNER_FT;
- enum mlx5e_traffic_types tt, max_tt;
+ enum mlx5_traffic_types tt, max_tt;
struct mlx5e_tir_builder *builder;
u32 indir_rqtn;
int err;
@@ -337,7 +337,7 @@ static int mlx5e_rx_res_ptp_init(struct mlx5e_rx_res *res)
static void mlx5e_rx_res_rss_destroy(struct mlx5e_rx_res *res)
{
- enum mlx5e_traffic_types tt;
+ enum mlx5_traffic_types tt;
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
mlx5e_tir_destroy(&res->rss[tt].indir_tir);
@@ -432,12 +432,12 @@ u32 mlx5e_rx_res_get_tirn_xsk(struct mlx5e_rx_res *res, unsigned int ix)
return mlx5e_tir_get_tirn(&res->channels[ix].xsk_tir);
}
-u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt)
{
return mlx5e_tir_get_tirn(&res->rss[tt].indir_tir);
}
-u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt)
{
WARN_ON(!(res->features & MLX5E_RX_RES_FEATURE_INNER_FT));
return mlx5e_tir_get_tirn(&res->rss[tt].inner_indir_tir);
@@ -608,7 +608,7 @@ int mlx5e_rx_res_xsk_deactivate(struct mlx5e_rx_res *res, unsigned int ix)
}
struct mlx5e_rss_params_traffic_type
-mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt)
{
struct mlx5e_rss_params_traffic_type rss_tt;
@@ -643,7 +643,7 @@ void mlx5e_rx_res_rss_get_rxfh(struct mlx5e_rx_res *res, u32 *indir, u8 *key, u8
*hfunc = res->rss_params.hash.hfunc;
}
-static int mlx5e_rx_res_rss_update_tir(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt,
+static int mlx5e_rx_res_rss_update_tir(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt,
bool inner)
{
struct mlx5e_rss_params_traffic_type rss_tt;
@@ -668,7 +668,7 @@ static int mlx5e_rx_res_rss_update_tir(struct mlx5e_rx_res *res, enum mlx5e_traf
int mlx5e_rx_res_rss_set_rxfh(struct mlx5e_rx_res *res, const u32 *indir,
const u8 *key, const u8 *hfunc)
{
- enum mlx5e_traffic_types tt;
+ enum mlx5_traffic_types tt;
bool changed_indir = false;
bool changed_hash = false;
int err;
@@ -730,12 +730,12 @@ int mlx5e_rx_res_rss_set_rxfh(struct mlx5e_rx_res *res, const u32 *indir,
return 0;
}
-u8 mlx5e_rx_res_rss_get_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt)
+u8 mlx5e_rx_res_rss_get_hash_fields(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt)
{
return res->rss_params.rx_hash_fields[tt];
}
-int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt,
+int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt,
u8 rx_hash_fields)
{
u8 old_rx_hash_fields;
@@ -778,7 +778,7 @@ int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffi
int mlx5e_rx_res_lro_set_param(struct mlx5e_rx_res *res, struct mlx5e_lro_param *lro_param)
{
struct mlx5e_tir_builder *builder;
- enum mlx5e_traffic_types tt;
+ enum mlx5_traffic_types tt;
int err, final_err;
unsigned int ix;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
index 934e41a0761f..1baeec5158a3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h
@@ -21,7 +21,7 @@ enum mlx5e_rx_res_features {
};
struct mlx5e_rss_params_traffic_type
-mlx5e_rss_get_default_tt_config(enum mlx5e_traffic_types tt);
+mlx5e_rss_get_default_tt_config(enum mlx5_traffic_types tt);
/* Setup */
struct mlx5e_rx_res *mlx5e_rx_res_alloc(void);
@@ -35,8 +35,8 @@ void mlx5e_rx_res_free(struct mlx5e_rx_res *res);
/* TIRN getters for flow steering */
u32 mlx5e_rx_res_get_tirn_direct(struct mlx5e_rx_res *res, unsigned int ix);
u32 mlx5e_rx_res_get_tirn_xsk(struct mlx5e_rx_res *res, unsigned int ix);
-u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
-u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
+u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt);
+u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt);
u32 mlx5e_rx_res_get_tirn_ptp(struct mlx5e_rx_res *res);
/* RQTN getters for modules that create their own TIRs */
@@ -51,13 +51,13 @@ int mlx5e_rx_res_xsk_deactivate(struct mlx5e_rx_res *res, unsigned int ix);
/* Configuration API */
struct mlx5e_rss_params_traffic_type
-mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
+mlx5e_rx_res_rss_get_current_tt_config(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt);
void mlx5e_rx_res_rss_set_indir_uniform(struct mlx5e_rx_res *res, unsigned int nch);
void mlx5e_rx_res_rss_get_rxfh(struct mlx5e_rx_res *res, u32 *indir, u8 *key, u8 *hfunc);
int mlx5e_rx_res_rss_set_rxfh(struct mlx5e_rx_res *res, const u32 *indir,
const u8 *key, const u8 *hfunc);
-u8 mlx5e_rx_res_rss_get_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt);
-int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5e_traffic_types tt,
+u8 mlx5e_rx_res_rss_get_hash_fields(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt);
+int mlx5e_rx_res_rss_set_hash_fields(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt,
u8 rx_hash_fields);
int mlx5e_rx_res_lro_set_param(struct mlx5e_rx_res *res, struct mlx5e_lro_param *lro_param);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
index e51f60b55daa..90095507a2ca 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
@@ -16,13 +16,13 @@ struct mlx5e_accel_fs_tcp {
struct mlx5_flow_handle *default_rules[ACCEL_FS_TCP_NUM_TYPES];
};
-static enum mlx5e_traffic_types fs_accel2tt(enum accel_fs_tcp_type i)
+static enum mlx5_traffic_types fs_accel2tt(enum accel_fs_tcp_type i)
{
switch (i) {
case ACCEL_FS_IPV4_TCP:
- return MLX5E_TT_IPV4_TCP;
+ return MLX5_TT_IPV4_TCP;
default: /* ACCEL_FS_IPV6_TCP */
- return MLX5E_TT_IPV6_TCP;
+ return MLX5_TT_IPV6_TCP;
}
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
index 34119ce92031..9d9e40a64d0c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
@@ -41,11 +41,11 @@ struct mlx5e_ipsec_tx {
};
/* IPsec RX flow steering */
-static enum mlx5e_traffic_types fs_esp2tt(enum accel_fs_esp_type i)
+static enum mlx5_traffic_types fs_esp2tt(enum accel_fs_esp_type i)
{
if (i == ACCEL_FS_ESP4)
- return MLX5E_TT_IPV4_IPSEC_ESP;
- return MLX5E_TT_IPV6_IPSEC_ESP;
+ return MLX5_TT_IPV4_IPSEC_ESP;
+ return MLX5_TT_IPV6_IPSEC_ESP;
}
static int rx_err_add_rule(struct mlx5e_priv *priv,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
index 5077367f3ea0..a9c984fb0447 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
@@ -98,17 +98,17 @@ struct arfs_rule {
for (j = 0; j < ARFS_HASH_SIZE; j++) \
hlist_for_each_entry_safe(hn, tmp, &hash[j], hlist)
-static enum mlx5e_traffic_types arfs_get_tt(enum arfs_type type)
+static enum mlx5_traffic_types arfs_get_tt(enum arfs_type type)
{
switch (type) {
case ARFS_IPV4_TCP:
- return MLX5E_TT_IPV4_TCP;
+ return MLX5_TT_IPV4_TCP;
case ARFS_IPV4_UDP:
- return MLX5E_TT_IPV4_UDP;
+ return MLX5_TT_IPV4_UDP;
case ARFS_IPV6_TCP:
- return MLX5E_TT_IPV6_TCP;
+ return MLX5_TT_IPV6_TCP;
case ARFS_IPV6_UDP:
- return MLX5E_TT_IPV6_UDP;
+ return MLX5_TT_IPV6_UDP;
default:
return -EINVAL;
}
@@ -194,7 +194,7 @@ static int arfs_add_default_rule(struct mlx5e_priv *priv,
struct arfs_table *arfs_t = &priv->fs.arfs->arfs_tables[type];
struct mlx5_flow_destination dest = {};
MLX5_DECLARE_FLOW_ACT(flow_act);
- enum mlx5e_traffic_types tt;
+ enum mlx5_traffic_types tt;
int err = 0;
dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index 776f73cb592b..65bc1b745bb8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -858,14 +858,14 @@ static void mlx5e_cleanup_ttc_rules(struct mlx5e_ttc_table *ttc)
{
int i;
- for (i = 0; i < MLX5E_NUM_TT; i++) {
+ for (i = 0; i < MLX5_NUM_TT; i++) {
if (!IS_ERR_OR_NULL(ttc->rules[i].rule)) {
mlx5_del_flow_rules(ttc->rules[i].rule);
ttc->rules[i].rule = NULL;
}
}
- for (i = 0; i < MLX5E_NUM_TUNNEL_TT; i++) {
+ for (i = 0; i < MLX5_NUM_TUNNEL_TT; i++) {
if (!IS_ERR_OR_NULL(ttc->tunnel_rules[i])) {
mlx5_del_flow_rules(ttc->tunnel_rules[i]);
ttc->tunnel_rules[i] = NULL;
@@ -879,81 +879,81 @@ struct mlx5e_etype_proto {
};
static struct mlx5e_etype_proto ttc_rules[] = {
- [MLX5E_TT_IPV4_TCP] = {
+ [MLX5_TT_IPV4_TCP] = {
.etype = ETH_P_IP,
.proto = IPPROTO_TCP,
},
- [MLX5E_TT_IPV6_TCP] = {
+ [MLX5_TT_IPV6_TCP] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_TCP,
},
- [MLX5E_TT_IPV4_UDP] = {
+ [MLX5_TT_IPV4_UDP] = {
.etype = ETH_P_IP,
.proto = IPPROTO_UDP,
},
- [MLX5E_TT_IPV6_UDP] = {
+ [MLX5_TT_IPV6_UDP] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_UDP,
},
- [MLX5E_TT_IPV4_IPSEC_AH] = {
+ [MLX5_TT_IPV4_IPSEC_AH] = {
.etype = ETH_P_IP,
.proto = IPPROTO_AH,
},
- [MLX5E_TT_IPV6_IPSEC_AH] = {
+ [MLX5_TT_IPV6_IPSEC_AH] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_AH,
},
- [MLX5E_TT_IPV4_IPSEC_ESP] = {
+ [MLX5_TT_IPV4_IPSEC_ESP] = {
.etype = ETH_P_IP,
.proto = IPPROTO_ESP,
},
- [MLX5E_TT_IPV6_IPSEC_ESP] = {
+ [MLX5_TT_IPV6_IPSEC_ESP] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_ESP,
},
- [MLX5E_TT_IPV4] = {
+ [MLX5_TT_IPV4] = {
.etype = ETH_P_IP,
.proto = 0,
},
- [MLX5E_TT_IPV6] = {
+ [MLX5_TT_IPV6] = {
.etype = ETH_P_IPV6,
.proto = 0,
},
- [MLX5E_TT_ANY] = {
+ [MLX5_TT_ANY] = {
.etype = 0,
.proto = 0,
},
};
static struct mlx5e_etype_proto ttc_tunnel_rules[] = {
- [MLX5E_TT_IPV4_GRE] = {
+ [MLX5_TT_IPV4_GRE] = {
.etype = ETH_P_IP,
.proto = IPPROTO_GRE,
},
- [MLX5E_TT_IPV6_GRE] = {
+ [MLX5_TT_IPV6_GRE] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_GRE,
},
- [MLX5E_TT_IPV4_IPIP] = {
+ [MLX5_TT_IPV4_IPIP] = {
.etype = ETH_P_IP,
.proto = IPPROTO_IPIP,
},
- [MLX5E_TT_IPV6_IPIP] = {
+ [MLX5_TT_IPV6_IPIP] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_IPIP,
},
- [MLX5E_TT_IPV4_IPV6] = {
+ [MLX5_TT_IPV4_IPV6] = {
.etype = ETH_P_IP,
.proto = IPPROTO_IPV6,
},
- [MLX5E_TT_IPV6_IPV6] = {
+ [MLX5_TT_IPV6_IPV6] = {
.etype = ETH_P_IPV6,
.proto = IPPROTO_IPV6,
},
};
-u8 mlx5e_get_proto_by_tunnel_type(enum mlx5e_tunnel_types tt)
+u8 mlx5e_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt)
{
return ttc_tunnel_rules[tt].proto;
}
@@ -976,7 +976,7 @@ static bool mlx5e_tunnel_any_rx_proto_supported(struct mlx5_core_dev *mdev)
{
int tt;
- for (tt = 0; tt < MLX5E_NUM_TUNNEL_TT; tt++) {
+ for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
if (mlx5e_tunnel_proto_supported_rx(mdev, ttc_tunnel_rules[tt].proto))
return true;
}
@@ -1060,10 +1060,10 @@ static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv,
rules = ttc->rules;
dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
- for (tt = 0; tt < MLX5E_NUM_TT; tt++) {
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
struct mlx5e_ttc_rule *rule = &rules[tt];
- if (tt == MLX5E_TT_ANY)
+ if (tt == MLX5_TT_ANY)
dest.tir_num = params->any_tt_tirn;
else
dest.tir_num = params->indir_tirn[tt];
@@ -1084,8 +1084,8 @@ static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv,
trules = ttc->tunnel_rules;
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- dest.ft = params->inner_ttc->ft.t;
- for (tt = 0; tt < MLX5E_NUM_TUNNEL_TT; tt++) {
+ dest.ft = params->inner_ttc->ft.t;
+ for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
if (!mlx5e_tunnel_proto_supported_rx(priv->mdev,
ttc_tunnel_rules[tt].proto))
continue;
@@ -1116,7 +1116,7 @@ static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc,
int err;
u8 *mc;
- ft->g = kcalloc(MLX5E_TTC_NUM_GROUPS,
+ ft->g = kcalloc(MLX5_TTC_NUM_GROUPS,
sizeof(*ft->g), GFP_KERNEL);
if (!ft->g)
return -ENOMEM;
@@ -1136,7 +1136,7 @@ static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc,
MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ethertype);
MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS);
MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5E_TTC_GROUP1_SIZE;
+ ix += MLX5_TTC_GROUP1_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
if (IS_ERR(ft->g[ft->num_groups]))
@@ -1146,7 +1146,7 @@ static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc,
/* L3 Group */
MLX5_SET(fte_match_param, mc, outer_headers.ip_protocol, 0);
MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5E_TTC_GROUP2_SIZE;
+ ix += MLX5_TTC_GROUP2_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
if (IS_ERR(ft->g[ft->num_groups]))
@@ -1156,7 +1156,7 @@ static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc,
/* Any Group */
memset(in, 0, inlen);
MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5E_TTC_GROUP3_SIZE;
+ ix += MLX5_TTC_GROUP3_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
if (IS_ERR(ft->g[ft->num_groups]))
@@ -1227,10 +1227,10 @@ static int mlx5e_generate_inner_ttc_table_rules(struct mlx5e_priv *priv,
rules = ttc->rules;
dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
- for (tt = 0; tt < MLX5E_NUM_TT; tt++) {
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
struct mlx5e_ttc_rule *rule = &rules[tt];
- if (tt == MLX5E_TT_ANY)
+ if (tt == MLX5_TT_ANY)
dest.tir_num = params->any_tt_tirn;
else
dest.tir_num = params->indir_tirn[tt];
@@ -1263,7 +1263,7 @@ static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
int err;
u8 *mc;
- ft->g = kcalloc(MLX5E_INNER_TTC_NUM_GROUPS, sizeof(*ft->g), GFP_KERNEL);
+ ft->g = kcalloc(MLX5_INNER_TTC_NUM_GROUPS, sizeof(*ft->g), GFP_KERNEL);
if (!ft->g)
return -ENOMEM;
in = kvzalloc(inlen, GFP_KERNEL);
@@ -1279,7 +1279,7 @@ static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
MLX5_SET_TO_ONES(fte_match_param, mc, inner_headers.ip_version);
MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_INNER_HEADERS);
MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5E_INNER_TTC_GROUP1_SIZE;
+ ix += MLX5_INNER_TTC_GROUP1_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
if (IS_ERR(ft->g[ft->num_groups]))
@@ -1289,7 +1289,7 @@ static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
/* L3 Group */
MLX5_SET(fte_match_param, mc, inner_headers.ip_protocol, 0);
MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5E_INNER_TTC_GROUP2_SIZE;
+ ix += MLX5_INNER_TTC_GROUP2_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
if (IS_ERR(ft->g[ft->num_groups]))
@@ -1299,7 +1299,7 @@ static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
/* Any Group */
memset(in, 0, inlen);
MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5E_INNER_TTC_GROUP3_SIZE;
+ ix += MLX5_INNER_TTC_GROUP3_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
if (IS_ERR(ft->g[ft->num_groups]))
@@ -1328,7 +1328,7 @@ static void mlx5e_set_inner_ttc_ft_params(struct ttc_params *ttc_params)
{
struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr;
- ft_attr->max_fte = MLX5E_INNER_TTC_TABLE_SIZE;
+ ft_attr->max_fte = MLX5_INNER_TTC_TABLE_SIZE;
ft_attr->level = MLX5E_INNER_TTC_FT_LEVEL;
ft_attr->prio = MLX5E_NIC_PRIO;
}
@@ -1338,7 +1338,7 @@ void mlx5e_set_ttc_ft_params(struct ttc_params *ttc_params)
{
struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr;
- ft_attr->max_fte = MLX5E_TTC_TABLE_SIZE;
+ ft_attr->max_fte = MLX5_TTC_TABLE_SIZE;
ft_attr->level = MLX5E_TTC_FT_LEVEL;
ft_attr->prio = MLX5E_NIC_PRIO;
}
@@ -1413,14 +1413,14 @@ int mlx5e_create_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params,
return err;
}
-int mlx5e_ttc_fwd_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types type,
+int mlx5e_ttc_fwd_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type,
struct mlx5_flow_destination *new_dest)
{
return mlx5_modify_rule_destination(priv->fs.ttc.rules[type].rule, new_dest, NULL);
}
struct mlx5_flow_destination
-mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types type)
+mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type)
{
struct mlx5_flow_destination *dest = &priv->fs.ttc.rules[type].default_dest;
@@ -1430,7 +1430,7 @@ mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types typ
return *dest;
}
-int mlx5e_ttc_fwd_default_dest(struct mlx5e_priv *priv, enum mlx5e_traffic_types type)
+int mlx5e_ttc_fwd_default_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type)
{
struct mlx5_flow_destination dest = mlx5e_ttc_get_default_dest(priv, type);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
index c057f830a15d..3d8918f9399e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
@@ -786,44 +786,44 @@ void mlx5e_ethtool_init_steering(struct mlx5e_priv *priv)
INIT_LIST_HEAD(&priv->fs.ethtool.rules);
}
-static enum mlx5e_traffic_types flow_type_to_traffic_type(u32 flow_type)
+static int flow_type_to_traffic_type(u32 flow_type)
{
switch (flow_type) {
case TCP_V4_FLOW:
- return MLX5E_TT_IPV4_TCP;
+ return MLX5_TT_IPV4_TCP;
case TCP_V6_FLOW:
- return MLX5E_TT_IPV6_TCP;
+ return MLX5_TT_IPV6_TCP;
case UDP_V4_FLOW:
- return MLX5E_TT_IPV4_UDP;
+ return MLX5_TT_IPV4_UDP;
case UDP_V6_FLOW:
- return MLX5E_TT_IPV6_UDP;
+ return MLX5_TT_IPV6_UDP;
case AH_V4_FLOW:
- return MLX5E_TT_IPV4_IPSEC_AH;
+ return MLX5_TT_IPV4_IPSEC_AH;
case AH_V6_FLOW:
- return MLX5E_TT_IPV6_IPSEC_AH;
+ return MLX5_TT_IPV6_IPSEC_AH;
case ESP_V4_FLOW:
- return MLX5E_TT_IPV4_IPSEC_ESP;
+ return MLX5_TT_IPV4_IPSEC_ESP;
case ESP_V6_FLOW:
- return MLX5E_TT_IPV6_IPSEC_ESP;
+ return MLX5_TT_IPV6_IPSEC_ESP;
case IPV4_FLOW:
- return MLX5E_TT_IPV4;
+ return MLX5_TT_IPV4;
case IPV6_FLOW:
- return MLX5E_TT_IPV6;
+ return MLX5_TT_IPV6;
default:
- return MLX5E_NUM_INDIR_TIRS;
+ return -EINVAL;
}
}
static int mlx5e_set_rss_hash_opt(struct mlx5e_priv *priv,
struct ethtool_rxnfc *nfc)
{
- enum mlx5e_traffic_types tt;
u8 rx_hash_field = 0;
int err;
+ int tt;
tt = flow_type_to_traffic_type(nfc->flow_type);
- if (tt == MLX5E_NUM_INDIR_TIRS)
- return -EINVAL;
+ if (tt < 0)
+ return tt;
/* RSS does not support anything other than hashing to queues
* on src IP, dest IP, TCP/UDP src port and TCP/UDP dest
@@ -858,12 +858,12 @@ static int mlx5e_set_rss_hash_opt(struct mlx5e_priv *priv,
static int mlx5e_get_rss_hash_opt(struct mlx5e_priv *priv,
struct ethtool_rxnfc *nfc)
{
- enum mlx5e_traffic_types tt;
u32 hash_field = 0;
+ int tt;
tt = flow_type_to_traffic_type(nfc->flow_type);
- if (tt == MLX5E_NUM_INDIR_TIRS)
- return -EINVAL;
+ if (tt < 0)
+ return tt;
hash_field = mlx5e_rx_res_rss_get_hash_fields(priv->rx_res, tt);
nfc->data = 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 6797328e0afd..c1469f5755b5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -4211,7 +4211,7 @@ static bool mlx5e_tunnel_any_tx_proto_supported(struct mlx5_core_dev *mdev)
{
int tt;
- for (tt = 0; tt < MLX5E_NUM_TUNNEL_TT; tt++) {
+ for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
if (mlx5e_tunnel_proto_supported_tx(mdev, mlx5e_get_proto_by_tunnel_type(tt)))
return true;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 04687ffaeffa..300a37c83c17 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -538,7 +538,7 @@ static int mlx5e_hairpin_create_indirect_tirs(struct mlx5e_hairpin *hp)
{
struct mlx5e_priv *priv = hp->func_priv;
struct mlx5e_rss_params_hash rss_hash;
- enum mlx5e_traffic_types tt, max_tt;
+ enum mlx5_traffic_types tt, max_tt;
struct mlx5e_tir_builder *builder;
int err = 0;
@@ -600,7 +600,7 @@ static void mlx5e_hairpin_set_ttc_params(struct mlx5e_hairpin *hp,
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
ttc_params->indir_tirn[tt] = mlx5e_tir_get_tirn(&hp->indir_tir[tt]);
- ft_attr->max_fte = MLX5E_TTC_TABLE_SIZE;
+ ft_attr->max_fte = MLX5_TTC_TABLE_SIZE;
ft_attr->level = MLX5E_TC_TTC_FT_LEVEL;
ft_attr->prio = MLX5E_TC_PRIO;
}
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 06/16] net/mlx5e: Rename some related TTC args and functions
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (4 preceding siblings ...)
2021-08-03 2:28 ` [net-next 05/16] net/mlx5e: Rename traffic type enums Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 07/16] net/mlx5e: Decouple TTC logic from mlx5e Saeed Mahameed
` (9 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maor Gottlieb, Tariq Toukan, Mark Bloch, Saeed Mahameed
From: Maor Gottlieb <maorg@nvidia.com>
Since TTC logic is going to be moved to a separate file, make the
relevant functions and arguments that used by TTC to be mlx5 generic.
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/en/fs.h | 4 +-
.../net/ethernet/mellanox/mlx5/core/en_fs.c | 41 ++++++++++---------
.../net/ethernet/mellanox/mlx5/core/en_main.c | 4 +-
3 files changed, 26 insertions(+), 23 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
index 77fe98c42ec4..6b01a28e1d93 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
@@ -104,7 +104,7 @@ enum mlx5_tunnel_types {
MLX5_NUM_TUNNEL_TT,
};
-bool mlx5e_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev);
+bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev);
struct mlx5e_ttc_rule {
struct mlx5_flow_handle *rule;
@@ -266,7 +266,7 @@ void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv);
int mlx5e_create_flow_steering(struct mlx5e_priv *priv);
void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv);
-u8 mlx5e_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt);
+u8 mlx5_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt);
int mlx5e_add_vlan_trap(struct mlx5e_priv *priv, int trap_id, int tir_num);
void mlx5e_remove_vlan_trap(struct mlx5e_priv *priv);
int mlx5e_add_mac_trap(struct mlx5e_priv *priv, int trap_id, int tir_num);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index 65bc1b745bb8..14a9011ea1a1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -873,12 +873,12 @@ static void mlx5e_cleanup_ttc_rules(struct mlx5e_ttc_table *ttc)
}
}
-struct mlx5e_etype_proto {
+struct mlx5_etype_proto {
u16 etype;
u8 proto;
};
-static struct mlx5e_etype_proto ttc_rules[] = {
+static struct mlx5_etype_proto ttc_rules[] = {
[MLX5_TT_IPV4_TCP] = {
.etype = ETH_P_IP,
.proto = IPPROTO_TCP,
@@ -925,7 +925,7 @@ static struct mlx5e_etype_proto ttc_rules[] = {
},
};
-static struct mlx5e_etype_proto ttc_tunnel_rules[] = {
+static struct mlx5_etype_proto ttc_tunnel_rules[] = {
[MLX5_TT_IPV4_GRE] = {
.etype = ETH_P_IP,
.proto = IPPROTO_GRE,
@@ -953,12 +953,13 @@ static struct mlx5e_etype_proto ttc_tunnel_rules[] = {
};
-u8 mlx5e_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt)
+u8 mlx5_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt)
{
return ttc_tunnel_rules[tt].proto;
}
-static bool mlx5e_tunnel_proto_supported_rx(struct mlx5_core_dev *mdev, u8 proto_type)
+static bool mlx5_tunnel_proto_supported_rx(struct mlx5_core_dev *mdev,
+ u8 proto_type)
{
switch (proto_type) {
case IPPROTO_GRE:
@@ -972,24 +973,26 @@ static bool mlx5e_tunnel_proto_supported_rx(struct mlx5_core_dev *mdev, u8 proto
}
}
-static bool mlx5e_tunnel_any_rx_proto_supported(struct mlx5_core_dev *mdev)
+static bool mlx5_tunnel_any_rx_proto_supported(struct mlx5_core_dev *mdev)
{
int tt;
for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
- if (mlx5e_tunnel_proto_supported_rx(mdev, ttc_tunnel_rules[tt].proto))
+ if (mlx5_tunnel_proto_supported_rx(mdev,
+ ttc_tunnel_rules[tt].proto))
return true;
}
return false;
}
-bool mlx5e_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev)
+bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev)
{
- return (mlx5e_tunnel_any_rx_proto_supported(mdev) &&
- MLX5_CAP_FLOWTABLE_NIC_RX(mdev, ft_field_support.inner_ip_version));
+ return (mlx5_tunnel_any_rx_proto_supported(mdev) &&
+ MLX5_CAP_FLOWTABLE_NIC_RX(mdev,
+ ft_field_support.inner_ip_version));
}
-static u8 mlx5e_etype_to_ipv(u16 ethertype)
+static u8 mlx5_etype_to_ipv(u16 ethertype)
{
if (ethertype == ETH_P_IP)
return 4;
@@ -1024,7 +1027,7 @@ mlx5e_generate_ttc_rule(struct mlx5e_priv *priv,
MLX5_SET(fte_match_param, spec->match_value, outer_headers.ip_protocol, proto);
}
- ipv = mlx5e_etype_to_ipv(etype);
+ ipv = mlx5_etype_to_ipv(etype);
if (match_ipv_outer && ipv) {
spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ip_version);
@@ -1079,15 +1082,15 @@ static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv,
rule->default_dest = dest;
}
- if (!params->inner_ttc || !mlx5e_tunnel_inner_ft_supported(priv->mdev))
+ if (!params->inner_ttc || !mlx5_tunnel_inner_ft_supported(priv->mdev))
return 0;
trules = ttc->tunnel_rules;
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
dest.ft = params->inner_ttc->ft.t;
for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
- if (!mlx5e_tunnel_proto_supported_rx(priv->mdev,
- ttc_tunnel_rules[tt].proto))
+ if (!mlx5_tunnel_proto_supported_rx(priv->mdev,
+ ttc_tunnel_rules[tt].proto))
continue;
trules[tt] = mlx5e_generate_ttc_rule(priv, ft, &dest,
ttc_tunnel_rules[tt].etype,
@@ -1190,7 +1193,7 @@ mlx5e_generate_inner_ttc_rule(struct mlx5e_priv *priv,
if (!spec)
return ERR_PTR(-ENOMEM);
- ipv = mlx5e_etype_to_ipv(etype);
+ ipv = mlx5_etype_to_ipv(etype);
if (etype && ipv) {
spec->match_criteria_enable = MLX5_MATCH_INNER_HEADERS;
MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, inner_headers.ip_version);
@@ -1783,7 +1786,7 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
mlx5e_set_ttc_basic_params(priv, &ttc_params);
- if (mlx5e_tunnel_inner_ft_supported(priv->mdev)) {
+ if (mlx5_tunnel_inner_ft_supported(priv->mdev)) {
mlx5e_set_inner_ttc_ft_params(&ttc_params);
for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
ttc_params.indir_tirn[tt] =
@@ -1837,7 +1840,7 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
err_destroy_ttc_table:
mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
err_destroy_inner_ttc_table:
- if (mlx5e_tunnel_inner_ft_supported(priv->mdev))
+ if (mlx5_tunnel_inner_ft_supported(priv->mdev))
mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc);
err_destroy_arfs_tables:
mlx5e_arfs_destroy_tables(priv);
@@ -1851,7 +1854,7 @@ void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv)
mlx5e_destroy_vlan_table(priv);
mlx5e_destroy_l2_table(priv);
mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
- if (mlx5e_tunnel_inner_ft_supported(priv->mdev))
+ if (mlx5_tunnel_inner_ft_supported(priv->mdev))
mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc);
mlx5e_arfs_destroy_tables(priv);
mlx5e_ethtool_cleanup_steering(priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index c1469f5755b5..25a0b5f0984a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -4151,7 +4151,7 @@ void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, u16
/* TX inline */
mlx5_query_min_inline(mdev, ¶ms->tx_min_inline_mode);
- params->tunneled_offload_en = mlx5e_tunnel_inner_ft_supported(mdev);
+ params->tunneled_offload_en = mlx5_tunnel_inner_ft_supported(mdev);
/* AF_XDP */
params->xsk = xsk;
@@ -4212,7 +4212,7 @@ static bool mlx5e_tunnel_any_tx_proto_supported(struct mlx5_core_dev *mdev)
int tt;
for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
- if (mlx5e_tunnel_proto_supported_tx(mdev, mlx5e_get_proto_by_tunnel_type(tt)))
+ if (mlx5e_tunnel_proto_supported_tx(mdev, mlx5_get_proto_by_tunnel_type(tt)))
return true;
}
return (mlx5_vxlan_allowed(mdev->vxlan) || mlx5_geneve_tx_allowed(mdev));
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 07/16] net/mlx5e: Decouple TTC logic from mlx5e
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (5 preceding siblings ...)
2021-08-03 2:28 ` [net-next 06/16] net/mlx5e: Rename some related TTC args and functions Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 08/16] net/mlx5: Move TTC logic to fs_ttc Saeed Mahameed
` (8 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maor Gottlieb, Tariq Toukan, Mark Bloch, Saeed Mahameed
From: Maor Gottlieb <maorg@nvidia.com>
Remove dependency in the mlx5e driver from the TTC implementation
by changing the TTC related functions to receive mlx5 generic arguments.
It allows to decouple TTC logic from mlx5e and reused by other parts of
mlx5 driver.
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/en/fs.h | 44 +--
.../mellanox/mlx5/core/en/fs_tt_redirect.c | 13 +-
.../mellanox/mlx5/core/en_accel/fs_tcp.c | 6 +-
.../mellanox/mlx5/core/en_accel/ipsec_fs.c | 6 +-
.../net/ethernet/mellanox/mlx5/core/en_arfs.c | 6 +-
.../net/ethernet/mellanox/mlx5/core/en_fs.c | 355 +++++++++---------
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 17 +-
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 26 +-
.../ethernet/mellanox/mlx5/core/ipoib/ipoib.c | 12 +-
9 files changed, 250 insertions(+), 235 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
index 6b01a28e1d93..c289f7004e10 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
@@ -106,15 +106,17 @@ enum mlx5_tunnel_types {
bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev);
-struct mlx5e_ttc_rule {
+struct mlx5_ttc_rule {
struct mlx5_flow_handle *rule;
struct mlx5_flow_destination default_dest;
};
/* L3/L4 traffic type classifier */
-struct mlx5e_ttc_table {
- struct mlx5e_flow_table ft;
- struct mlx5e_ttc_rule rules[MLX5_NUM_TT];
+struct mlx5_ttc_table {
+ int num_groups;
+ struct mlx5_flow_table *t;
+ struct mlx5_flow_group **g;
+ struct mlx5_ttc_rule rules[MLX5_NUM_TT];
struct mlx5_flow_handle *tunnel_rules[MLX5_NUM_TUNNEL_TT];
};
@@ -223,8 +225,8 @@ struct mlx5e_flow_steering {
struct mlx5e_promisc_table promisc;
struct mlx5e_vlan_table *vlan;
struct mlx5e_l2_table l2;
- struct mlx5e_ttc_table ttc;
- struct mlx5e_ttc_table inner_ttc;
+ struct mlx5_ttc_table ttc;
+ struct mlx5_ttc_table inner_ttc;
#ifdef CONFIG_MLX5_EN_ARFS
struct mlx5e_arfs_tables *arfs;
#endif
@@ -237,28 +239,28 @@ struct mlx5e_flow_steering {
};
struct ttc_params {
+ struct mlx5_flow_namespace *ns;
struct mlx5_flow_table_attr ft_attr;
- u32 any_tt_tirn;
- u32 indir_tirn[MLX5E_NUM_INDIR_TIRS];
- struct mlx5e_ttc_table *inner_ttc;
+ struct mlx5_flow_destination dests[MLX5_NUM_TT];
+ bool inner_ttc;
+ struct mlx5_flow_destination tunnel_dests[MLX5_NUM_TUNNEL_TT];
};
-void mlx5e_set_ttc_basic_params(struct mlx5e_priv *priv, struct ttc_params *ttc_params);
-void mlx5e_set_ttc_ft_params(struct ttc_params *ttc_params);
+void mlx5e_set_ttc_params(struct mlx5e_priv *priv,
+ struct ttc_params *ttc_params, bool tunnel);
-int mlx5e_create_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params,
- struct mlx5e_ttc_table *ttc);
-void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv,
- struct mlx5e_ttc_table *ttc);
+int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
+ struct mlx5_ttc_table *ttc);
+void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc);
void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft);
-int mlx5e_ttc_fwd_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type,
- struct mlx5_flow_destination *new_dest);
+int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
+ struct mlx5_flow_destination *new_dest);
struct mlx5_flow_destination
-mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv,
- enum mlx5_traffic_types type);
-int mlx5e_ttc_fwd_default_dest(struct mlx5e_priv *priv,
- enum mlx5_traffic_types type);
+mlx5_ttc_get_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type);
+int mlx5_ttc_fwd_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type);
void mlx5e_enable_cvlan_filter(struct mlx5e_priv *priv);
void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
index 5645e8032218..68cc3a8fd6b7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
@@ -124,7 +124,7 @@ static int fs_udp_add_default_rule(struct mlx5e_priv *priv, enum fs_udp_type typ
fs_udp = priv->fs.udp;
fs_udp_t = &fs_udp->tables[type];
- dest = mlx5e_ttc_get_default_dest(priv, fs_udp2tt(type));
+ dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, fs_udp2tt(type));
rule = mlx5_add_flow_rules(fs_udp_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -259,7 +259,7 @@ static int fs_udp_disable(struct mlx5e_priv *priv)
for (i = 0; i < FS_UDP_NUM_TYPES; i++) {
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5e_ttc_fwd_default_dest(priv, fs_udp2tt(i));
+ err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, fs_udp2tt(i));
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -281,7 +281,8 @@ static int fs_udp_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.udp->tables[i].t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5e_ttc_fwd_dest(priv, fs_udp2tt(i), &dest);
+ err = mlx5_ttc_fwd_dest(&priv->fs.ttc, fs_udp2tt(i),
+ &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
@@ -401,7 +402,7 @@ static int fs_any_add_default_rule(struct mlx5e_priv *priv)
fs_any = priv->fs.any;
fs_any_t = &fs_any->table;
- dest = mlx5e_ttc_get_default_dest(priv, MLX5_TT_ANY);
+ dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, MLX5_TT_ANY);
rule = mlx5_add_flow_rules(fs_any_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -514,7 +515,7 @@ static int fs_any_disable(struct mlx5e_priv *priv)
int err;
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5e_ttc_fwd_default_dest(priv, MLX5_TT_ANY);
+ err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, MLX5_TT_ANY);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -533,7 +534,7 @@ static int fs_any_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.any->table.t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5e_ttc_fwd_dest(priv, MLX5_TT_ANY, &dest);
+ err = mlx5_ttc_fwd_dest(&priv->fs.ttc, MLX5_TT_ANY, &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
index 90095507a2ca..a82be377e9f7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
@@ -161,7 +161,7 @@ static int accel_fs_tcp_add_default_rule(struct mlx5e_priv *priv,
fs_tcp = priv->fs.accel_tcp;
accel_fs_t = &fs_tcp->tables[type];
- dest = mlx5e_ttc_get_default_dest(priv, fs_accel2tt(type));
+ dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, fs_accel2tt(type));
rule = mlx5_add_flow_rules(accel_fs_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -307,7 +307,7 @@ static int accel_fs_tcp_disable(struct mlx5e_priv *priv)
for (i = 0; i < ACCEL_FS_TCP_NUM_TYPES; i++) {
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5e_ttc_fwd_default_dest(priv, fs_accel2tt(i));
+ err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, fs_accel2tt(i));
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -329,7 +329,7 @@ static int accel_fs_tcp_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.accel_tcp->tables[i].t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5e_ttc_fwd_dest(priv, fs_accel2tt(i), &dest);
+ err = mlx5_ttc_fwd_dest(&priv->fs.ttc, fs_accel2tt(i), &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
index 9d9e40a64d0c..ff177bb74bb4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
@@ -265,7 +265,7 @@ static int rx_create(struct mlx5e_priv *priv, enum accel_fs_esp_type type)
accel_esp = priv->ipsec->rx_fs;
fs_prot = &accel_esp->fs_prot[type];
- fs_prot->default_dest = mlx5e_ttc_get_default_dest(priv, fs_esp2tt(type));
+ fs_prot->default_dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, fs_esp2tt(type));
err = rx_err_create_ft(priv, fs_prot, &fs_prot->rx_err);
if (err)
@@ -301,7 +301,7 @@ static int rx_ft_get(struct mlx5e_priv *priv, enum accel_fs_esp_type type)
/* connect */
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
dest.ft = fs_prot->ft;
- mlx5e_ttc_fwd_dest(priv, fs_esp2tt(type), &dest);
+ mlx5_ttc_fwd_dest(&priv->fs.ttc, fs_esp2tt(type), &dest);
out:
mutex_unlock(&fs_prot->prot_mutex);
@@ -320,7 +320,7 @@ static void rx_ft_put(struct mlx5e_priv *priv, enum accel_fs_esp_type type)
goto out;
/* disconnect */
- mlx5e_ttc_fwd_default_dest(priv, fs_esp2tt(type));
+ mlx5_ttc_fwd_default_dest(&priv->fs.ttc, fs_esp2tt(type));
/* remove FT */
rx_destroy(priv, type);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
index a9c984fb0447..374e262d9917 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
@@ -120,7 +120,7 @@ static int arfs_disable(struct mlx5e_priv *priv)
for (i = 0; i < ARFS_NUM_TYPES; i++) {
/* Modify ttc rules destination back to their default */
- err = mlx5e_ttc_fwd_default_dest(priv, arfs_get_tt(i));
+ err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, arfs_get_tt(i));
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -149,7 +149,7 @@ int mlx5e_arfs_enable(struct mlx5e_priv *priv)
for (i = 0; i < ARFS_NUM_TYPES; i++) {
dest.ft = priv->fs.arfs->arfs_tables[i].ft.t;
/* Modify ttc rules destination to point on the aRFS FTs */
- err = mlx5e_ttc_fwd_dest(priv, arfs_get_tt(i), &dest);
+ err = mlx5_ttc_fwd_dest(&priv->fs.ttc, arfs_get_tt(i), &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] dest to arfs, failed err(%d)\n",
@@ -205,7 +205,7 @@ static int arfs_add_default_rule(struct mlx5e_priv *priv,
return -EINVAL;
}
- /* FIXME: Must use mlx5e_ttc_get_default_dest(),
+ /* FIXME: Must use mlx5_ttc_get_default_dest(),
* but can't since TTC default is not setup yet !
*/
dest.tir_num = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index 14a9011ea1a1..a03842d132f6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -718,7 +718,7 @@ static int mlx5e_add_promisc_rule(struct mlx5e_priv *priv)
if (!spec)
return -ENOMEM;
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- dest.ft = priv->fs.ttc.ft.t;
+ dest.ft = priv->fs.ttc.t;
rule_p = &priv->fs.promisc.rule;
*rule_p = mlx5_add_flow_rules(ft, spec, &flow_act, &dest, 1);
@@ -854,7 +854,7 @@ void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft)
ft->t = NULL;
}
-static void mlx5e_cleanup_ttc_rules(struct mlx5e_ttc_table *ttc)
+static void mlx5_cleanup_ttc_rules(struct mlx5_ttc_table *ttc)
{
int i;
@@ -1004,13 +1004,12 @@ static u8 mlx5_etype_to_ipv(u16 ethertype)
}
static struct mlx5_flow_handle *
-mlx5e_generate_ttc_rule(struct mlx5e_priv *priv,
- struct mlx5_flow_table *ft,
- struct mlx5_flow_destination *dest,
- u16 etype,
- u8 proto)
+mlx5_generate_ttc_rule(struct mlx5_core_dev *dev, struct mlx5_flow_table *ft,
+ struct mlx5_flow_destination *dest, u16 etype, u8 proto)
{
- int match_ipv_outer = MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ft_field_support.outer_ip_version);
+ int match_ipv_outer =
+ MLX5_CAP_FLOWTABLE_NIC_RX(dev,
+ ft_field_support.outer_ip_version);
MLX5_DECLARE_FLOW_ACT(flow_act);
struct mlx5_flow_handle *rule;
struct mlx5_flow_spec *spec;
@@ -1041,60 +1040,51 @@ mlx5e_generate_ttc_rule(struct mlx5e_priv *priv,
rule = mlx5_add_flow_rules(ft, spec, &flow_act, dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
- netdev_err(priv->netdev, "%s: add rule failed\n", __func__);
+ mlx5_core_err(dev, "%s: add rule failed\n", __func__);
}
kvfree(spec);
return err ? ERR_PTR(err) : rule;
}
-static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv,
- struct ttc_params *params,
- struct mlx5e_ttc_table *ttc)
+static int mlx5_generate_ttc_table_rules(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
{
- struct mlx5_flow_destination dest = {};
struct mlx5_flow_handle **trules;
- struct mlx5e_ttc_rule *rules;
+ struct mlx5_ttc_rule *rules;
struct mlx5_flow_table *ft;
int tt;
int err;
- ft = ttc->ft.t;
+ ft = ttc->t;
rules = ttc->rules;
-
- dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
for (tt = 0; tt < MLX5_NUM_TT; tt++) {
- struct mlx5e_ttc_rule *rule = &rules[tt];
-
- if (tt == MLX5_TT_ANY)
- dest.tir_num = params->any_tt_tirn;
- else
- dest.tir_num = params->indir_tirn[tt];
+ struct mlx5_ttc_rule *rule = &rules[tt];
- rule->rule = mlx5e_generate_ttc_rule(priv, ft, &dest,
- ttc_rules[tt].etype,
- ttc_rules[tt].proto);
+ rule->rule = mlx5_generate_ttc_rule(dev, ft, ¶ms->dests[tt],
+ ttc_rules[tt].etype,
+ ttc_rules[tt].proto);
if (IS_ERR(rule->rule)) {
err = PTR_ERR(rule->rule);
rule->rule = NULL;
goto del_rules;
}
- rule->default_dest = dest;
+ rule->default_dest = params->dests[tt];
}
- if (!params->inner_ttc || !mlx5_tunnel_inner_ft_supported(priv->mdev))
+ if (!params->inner_ttc || !mlx5_tunnel_inner_ft_supported(dev))
return 0;
trules = ttc->tunnel_rules;
- dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- dest.ft = params->inner_ttc->ft.t;
for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
- if (!mlx5_tunnel_proto_supported_rx(priv->mdev,
+ if (!mlx5_tunnel_proto_supported_rx(dev,
ttc_tunnel_rules[tt].proto))
continue;
- trules[tt] = mlx5e_generate_ttc_rule(priv, ft, &dest,
- ttc_tunnel_rules[tt].etype,
- ttc_tunnel_rules[tt].proto);
+ trules[tt] = mlx5_generate_ttc_rule(dev, ft,
+ ¶ms->tunnel_dests[tt],
+ ttc_tunnel_rules[tt].etype,
+ ttc_tunnel_rules[tt].proto);
if (IS_ERR(trules[tt])) {
err = PTR_ERR(trules[tt]);
trules[tt] = NULL;
@@ -1105,28 +1095,26 @@ static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv,
return 0;
del_rules:
- mlx5e_cleanup_ttc_rules(ttc);
+ mlx5_cleanup_ttc_rules(ttc);
return err;
}
-static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc,
- bool use_ipv)
+static int mlx5_create_ttc_table_groups(struct mlx5_ttc_table *ttc,
+ bool use_ipv)
{
int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
- struct mlx5e_flow_table *ft = &ttc->ft;
int ix = 0;
u32 *in;
int err;
u8 *mc;
- ft->g = kcalloc(MLX5_TTC_NUM_GROUPS,
- sizeof(*ft->g), GFP_KERNEL);
- if (!ft->g)
+ ttc->g = kcalloc(MLX5_TTC_NUM_GROUPS, sizeof(*ttc->g), GFP_KERNEL);
+ if (!ttc->g)
return -ENOMEM;
in = kvzalloc(inlen, GFP_KERNEL);
if (!in) {
- kfree(ft->g);
- ft->g = NULL;
+ kfree(ttc->g);
+ ttc->g = NULL;
return -ENOMEM;
}
@@ -1141,47 +1129,47 @@ static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc,
MLX5_SET_CFG(in, start_flow_index, ix);
ix += MLX5_TTC_GROUP1_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
- if (IS_ERR(ft->g[ft->num_groups]))
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
goto err;
- ft->num_groups++;
+ ttc->num_groups++;
/* L3 Group */
MLX5_SET(fte_match_param, mc, outer_headers.ip_protocol, 0);
MLX5_SET_CFG(in, start_flow_index, ix);
ix += MLX5_TTC_GROUP2_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
- if (IS_ERR(ft->g[ft->num_groups]))
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
goto err;
- ft->num_groups++;
+ ttc->num_groups++;
/* Any Group */
memset(in, 0, inlen);
MLX5_SET_CFG(in, start_flow_index, ix);
ix += MLX5_TTC_GROUP3_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
- if (IS_ERR(ft->g[ft->num_groups]))
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
goto err;
- ft->num_groups++;
+ ttc->num_groups++;
kvfree(in);
return 0;
err:
- err = PTR_ERR(ft->g[ft->num_groups]);
- ft->g[ft->num_groups] = NULL;
+ err = PTR_ERR(ttc->g[ttc->num_groups]);
+ ttc->g[ttc->num_groups] = NULL;
kvfree(in);
return err;
}
static struct mlx5_flow_handle *
-mlx5e_generate_inner_ttc_rule(struct mlx5e_priv *priv,
- struct mlx5_flow_table *ft,
- struct mlx5_flow_destination *dest,
- u16 etype, u8 proto)
+mlx5_generate_inner_ttc_rule(struct mlx5_core_dev *dev,
+ struct mlx5_flow_table *ft,
+ struct mlx5_flow_destination *dest,
+ u16 etype, u8 proto)
{
MLX5_DECLARE_FLOW_ACT(flow_act);
struct mlx5_flow_handle *rule;
@@ -1209,70 +1197,64 @@ mlx5e_generate_inner_ttc_rule(struct mlx5e_priv *priv,
rule = mlx5_add_flow_rules(ft, spec, &flow_act, dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
- netdev_err(priv->netdev, "%s: add rule failed\n", __func__);
+ mlx5_core_err(dev, "%s: add inner TTC rule failed\n", __func__);
}
kvfree(spec);
return err ? ERR_PTR(err) : rule;
}
-static int mlx5e_generate_inner_ttc_table_rules(struct mlx5e_priv *priv,
- struct ttc_params *params,
- struct mlx5e_ttc_table *ttc)
+static int mlx5_generate_inner_ttc_table_rules(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
{
- struct mlx5_flow_destination dest = {};
- struct mlx5e_ttc_rule *rules;
+ struct mlx5_ttc_rule *rules;
struct mlx5_flow_table *ft;
int err;
int tt;
- ft = ttc->ft.t;
+ ft = ttc->t;
rules = ttc->rules;
- dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR;
for (tt = 0; tt < MLX5_NUM_TT; tt++) {
- struct mlx5e_ttc_rule *rule = &rules[tt];
-
- if (tt == MLX5_TT_ANY)
- dest.tir_num = params->any_tt_tirn;
- else
- dest.tir_num = params->indir_tirn[tt];
+ struct mlx5_ttc_rule *rule = &rules[tt];
- rule->rule = mlx5e_generate_inner_ttc_rule(priv, ft, &dest,
- ttc_rules[tt].etype,
- ttc_rules[tt].proto);
+ rule->rule = mlx5_generate_inner_ttc_rule(dev, ft,
+ ¶ms->dests[tt],
+ ttc_rules[tt].etype,
+ ttc_rules[tt].proto);
if (IS_ERR(rule->rule)) {
err = PTR_ERR(rule->rule);
rule->rule = NULL;
goto del_rules;
}
- rule->default_dest = dest;
+ rule->default_dest = params->dests[tt];
}
return 0;
del_rules:
- mlx5e_cleanup_ttc_rules(ttc);
+ mlx5_cleanup_ttc_rules(ttc);
return err;
}
-static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
+static int mlx5_create_inner_ttc_table_groups(struct mlx5_ttc_table *ttc)
{
int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
- struct mlx5e_flow_table *ft = &ttc->ft;
int ix = 0;
u32 *in;
int err;
u8 *mc;
- ft->g = kcalloc(MLX5_INNER_TTC_NUM_GROUPS, sizeof(*ft->g), GFP_KERNEL);
- if (!ft->g)
+ ttc->g = kcalloc(MLX5_INNER_TTC_NUM_GROUPS, sizeof(*ttc->g),
+ GFP_KERNEL);
+ if (!ttc->g)
return -ENOMEM;
in = kvzalloc(inlen, GFP_KERNEL);
if (!in) {
- kfree(ft->g);
- ft->g = NULL;
+ kfree(ttc->g);
+ ttc->g = NULL;
return -ENOMEM;
}
@@ -1284,148 +1266,191 @@ static int mlx5e_create_inner_ttc_table_groups(struct mlx5e_ttc_table *ttc)
MLX5_SET_CFG(in, start_flow_index, ix);
ix += MLX5_INNER_TTC_GROUP1_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
- if (IS_ERR(ft->g[ft->num_groups]))
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
goto err;
- ft->num_groups++;
+ ttc->num_groups++;
/* L3 Group */
MLX5_SET(fte_match_param, mc, inner_headers.ip_protocol, 0);
MLX5_SET_CFG(in, start_flow_index, ix);
ix += MLX5_INNER_TTC_GROUP2_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
- if (IS_ERR(ft->g[ft->num_groups]))
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
goto err;
- ft->num_groups++;
+ ttc->num_groups++;
/* Any Group */
memset(in, 0, inlen);
MLX5_SET_CFG(in, start_flow_index, ix);
ix += MLX5_INNER_TTC_GROUP3_SIZE;
MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in);
- if (IS_ERR(ft->g[ft->num_groups]))
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
goto err;
- ft->num_groups++;
+ ttc->num_groups++;
kvfree(in);
return 0;
err:
- err = PTR_ERR(ft->g[ft->num_groups]);
- ft->g[ft->num_groups] = NULL;
+ err = PTR_ERR(ttc->g[ttc->num_groups]);
+ ttc->g[ttc->num_groups] = NULL;
kvfree(in);
return err;
}
-void mlx5e_set_ttc_basic_params(struct mlx5e_priv *priv,
- struct ttc_params *ttc_params)
-{
- ttc_params->any_tt_tirn = mlx5e_rx_res_get_tirn_direct(priv->rx_res, 0);
- ttc_params->inner_ttc = &priv->fs.inner_ttc;
-}
-
-static void mlx5e_set_inner_ttc_ft_params(struct ttc_params *ttc_params)
+static void mlx5e_set_inner_ttc_params(struct mlx5e_priv *priv,
+ struct ttc_params *ttc_params)
{
struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr;
+ int tt;
- ft_attr->max_fte = MLX5_INNER_TTC_TABLE_SIZE;
+ memset(ttc_params, 0, sizeof(*ttc_params));
+ ttc_params->ns = mlx5_get_flow_namespace(priv->mdev,
+ MLX5_FLOW_NAMESPACE_KERNEL);
ft_attr->level = MLX5E_INNER_TTC_FT_LEVEL;
ft_attr->prio = MLX5E_NIC_PRIO;
+
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
+ ttc_params->dests[tt].type = MLX5_FLOW_DESTINATION_TYPE_TIR;
+ ttc_params->dests[tt].tir_num =
+ tt == MLX5_TT_ANY ?
+ mlx5e_rx_res_get_tirn_direct(priv->rx_res, 0) :
+ mlx5e_rx_res_get_tirn_rss_inner(priv->rx_res,
+ tt);
+ }
}
-void mlx5e_set_ttc_ft_params(struct ttc_params *ttc_params)
+void mlx5e_set_ttc_params(struct mlx5e_priv *priv,
+ struct ttc_params *ttc_params, bool tunnel)
{
struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr;
+ int tt;
- ft_attr->max_fte = MLX5_TTC_TABLE_SIZE;
+ memset(ttc_params, 0, sizeof(*ttc_params));
+ ttc_params->ns = mlx5_get_flow_namespace(priv->mdev,
+ MLX5_FLOW_NAMESPACE_KERNEL);
ft_attr->level = MLX5E_TTC_FT_LEVEL;
ft_attr->prio = MLX5E_NIC_PRIO;
+
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
+ ttc_params->dests[tt].type = MLX5_FLOW_DESTINATION_TYPE_TIR;
+ ttc_params->dests[tt].tir_num =
+ tt == MLX5_TT_ANY ?
+ mlx5e_rx_res_get_tirn_direct(priv->rx_res, 0) :
+ mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
+ }
+
+ ttc_params->inner_ttc = tunnel;
+ if (!tunnel || !mlx5_tunnel_inner_ft_supported(priv->mdev))
+ return;
+
+ for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
+ ttc_params->tunnel_dests[tt].type =
+ MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
+ ttc_params->tunnel_dests[tt].ft = priv->fs.inner_ttc.t;
+ }
}
-static int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params,
- struct mlx5e_ttc_table *ttc)
+static int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
{
- struct mlx5e_flow_table *ft = &ttc->ft;
int err;
- ft->t = mlx5_create_flow_table(priv->fs.ns, ¶ms->ft_attr);
- if (IS_ERR(ft->t)) {
- err = PTR_ERR(ft->t);
- ft->t = NULL;
+ WARN_ON_ONCE(params->ft_attr.max_fte);
+ params->ft_attr.max_fte = MLX5_INNER_TTC_TABLE_SIZE;
+ ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
+ if (IS_ERR(ttc->t)) {
+ err = PTR_ERR(ttc->t);
+ ttc->t = NULL;
return err;
}
- err = mlx5e_create_inner_ttc_table_groups(ttc);
+ err = mlx5_create_inner_ttc_table_groups(ttc);
if (err)
- goto err;
+ goto destroy_ttc;
- err = mlx5e_generate_inner_ttc_table_rules(priv, params, ttc);
+ err = mlx5_generate_inner_ttc_table_rules(dev, params, ttc);
if (err)
- goto err;
+ goto destroy_ttc;
return 0;
-err:
- mlx5e_destroy_flow_table(ft);
+destroy_ttc:
+ mlx5_destroy_ttc_table(ttc);
return err;
}
-static void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv,
- struct mlx5e_ttc_table *ttc)
+void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc)
{
- mlx5e_cleanup_ttc_rules(ttc);
- mlx5e_destroy_flow_table(&ttc->ft);
+ int i;
+
+ mlx5_cleanup_ttc_rules(ttc);
+ for (i = ttc->num_groups - 1; i >= 0; i--) {
+ if (!IS_ERR_OR_NULL(ttc->g[i]))
+ mlx5_destroy_flow_group(ttc->g[i]);
+ ttc->g[i] = NULL;
+ }
+
+ ttc->num_groups = 0;
+ kfree(ttc->g);
+ mlx5_destroy_flow_table(ttc->t);
+ ttc->t = NULL;
}
-void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv,
- struct mlx5e_ttc_table *ttc)
+static void mlx5_destroy_inner_ttc_table(struct mlx5_ttc_table *ttc)
{
- mlx5e_cleanup_ttc_rules(ttc);
- mlx5e_destroy_flow_table(&ttc->ft);
+ mlx5_destroy_ttc_table(ttc);
}
-int mlx5e_create_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params,
- struct mlx5e_ttc_table *ttc)
+int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
{
- bool match_ipv_outer = MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ft_field_support.outer_ip_version);
- struct mlx5e_flow_table *ft = &ttc->ft;
+ bool match_ipv_outer =
+ MLX5_CAP_FLOWTABLE_NIC_RX(dev,
+ ft_field_support.outer_ip_version);
int err;
- ft->t = mlx5_create_flow_table(priv->fs.ns, ¶ms->ft_attr);
- if (IS_ERR(ft->t)) {
- err = PTR_ERR(ft->t);
- ft->t = NULL;
+ WARN_ON_ONCE(params->ft_attr.max_fte);
+ params->ft_attr.max_fte = MLX5_TTC_TABLE_SIZE;
+ ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
+ if (IS_ERR(ttc->t)) {
+ err = PTR_ERR(ttc->t);
+ ttc->t = NULL;
return err;
}
- err = mlx5e_create_ttc_table_groups(ttc, match_ipv_outer);
+ err = mlx5_create_ttc_table_groups(ttc, match_ipv_outer);
if (err)
- goto err;
+ goto destroy_ttc;
- err = mlx5e_generate_ttc_table_rules(priv, params, ttc);
+ err = mlx5_generate_ttc_table_rules(dev, params, ttc);
if (err)
- goto err;
+ goto destroy_ttc;
return 0;
-err:
- mlx5e_destroy_flow_table(ft);
+destroy_ttc:
+ mlx5_destroy_ttc_table(ttc);
return err;
}
-int mlx5e_ttc_fwd_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type,
- struct mlx5_flow_destination *new_dest)
+int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
+ struct mlx5_flow_destination *new_dest)
{
- return mlx5_modify_rule_destination(priv->fs.ttc.rules[type].rule, new_dest, NULL);
+ return mlx5_modify_rule_destination(ttc->rules[type].rule, new_dest,
+ NULL);
}
struct mlx5_flow_destination
-mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type)
+mlx5_ttc_get_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type)
{
- struct mlx5_flow_destination *dest = &priv->fs.ttc.rules[type].default_dest;
+ struct mlx5_flow_destination *dest = &ttc->rules[type].default_dest;
WARN_ONCE(dest->type != MLX5_FLOW_DESTINATION_TYPE_TIR,
"TTC[%d] default dest is not setup yet", type);
@@ -1433,11 +1458,12 @@ mlx5e_ttc_get_default_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type
return *dest;
}
-int mlx5e_ttc_fwd_default_dest(struct mlx5e_priv *priv, enum mlx5_traffic_types type)
+int mlx5_ttc_fwd_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type)
{
- struct mlx5_flow_destination dest = mlx5e_ttc_get_default_dest(priv, type);
+ struct mlx5_flow_destination dest = mlx5_ttc_get_default_dest(ttc, type);
- return mlx5e_ttc_fwd_dest(priv, type, &dest);
+ return mlx5_ttc_fwd_dest(ttc, type, &dest);
}
static void mlx5e_del_l2_flow_rule(struct mlx5e_priv *priv,
@@ -1470,7 +1496,7 @@ static int mlx5e_add_l2_flow_rule(struct mlx5e_priv *priv,
outer_headers.dmac_47_16);
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- dest.ft = priv->fs.ttc.ft.t;
+ dest.ft = priv->fs.ttc.t;
switch (type) {
case MLX5E_FULLMATCH:
@@ -1769,7 +1795,7 @@ static void mlx5e_destroy_vlan_table(struct mlx5e_priv *priv)
int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
{
struct ttc_params ttc_params = {};
- int tt, err;
+ int err;
priv->fs.ns = mlx5_get_flow_namespace(priv->mdev,
MLX5_FLOW_NAMESPACE_KERNEL);
@@ -1784,27 +1810,20 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
priv->netdev->hw_features &= ~NETIF_F_NTUPLE;
}
- mlx5e_set_ttc_basic_params(priv, &ttc_params);
-
if (mlx5_tunnel_inner_ft_supported(priv->mdev)) {
- mlx5e_set_inner_ttc_ft_params(&ttc_params);
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] =
- mlx5e_rx_res_get_tirn_rss_inner(priv->rx_res, tt);
-
- err = mlx5e_create_inner_ttc_table(priv, &ttc_params, &priv->fs.inner_ttc);
+ mlx5e_set_inner_ttc_params(priv, &ttc_params);
+ err = mlx5_create_inner_ttc_table(priv->mdev, &ttc_params,
+ &priv->fs.inner_ttc);
if (err) {
- netdev_err(priv->netdev, "Failed to create inner ttc table, err=%d\n",
+ netdev_err(priv->netdev,
+ "Failed to create inner ttc table, err=%d\n",
err);
goto err_destroy_arfs_tables;
}
}
- mlx5e_set_ttc_ft_params(&ttc_params);
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
-
- err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
+ mlx5e_set_ttc_params(priv, &ttc_params, true);
+ err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &priv->fs.ttc);
if (err) {
netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n",
err);
@@ -1838,10 +1857,10 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
err_destroy_l2_table:
mlx5e_destroy_l2_table(priv);
err_destroy_ttc_table:
- mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
+ mlx5_destroy_ttc_table(&priv->fs.ttc);
err_destroy_inner_ttc_table:
if (mlx5_tunnel_inner_ft_supported(priv->mdev))
- mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc);
+ mlx5_destroy_inner_ttc_table(&priv->fs.inner_ttc);
err_destroy_arfs_tables:
mlx5e_arfs_destroy_tables(priv);
@@ -1853,9 +1872,9 @@ void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv)
mlx5e_ptp_free_rx_fs(priv);
mlx5e_destroy_vlan_table(priv);
mlx5e_destroy_l2_table(priv);
- mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
+ mlx5_destroy_ttc_table(&priv->fs.ttc);
if (mlx5_tunnel_inner_ft_supported(priv->mdev))
- mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc);
+ mlx5_destroy_inner_ttc_table(&priv->fs.inner_ttc);
mlx5e_arfs_destroy_tables(priv);
mlx5e_ethtool_cleanup_steering(priv);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index f6e96b7d4698..9817a176916a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -647,25 +647,20 @@ static int mlx5e_create_rep_ttc_table(struct mlx5e_priv *priv)
{
struct mlx5e_rep_priv *rpriv = priv->ppriv;
struct mlx5_eswitch_rep *rep = rpriv->rep;
- struct mlx5e_rx_res *res = priv->rx_res;
struct ttc_params ttc_params = {};
- int tt, err;
+ int err;
priv->fs.ns = mlx5_get_flow_namespace(priv->mdev,
MLX5_FLOW_NAMESPACE_KERNEL);
/* The inner_ttc in the ttc params is intentionally not set */
- ttc_params.any_tt_tirn = mlx5e_rx_res_get_tirn_direct(res, 0);
- mlx5e_set_ttc_ft_params(&ttc_params);
+ mlx5e_set_ttc_params(priv, &ttc_params, false);
if (rep->vport != MLX5_VPORT_UPLINK)
/* To give uplik rep TTC a lower level for chaining from root ft */
ttc_params.ft_attr.level = MLX5E_TTC_FT_LEVEL + 1;
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = mlx5e_rx_res_get_tirn_rss(res, tt);
-
- err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
+ err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &priv->fs.ttc);
if (err) {
netdev_err(priv->netdev, "Failed to create rep ttc table, err=%d\n", err);
return err;
@@ -685,7 +680,7 @@ static int mlx5e_create_rep_root_ft(struct mlx5e_priv *priv)
/* non uplik reps will skip any bypass tables and go directly to
* their own ttc
*/
- rpriv->root_ft = priv->fs.ttc.ft.t;
+ rpriv->root_ft = priv->fs.ttc.t;
return 0;
}
@@ -799,7 +794,7 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv)
err_destroy_root_ft:
mlx5e_destroy_rep_root_ft(priv);
err_destroy_ttc_table:
- mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
+ mlx5_destroy_ttc_table(&priv->fs.ttc);
err_destroy_rx_res:
mlx5e_rx_res_destroy(priv->rx_res);
err_close_drop_rq:
@@ -814,7 +809,7 @@ static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv)
mlx5e_ethtool_cleanup_steering(priv);
rep_vport_rx_rule_destroy(priv);
mlx5e_destroy_rep_root_ft(priv);
- mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
+ mlx5_destroy_ttc_table(&priv->fs.ttc);
mlx5e_rx_res_destroy(priv->rx_res);
mlx5e_close_drop_rq(&priv->drop_rq);
mlx5e_rx_res_free(priv->rx_res);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 300a37c83c17..afbd0caf31ae 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -345,7 +345,7 @@ struct mlx5e_hairpin {
int num_channels;
struct mlx5e_rqt indir_rqt;
struct mlx5e_tir indir_tir[MLX5E_NUM_INDIR_TIRS];
- struct mlx5e_ttc_table ttc;
+ struct mlx5_ttc_table ttc;
};
struct mlx5e_hairpin_entry {
@@ -595,12 +595,16 @@ static void mlx5e_hairpin_set_ttc_params(struct mlx5e_hairpin *hp,
memset(ttc_params, 0, sizeof(*ttc_params));
- ttc_params->any_tt_tirn = mlx5e_tir_get_tirn(&hp->direct_tir);
-
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params->indir_tirn[tt] = mlx5e_tir_get_tirn(&hp->indir_tir[tt]);
+ ttc_params->ns = mlx5_get_flow_namespace(hp->func_mdev,
+ MLX5_FLOW_NAMESPACE_KERNEL);
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
+ ttc_params->dests[tt].type = MLX5_FLOW_DESTINATION_TYPE_TIR;
+ ttc_params->dests[tt].tir_num =
+ tt == MLX5_TT_ANY ?
+ mlx5e_tir_get_tirn(&hp->direct_tir) :
+ mlx5e_tir_get_tirn(&hp->indir_tir[tt]);
+ }
- ft_attr->max_fte = MLX5_TTC_TABLE_SIZE;
ft_attr->level = MLX5E_TC_TTC_FT_LEVEL;
ft_attr->prio = MLX5E_TC_PRIO;
}
@@ -620,12 +624,12 @@ static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp)
goto err_create_indirect_tirs;
mlx5e_hairpin_set_ttc_params(hp, &ttc_params);
- err = mlx5e_create_ttc_table(priv, &ttc_params, &hp->ttc);
+ err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &hp->ttc);
if (err)
goto err_create_ttc_table;
netdev_dbg(priv->netdev, "add hairpin: using %d channels rss ttc table id %x\n",
- hp->num_channels, hp->ttc.ft.t->id);
+ hp->num_channels, hp->ttc.t->id);
return 0;
@@ -639,9 +643,7 @@ static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp)
static void mlx5e_hairpin_rss_cleanup(struct mlx5e_hairpin *hp)
{
- struct mlx5e_priv *priv = hp->func_priv;
-
- mlx5e_destroy_ttc_table(priv, &hp->ttc);
+ mlx5_destroy_ttc_table(&hp->ttc);
mlx5e_hairpin_destroy_indirect_tirs(hp);
mlx5e_rqt_destroy(&hp->indir_rqt);
}
@@ -885,7 +887,7 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv,
attach_flow:
if (hpe->hp->num_channels > 1) {
flow_flag_set(flow, HAIRPIN_RSS);
- flow->attr->nic_attr->hairpin_ft = hpe->hp->ttc.ft.t;
+ flow->attr->nic_attr->hairpin_ft = hpe->hp->ttc.t;
} else {
flow->attr->nic_attr->hairpin_tirn = mlx5e_tir_get_tirn(&hpe->hp->direct_tir);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
index 1f118678ea9d..e04b758f20e3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
@@ -315,7 +315,7 @@ static void mlx5i_cleanup_tx(struct mlx5e_priv *priv)
static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
{
struct ttc_params ttc_params = {};
- int tt, err;
+ int err;
priv->fs.ns = mlx5_get_flow_namespace(priv->mdev,
MLX5_FLOW_NAMESPACE_KERNEL);
@@ -330,12 +330,8 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
priv->netdev->hw_features &= ~NETIF_F_NTUPLE;
}
- mlx5e_set_ttc_basic_params(priv, &ttc_params);
- mlx5e_set_ttc_ft_params(&ttc_params);
- for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
- ttc_params.indir_tirn[tt] = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt);
-
- err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
+ mlx5e_set_ttc_params(priv, &ttc_params, true);
+ err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &priv->fs.ttc);
if (err) {
netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n",
err);
@@ -352,7 +348,7 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv)
{
- mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
+ mlx5_destroy_ttc_table(&priv->fs.ttc);
mlx5e_arfs_destroy_tables(priv);
}
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 08/16] net/mlx5: Move TTC logic to fs_ttc
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (6 preceding siblings ...)
2021-08-03 2:28 ` [net-next 07/16] net/mlx5e: Decouple TTC logic from mlx5e Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 09/16] net/mlx5: Embed mlx5_ttc_table Saeed Mahameed
` (7 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maor Gottlieb, Tariq Toukan, Mark Bloch, Saeed Mahameed
From: Maor Gottlieb <maorg@nvidia.com>
Now that TTC logic is not dependent on mlx5e structs, move it to
lib/fs_ttc.c so it could be used other part of the mlx5 driver.
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/en.h | 2 -
.../net/ethernet/mellanox/mlx5/core/en/fs.h | 78 +--
.../net/ethernet/mellanox/mlx5/core/en_fs.c | 558 -----------------
.../ethernet/mellanox/mlx5/core/lib/fs_ttc.c | 584 ++++++++++++++++++
.../ethernet/mellanox/mlx5/core/lib/fs_ttc.h | 77 +++
include/linux/mlx5/fs.h | 2 +
7 files changed, 665 insertions(+), 638 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
index e8522ccb3519..33e550d77fa6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
@@ -15,7 +15,7 @@ mlx5_core-y := main.o cmd.o debugfs.o fw.o eq.o uar.o pagealloc.o \
health.o mcg.o cq.o alloc.o port.o mr.o pd.o \
transobj.o vport.o sriov.o fs_cmd.o fs_core.o pci_irq.o \
fs_counters.o fs_ft_pool.o rl.o lag.o dev.o events.o wq.o lib/gid.o \
- lib/devcom.o lib/pci_vsc.o lib/dm.o diag/fs_tracepoint.o \
+ lib/devcom.o lib/pci_vsc.o lib/dm.o lib/fs_ttc.o diag/fs_tracepoint.o \
diag/fw_tracer.o diag/crdump.o devlink.o diag/rsc_dump.o \
fw_reset.o qos.o
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 594b7971caf9..4f6897c1ea8d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -66,8 +66,6 @@ struct page_pool;
#define MLX5E_METADATA_ETHER_TYPE (0x8CE4)
#define MLX5E_METADATA_ETHER_LEN 8
-#define MLX5_SET_CFG(p, f, v) MLX5_SET(create_flow_group_in, p, f, v)
-
#define MLX5E_ETH_HARD_MTU (ETH_HLEN + VLAN_HLEN + ETH_FCS_LEN)
#define MLX5E_HW2SW_MTU(params, hwmtu) ((hwmtu) - ((params)->hard_mtu))
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
index c289f7004e10..8e7794c3d330 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
@@ -5,6 +5,7 @@
#define __MLX5E_FLOW_STEER_H__
#include "mod_hdr.h"
+#include "lib/fs_ttc.h"
enum {
MLX5E_TC_FT_LEVEL = 0,
@@ -67,21 +68,6 @@ struct mlx5e_l2_table {
bool promisc_enabled;
};
-enum mlx5_traffic_types {
- MLX5_TT_IPV4_TCP,
- MLX5_TT_IPV6_TCP,
- MLX5_TT_IPV4_UDP,
- MLX5_TT_IPV6_UDP,
- MLX5_TT_IPV4_IPSEC_AH,
- MLX5_TT_IPV6_IPSEC_AH,
- MLX5_TT_IPV4_IPSEC_ESP,
- MLX5_TT_IPV6_IPSEC_ESP,
- MLX5_TT_IPV4,
- MLX5_TT_IPV6,
- MLX5_TT_ANY,
- MLX5_NUM_TT,
-};
-
#define MLX5E_NUM_INDIR_TIRS (MLX5_NUM_TT - 1)
#define MLX5_HASH_IP (MLX5_HASH_FIELD_SEL_SRC_IP |\
@@ -94,32 +80,6 @@ enum mlx5_traffic_types {
MLX5_HASH_FIELD_SEL_DST_IP |\
MLX5_HASH_FIELD_SEL_IPSEC_SPI)
-enum mlx5_tunnel_types {
- MLX5_TT_IPV4_GRE,
- MLX5_TT_IPV6_GRE,
- MLX5_TT_IPV4_IPIP,
- MLX5_TT_IPV6_IPIP,
- MLX5_TT_IPV4_IPV6,
- MLX5_TT_IPV6_IPV6,
- MLX5_NUM_TUNNEL_TT,
-};
-
-bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev);
-
-struct mlx5_ttc_rule {
- struct mlx5_flow_handle *rule;
- struct mlx5_flow_destination default_dest;
-};
-
-/* L3/L4 traffic type classifier */
-struct mlx5_ttc_table {
- int num_groups;
- struct mlx5_flow_table *t;
- struct mlx5_flow_group **g;
- struct mlx5_ttc_rule rules[MLX5_NUM_TT];
- struct mlx5_flow_handle *tunnel_rules[MLX5_NUM_TUNNEL_TT];
-};
-
/* NIC prio FTS */
enum {
MLX5E_PROMISC_FT_LEVEL,
@@ -141,22 +101,6 @@ enum {
#endif
};
-#define MLX5_TTC_NUM_GROUPS 3
-#define MLX5_TTC_GROUP1_SIZE (BIT(3) + MLX5_NUM_TUNNEL_TT)
-#define MLX5_TTC_GROUP2_SIZE BIT(1)
-#define MLX5_TTC_GROUP3_SIZE BIT(0)
-#define MLX5_TTC_TABLE_SIZE (MLX5_TTC_GROUP1_SIZE +\
- MLX5_TTC_GROUP2_SIZE +\
- MLX5_TTC_GROUP3_SIZE)
-
-#define MLX5_INNER_TTC_NUM_GROUPS 3
-#define MLX5_INNER_TTC_GROUP1_SIZE BIT(3)
-#define MLX5_INNER_TTC_GROUP2_SIZE BIT(1)
-#define MLX5_INNER_TTC_GROUP3_SIZE BIT(0)
-#define MLX5_INNER_TTC_TABLE_SIZE (MLX5_INNER_TTC_GROUP1_SIZE +\
- MLX5_INNER_TTC_GROUP2_SIZE +\
- MLX5_INNER_TTC_GROUP3_SIZE)
-
struct mlx5e_priv;
#ifdef CONFIG_MLX5_EN_RXNFC
@@ -238,29 +182,10 @@ struct mlx5e_flow_steering {
struct mlx5e_ptp_fs *ptp_fs;
};
-struct ttc_params {
- struct mlx5_flow_namespace *ns;
- struct mlx5_flow_table_attr ft_attr;
- struct mlx5_flow_destination dests[MLX5_NUM_TT];
- bool inner_ttc;
- struct mlx5_flow_destination tunnel_dests[MLX5_NUM_TUNNEL_TT];
-};
-
void mlx5e_set_ttc_params(struct mlx5e_priv *priv,
struct ttc_params *ttc_params, bool tunnel);
-int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
- struct mlx5_ttc_table *ttc);
-void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc);
-
void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft);
-int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
- struct mlx5_flow_destination *new_dest);
-struct mlx5_flow_destination
-mlx5_ttc_get_default_dest(struct mlx5_ttc_table *ttc,
- enum mlx5_traffic_types type);
-int mlx5_ttc_fwd_default_dest(struct mlx5_ttc_table *ttc,
- enum mlx5_traffic_types type);
void mlx5e_enable_cvlan_filter(struct mlx5e_priv *priv);
void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv);
@@ -268,7 +193,6 @@ void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv);
int mlx5e_create_flow_steering(struct mlx5e_priv *priv);
void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv);
-u8 mlx5_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt);
int mlx5e_add_vlan_trap(struct mlx5e_priv *priv, int trap_id, int tir_num);
void mlx5e_remove_vlan_trap(struct mlx5e_priv *priv);
int mlx5e_add_mac_trap(struct mlx5e_priv *priv, int trap_id, int tir_num);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index a03842d132f6..cbad05760551 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -854,454 +854,6 @@ void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft)
ft->t = NULL;
}
-static void mlx5_cleanup_ttc_rules(struct mlx5_ttc_table *ttc)
-{
- int i;
-
- for (i = 0; i < MLX5_NUM_TT; i++) {
- if (!IS_ERR_OR_NULL(ttc->rules[i].rule)) {
- mlx5_del_flow_rules(ttc->rules[i].rule);
- ttc->rules[i].rule = NULL;
- }
- }
-
- for (i = 0; i < MLX5_NUM_TUNNEL_TT; i++) {
- if (!IS_ERR_OR_NULL(ttc->tunnel_rules[i])) {
- mlx5_del_flow_rules(ttc->tunnel_rules[i]);
- ttc->tunnel_rules[i] = NULL;
- }
- }
-}
-
-struct mlx5_etype_proto {
- u16 etype;
- u8 proto;
-};
-
-static struct mlx5_etype_proto ttc_rules[] = {
- [MLX5_TT_IPV4_TCP] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_TCP,
- },
- [MLX5_TT_IPV6_TCP] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_TCP,
- },
- [MLX5_TT_IPV4_UDP] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_UDP,
- },
- [MLX5_TT_IPV6_UDP] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_UDP,
- },
- [MLX5_TT_IPV4_IPSEC_AH] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_AH,
- },
- [MLX5_TT_IPV6_IPSEC_AH] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_AH,
- },
- [MLX5_TT_IPV4_IPSEC_ESP] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_ESP,
- },
- [MLX5_TT_IPV6_IPSEC_ESP] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_ESP,
- },
- [MLX5_TT_IPV4] = {
- .etype = ETH_P_IP,
- .proto = 0,
- },
- [MLX5_TT_IPV6] = {
- .etype = ETH_P_IPV6,
- .proto = 0,
- },
- [MLX5_TT_ANY] = {
- .etype = 0,
- .proto = 0,
- },
-};
-
-static struct mlx5_etype_proto ttc_tunnel_rules[] = {
- [MLX5_TT_IPV4_GRE] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_GRE,
- },
- [MLX5_TT_IPV6_GRE] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_GRE,
- },
- [MLX5_TT_IPV4_IPIP] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_IPIP,
- },
- [MLX5_TT_IPV6_IPIP] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_IPIP,
- },
- [MLX5_TT_IPV4_IPV6] = {
- .etype = ETH_P_IP,
- .proto = IPPROTO_IPV6,
- },
- [MLX5_TT_IPV6_IPV6] = {
- .etype = ETH_P_IPV6,
- .proto = IPPROTO_IPV6,
- },
-
-};
-
-u8 mlx5_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt)
-{
- return ttc_tunnel_rules[tt].proto;
-}
-
-static bool mlx5_tunnel_proto_supported_rx(struct mlx5_core_dev *mdev,
- u8 proto_type)
-{
- switch (proto_type) {
- case IPPROTO_GRE:
- return MLX5_CAP_ETH(mdev, tunnel_stateless_gre);
- case IPPROTO_IPIP:
- case IPPROTO_IPV6:
- return (MLX5_CAP_ETH(mdev, tunnel_stateless_ip_over_ip) ||
- MLX5_CAP_ETH(mdev, tunnel_stateless_ip_over_ip_rx));
- default:
- return false;
- }
-}
-
-static bool mlx5_tunnel_any_rx_proto_supported(struct mlx5_core_dev *mdev)
-{
- int tt;
-
- for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
- if (mlx5_tunnel_proto_supported_rx(mdev,
- ttc_tunnel_rules[tt].proto))
- return true;
- }
- return false;
-}
-
-bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev)
-{
- return (mlx5_tunnel_any_rx_proto_supported(mdev) &&
- MLX5_CAP_FLOWTABLE_NIC_RX(mdev,
- ft_field_support.inner_ip_version));
-}
-
-static u8 mlx5_etype_to_ipv(u16 ethertype)
-{
- if (ethertype == ETH_P_IP)
- return 4;
-
- if (ethertype == ETH_P_IPV6)
- return 6;
-
- return 0;
-}
-
-static struct mlx5_flow_handle *
-mlx5_generate_ttc_rule(struct mlx5_core_dev *dev, struct mlx5_flow_table *ft,
- struct mlx5_flow_destination *dest, u16 etype, u8 proto)
-{
- int match_ipv_outer =
- MLX5_CAP_FLOWTABLE_NIC_RX(dev,
- ft_field_support.outer_ip_version);
- MLX5_DECLARE_FLOW_ACT(flow_act);
- struct mlx5_flow_handle *rule;
- struct mlx5_flow_spec *spec;
- int err = 0;
- u8 ipv;
-
- spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
- if (!spec)
- return ERR_PTR(-ENOMEM);
-
- if (proto) {
- spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
- MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ip_protocol);
- MLX5_SET(fte_match_param, spec->match_value, outer_headers.ip_protocol, proto);
- }
-
- ipv = mlx5_etype_to_ipv(etype);
- if (match_ipv_outer && ipv) {
- spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
- MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ip_version);
- MLX5_SET(fte_match_param, spec->match_value, outer_headers.ip_version, ipv);
- } else if (etype) {
- spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
- MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ethertype);
- MLX5_SET(fte_match_param, spec->match_value, outer_headers.ethertype, etype);
- }
-
- rule = mlx5_add_flow_rules(ft, spec, &flow_act, dest, 1);
- if (IS_ERR(rule)) {
- err = PTR_ERR(rule);
- mlx5_core_err(dev, "%s: add rule failed\n", __func__);
- }
-
- kvfree(spec);
- return err ? ERR_PTR(err) : rule;
-}
-
-static int mlx5_generate_ttc_table_rules(struct mlx5_core_dev *dev,
- struct ttc_params *params,
- struct mlx5_ttc_table *ttc)
-{
- struct mlx5_flow_handle **trules;
- struct mlx5_ttc_rule *rules;
- struct mlx5_flow_table *ft;
- int tt;
- int err;
-
- ft = ttc->t;
- rules = ttc->rules;
- for (tt = 0; tt < MLX5_NUM_TT; tt++) {
- struct mlx5_ttc_rule *rule = &rules[tt];
-
- rule->rule = mlx5_generate_ttc_rule(dev, ft, ¶ms->dests[tt],
- ttc_rules[tt].etype,
- ttc_rules[tt].proto);
- if (IS_ERR(rule->rule)) {
- err = PTR_ERR(rule->rule);
- rule->rule = NULL;
- goto del_rules;
- }
- rule->default_dest = params->dests[tt];
- }
-
- if (!params->inner_ttc || !mlx5_tunnel_inner_ft_supported(dev))
- return 0;
-
- trules = ttc->tunnel_rules;
- for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
- if (!mlx5_tunnel_proto_supported_rx(dev,
- ttc_tunnel_rules[tt].proto))
- continue;
- trules[tt] = mlx5_generate_ttc_rule(dev, ft,
- ¶ms->tunnel_dests[tt],
- ttc_tunnel_rules[tt].etype,
- ttc_tunnel_rules[tt].proto);
- if (IS_ERR(trules[tt])) {
- err = PTR_ERR(trules[tt]);
- trules[tt] = NULL;
- goto del_rules;
- }
- }
-
- return 0;
-
-del_rules:
- mlx5_cleanup_ttc_rules(ttc);
- return err;
-}
-
-static int mlx5_create_ttc_table_groups(struct mlx5_ttc_table *ttc,
- bool use_ipv)
-{
- int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
- int ix = 0;
- u32 *in;
- int err;
- u8 *mc;
-
- ttc->g = kcalloc(MLX5_TTC_NUM_GROUPS, sizeof(*ttc->g), GFP_KERNEL);
- if (!ttc->g)
- return -ENOMEM;
- in = kvzalloc(inlen, GFP_KERNEL);
- if (!in) {
- kfree(ttc->g);
- ttc->g = NULL;
- return -ENOMEM;
- }
-
- /* L4 Group */
- mc = MLX5_ADDR_OF(create_flow_group_in, in, match_criteria);
- MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ip_protocol);
- if (use_ipv)
- MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ip_version);
- else
- MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ethertype);
- MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS);
- MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5_TTC_GROUP1_SIZE;
- MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
- if (IS_ERR(ttc->g[ttc->num_groups]))
- goto err;
- ttc->num_groups++;
-
- /* L3 Group */
- MLX5_SET(fte_match_param, mc, outer_headers.ip_protocol, 0);
- MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5_TTC_GROUP2_SIZE;
- MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
- if (IS_ERR(ttc->g[ttc->num_groups]))
- goto err;
- ttc->num_groups++;
-
- /* Any Group */
- memset(in, 0, inlen);
- MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5_TTC_GROUP3_SIZE;
- MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
- if (IS_ERR(ttc->g[ttc->num_groups]))
- goto err;
- ttc->num_groups++;
-
- kvfree(in);
- return 0;
-
-err:
- err = PTR_ERR(ttc->g[ttc->num_groups]);
- ttc->g[ttc->num_groups] = NULL;
- kvfree(in);
-
- return err;
-}
-
-static struct mlx5_flow_handle *
-mlx5_generate_inner_ttc_rule(struct mlx5_core_dev *dev,
- struct mlx5_flow_table *ft,
- struct mlx5_flow_destination *dest,
- u16 etype, u8 proto)
-{
- MLX5_DECLARE_FLOW_ACT(flow_act);
- struct mlx5_flow_handle *rule;
- struct mlx5_flow_spec *spec;
- int err = 0;
- u8 ipv;
-
- spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
- if (!spec)
- return ERR_PTR(-ENOMEM);
-
- ipv = mlx5_etype_to_ipv(etype);
- if (etype && ipv) {
- spec->match_criteria_enable = MLX5_MATCH_INNER_HEADERS;
- MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, inner_headers.ip_version);
- MLX5_SET(fte_match_param, spec->match_value, inner_headers.ip_version, ipv);
- }
-
- if (proto) {
- spec->match_criteria_enable = MLX5_MATCH_INNER_HEADERS;
- MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, inner_headers.ip_protocol);
- MLX5_SET(fte_match_param, spec->match_value, inner_headers.ip_protocol, proto);
- }
-
- rule = mlx5_add_flow_rules(ft, spec, &flow_act, dest, 1);
- if (IS_ERR(rule)) {
- err = PTR_ERR(rule);
- mlx5_core_err(dev, "%s: add inner TTC rule failed\n", __func__);
- }
-
- kvfree(spec);
- return err ? ERR_PTR(err) : rule;
-}
-
-static int mlx5_generate_inner_ttc_table_rules(struct mlx5_core_dev *dev,
- struct ttc_params *params,
- struct mlx5_ttc_table *ttc)
-{
- struct mlx5_ttc_rule *rules;
- struct mlx5_flow_table *ft;
- int err;
- int tt;
-
- ft = ttc->t;
- rules = ttc->rules;
-
- for (tt = 0; tt < MLX5_NUM_TT; tt++) {
- struct mlx5_ttc_rule *rule = &rules[tt];
-
- rule->rule = mlx5_generate_inner_ttc_rule(dev, ft,
- ¶ms->dests[tt],
- ttc_rules[tt].etype,
- ttc_rules[tt].proto);
- if (IS_ERR(rule->rule)) {
- err = PTR_ERR(rule->rule);
- rule->rule = NULL;
- goto del_rules;
- }
- rule->default_dest = params->dests[tt];
- }
-
- return 0;
-
-del_rules:
-
- mlx5_cleanup_ttc_rules(ttc);
- return err;
-}
-
-static int mlx5_create_inner_ttc_table_groups(struct mlx5_ttc_table *ttc)
-{
- int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
- int ix = 0;
- u32 *in;
- int err;
- u8 *mc;
-
- ttc->g = kcalloc(MLX5_INNER_TTC_NUM_GROUPS, sizeof(*ttc->g),
- GFP_KERNEL);
- if (!ttc->g)
- return -ENOMEM;
- in = kvzalloc(inlen, GFP_KERNEL);
- if (!in) {
- kfree(ttc->g);
- ttc->g = NULL;
- return -ENOMEM;
- }
-
- /* L4 Group */
- mc = MLX5_ADDR_OF(create_flow_group_in, in, match_criteria);
- MLX5_SET_TO_ONES(fte_match_param, mc, inner_headers.ip_protocol);
- MLX5_SET_TO_ONES(fte_match_param, mc, inner_headers.ip_version);
- MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_INNER_HEADERS);
- MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5_INNER_TTC_GROUP1_SIZE;
- MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
- if (IS_ERR(ttc->g[ttc->num_groups]))
- goto err;
- ttc->num_groups++;
-
- /* L3 Group */
- MLX5_SET(fte_match_param, mc, inner_headers.ip_protocol, 0);
- MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5_INNER_TTC_GROUP2_SIZE;
- MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
- if (IS_ERR(ttc->g[ttc->num_groups]))
- goto err;
- ttc->num_groups++;
-
- /* Any Group */
- memset(in, 0, inlen);
- MLX5_SET_CFG(in, start_flow_index, ix);
- ix += MLX5_INNER_TTC_GROUP3_SIZE;
- MLX5_SET_CFG(in, end_flow_index, ix - 1);
- ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
- if (IS_ERR(ttc->g[ttc->num_groups]))
- goto err;
- ttc->num_groups++;
-
- kvfree(in);
- return 0;
-
-err:
- err = PTR_ERR(ttc->g[ttc->num_groups]);
- ttc->g[ttc->num_groups] = NULL;
- kvfree(in);
-
- return err;
-}
-
static void mlx5e_set_inner_ttc_params(struct mlx5e_priv *priv,
struct ttc_params *ttc_params)
{
@@ -1356,116 +908,6 @@ void mlx5e_set_ttc_params(struct mlx5e_priv *priv,
}
}
-static int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
- struct ttc_params *params,
- struct mlx5_ttc_table *ttc)
-{
- int err;
-
- WARN_ON_ONCE(params->ft_attr.max_fte);
- params->ft_attr.max_fte = MLX5_INNER_TTC_TABLE_SIZE;
- ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
- if (IS_ERR(ttc->t)) {
- err = PTR_ERR(ttc->t);
- ttc->t = NULL;
- return err;
- }
-
- err = mlx5_create_inner_ttc_table_groups(ttc);
- if (err)
- goto destroy_ttc;
-
- err = mlx5_generate_inner_ttc_table_rules(dev, params, ttc);
- if (err)
- goto destroy_ttc;
-
- return 0;
-
-destroy_ttc:
- mlx5_destroy_ttc_table(ttc);
- return err;
-}
-
-void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc)
-{
- int i;
-
- mlx5_cleanup_ttc_rules(ttc);
- for (i = ttc->num_groups - 1; i >= 0; i--) {
- if (!IS_ERR_OR_NULL(ttc->g[i]))
- mlx5_destroy_flow_group(ttc->g[i]);
- ttc->g[i] = NULL;
- }
-
- ttc->num_groups = 0;
- kfree(ttc->g);
- mlx5_destroy_flow_table(ttc->t);
- ttc->t = NULL;
-}
-
-static void mlx5_destroy_inner_ttc_table(struct mlx5_ttc_table *ttc)
-{
- mlx5_destroy_ttc_table(ttc);
-}
-
-int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
- struct mlx5_ttc_table *ttc)
-{
- bool match_ipv_outer =
- MLX5_CAP_FLOWTABLE_NIC_RX(dev,
- ft_field_support.outer_ip_version);
- int err;
-
- WARN_ON_ONCE(params->ft_attr.max_fte);
- params->ft_attr.max_fte = MLX5_TTC_TABLE_SIZE;
- ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
- if (IS_ERR(ttc->t)) {
- err = PTR_ERR(ttc->t);
- ttc->t = NULL;
- return err;
- }
-
- err = mlx5_create_ttc_table_groups(ttc, match_ipv_outer);
- if (err)
- goto destroy_ttc;
-
- err = mlx5_generate_ttc_table_rules(dev, params, ttc);
- if (err)
- goto destroy_ttc;
-
- return 0;
-destroy_ttc:
- mlx5_destroy_ttc_table(ttc);
- return err;
-}
-
-int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
- struct mlx5_flow_destination *new_dest)
-{
- return mlx5_modify_rule_destination(ttc->rules[type].rule, new_dest,
- NULL);
-}
-
-struct mlx5_flow_destination
-mlx5_ttc_get_default_dest(struct mlx5_ttc_table *ttc,
- enum mlx5_traffic_types type)
-{
- struct mlx5_flow_destination *dest = &ttc->rules[type].default_dest;
-
- WARN_ONCE(dest->type != MLX5_FLOW_DESTINATION_TYPE_TIR,
- "TTC[%d] default dest is not setup yet", type);
-
- return *dest;
-}
-
-int mlx5_ttc_fwd_default_dest(struct mlx5_ttc_table *ttc,
- enum mlx5_traffic_types type)
-{
- struct mlx5_flow_destination dest = mlx5_ttc_get_default_dest(ttc, type);
-
- return mlx5_ttc_fwd_dest(ttc, type, &dest);
-}
-
static void mlx5e_del_l2_flow_rule(struct mlx5e_priv *priv,
struct mlx5e_l2_rule *ai)
{
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
new file mode 100644
index 000000000000..4b54b4127d33
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
@@ -0,0 +1,584 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+// Copyright (c) 2021 NVIDIA CORPORATION & AFFILIATES.
+
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <linux/tcp.h>
+#include <linux/mlx5/fs.h>
+#include <linux/mlx5/driver.h>
+#include "mlx5_core.h"
+#include "lib/fs_ttc.h"
+
+#define MLX5_TTC_NUM_GROUPS 3
+#define MLX5_TTC_GROUP1_SIZE (BIT(3) + MLX5_NUM_TUNNEL_TT)
+#define MLX5_TTC_GROUP2_SIZE BIT(1)
+#define MLX5_TTC_GROUP3_SIZE BIT(0)
+#define MLX5_TTC_TABLE_SIZE (MLX5_TTC_GROUP1_SIZE +\
+ MLX5_TTC_GROUP2_SIZE +\
+ MLX5_TTC_GROUP3_SIZE)
+
+#define MLX5_INNER_TTC_NUM_GROUPS 3
+#define MLX5_INNER_TTC_GROUP1_SIZE BIT(3)
+#define MLX5_INNER_TTC_GROUP2_SIZE BIT(1)
+#define MLX5_INNER_TTC_GROUP3_SIZE BIT(0)
+#define MLX5_INNER_TTC_TABLE_SIZE (MLX5_INNER_TTC_GROUP1_SIZE +\
+ MLX5_INNER_TTC_GROUP2_SIZE +\
+ MLX5_INNER_TTC_GROUP3_SIZE)
+
+static void mlx5_cleanup_ttc_rules(struct mlx5_ttc_table *ttc)
+{
+ int i;
+
+ for (i = 0; i < MLX5_NUM_TT; i++) {
+ if (!IS_ERR_OR_NULL(ttc->rules[i].rule)) {
+ mlx5_del_flow_rules(ttc->rules[i].rule);
+ ttc->rules[i].rule = NULL;
+ }
+ }
+
+ for (i = 0; i < MLX5_NUM_TUNNEL_TT; i++) {
+ if (!IS_ERR_OR_NULL(ttc->tunnel_rules[i])) {
+ mlx5_del_flow_rules(ttc->tunnel_rules[i]);
+ ttc->tunnel_rules[i] = NULL;
+ }
+ }
+}
+
+struct mlx5_etype_proto {
+ u16 etype;
+ u8 proto;
+};
+
+static struct mlx5_etype_proto ttc_rules[] = {
+ [MLX5_TT_IPV4_TCP] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_TCP,
+ },
+ [MLX5_TT_IPV6_TCP] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_TCP,
+ },
+ [MLX5_TT_IPV4_UDP] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_UDP,
+ },
+ [MLX5_TT_IPV6_UDP] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_UDP,
+ },
+ [MLX5_TT_IPV4_IPSEC_AH] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_AH,
+ },
+ [MLX5_TT_IPV6_IPSEC_AH] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_AH,
+ },
+ [MLX5_TT_IPV4_IPSEC_ESP] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_ESP,
+ },
+ [MLX5_TT_IPV6_IPSEC_ESP] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_ESP,
+ },
+ [MLX5_TT_IPV4] = {
+ .etype = ETH_P_IP,
+ .proto = 0,
+ },
+ [MLX5_TT_IPV6] = {
+ .etype = ETH_P_IPV6,
+ .proto = 0,
+ },
+ [MLX5_TT_ANY] = {
+ .etype = 0,
+ .proto = 0,
+ },
+};
+
+static struct mlx5_etype_proto ttc_tunnel_rules[] = {
+ [MLX5_TT_IPV4_GRE] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_GRE,
+ },
+ [MLX5_TT_IPV6_GRE] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_GRE,
+ },
+ [MLX5_TT_IPV4_IPIP] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_IPIP,
+ },
+ [MLX5_TT_IPV6_IPIP] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_IPIP,
+ },
+ [MLX5_TT_IPV4_IPV6] = {
+ .etype = ETH_P_IP,
+ .proto = IPPROTO_IPV6,
+ },
+ [MLX5_TT_IPV6_IPV6] = {
+ .etype = ETH_P_IPV6,
+ .proto = IPPROTO_IPV6,
+ },
+
+};
+
+u8 mlx5_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt)
+{
+ return ttc_tunnel_rules[tt].proto;
+}
+
+static bool mlx5_tunnel_proto_supported_rx(struct mlx5_core_dev *mdev,
+ u8 proto_type)
+{
+ switch (proto_type) {
+ case IPPROTO_GRE:
+ return MLX5_CAP_ETH(mdev, tunnel_stateless_gre);
+ case IPPROTO_IPIP:
+ case IPPROTO_IPV6:
+ return (MLX5_CAP_ETH(mdev, tunnel_stateless_ip_over_ip) ||
+ MLX5_CAP_ETH(mdev, tunnel_stateless_ip_over_ip_rx));
+ default:
+ return false;
+ }
+}
+
+static bool mlx5_tunnel_any_rx_proto_supported(struct mlx5_core_dev *mdev)
+{
+ int tt;
+
+ for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
+ if (mlx5_tunnel_proto_supported_rx(mdev,
+ ttc_tunnel_rules[tt].proto))
+ return true;
+ }
+ return false;
+}
+
+bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev)
+{
+ return (mlx5_tunnel_any_rx_proto_supported(mdev) &&
+ MLX5_CAP_FLOWTABLE_NIC_RX(mdev,
+ ft_field_support.inner_ip_version));
+}
+
+static u8 mlx5_etype_to_ipv(u16 ethertype)
+{
+ if (ethertype == ETH_P_IP)
+ return 4;
+
+ if (ethertype == ETH_P_IPV6)
+ return 6;
+
+ return 0;
+}
+
+static struct mlx5_flow_handle *
+mlx5_generate_ttc_rule(struct mlx5_core_dev *dev, struct mlx5_flow_table *ft,
+ struct mlx5_flow_destination *dest, u16 etype, u8 proto)
+{
+ int match_ipv_outer =
+ MLX5_CAP_FLOWTABLE_NIC_RX(dev,
+ ft_field_support.outer_ip_version);
+ MLX5_DECLARE_FLOW_ACT(flow_act);
+ struct mlx5_flow_handle *rule;
+ struct mlx5_flow_spec *spec;
+ int err = 0;
+ u8 ipv;
+
+ spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
+ if (!spec)
+ return ERR_PTR(-ENOMEM);
+
+ if (proto) {
+ spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
+ MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ip_protocol);
+ MLX5_SET(fte_match_param, spec->match_value, outer_headers.ip_protocol, proto);
+ }
+
+ ipv = mlx5_etype_to_ipv(etype);
+ if (match_ipv_outer && ipv) {
+ spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
+ MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ip_version);
+ MLX5_SET(fte_match_param, spec->match_value, outer_headers.ip_version, ipv);
+ } else if (etype) {
+ spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
+ MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ethertype);
+ MLX5_SET(fte_match_param, spec->match_value, outer_headers.ethertype, etype);
+ }
+
+ rule = mlx5_add_flow_rules(ft, spec, &flow_act, dest, 1);
+ if (IS_ERR(rule)) {
+ err = PTR_ERR(rule);
+ mlx5_core_err(dev, "%s: add rule failed\n", __func__);
+ }
+
+ kvfree(spec);
+ return err ? ERR_PTR(err) : rule;
+}
+
+static int mlx5_generate_ttc_table_rules(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
+{
+ struct mlx5_flow_handle **trules;
+ struct mlx5_ttc_rule *rules;
+ struct mlx5_flow_table *ft;
+ int tt;
+ int err;
+
+ ft = ttc->t;
+ rules = ttc->rules;
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
+ struct mlx5_ttc_rule *rule = &rules[tt];
+
+ rule->rule = mlx5_generate_ttc_rule(dev, ft, ¶ms->dests[tt],
+ ttc_rules[tt].etype,
+ ttc_rules[tt].proto);
+ if (IS_ERR(rule->rule)) {
+ err = PTR_ERR(rule->rule);
+ rule->rule = NULL;
+ goto del_rules;
+ }
+ rule->default_dest = params->dests[tt];
+ }
+
+ if (!params->inner_ttc || !mlx5_tunnel_inner_ft_supported(dev))
+ return 0;
+
+ trules = ttc->tunnel_rules;
+ for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
+ if (!mlx5_tunnel_proto_supported_rx(dev,
+ ttc_tunnel_rules[tt].proto))
+ continue;
+ trules[tt] = mlx5_generate_ttc_rule(dev, ft,
+ ¶ms->tunnel_dests[tt],
+ ttc_tunnel_rules[tt].etype,
+ ttc_tunnel_rules[tt].proto);
+ if (IS_ERR(trules[tt])) {
+ err = PTR_ERR(trules[tt]);
+ trules[tt] = NULL;
+ goto del_rules;
+ }
+ }
+
+ return 0;
+
+del_rules:
+ mlx5_cleanup_ttc_rules(ttc);
+ return err;
+}
+
+static int mlx5_create_ttc_table_groups(struct mlx5_ttc_table *ttc,
+ bool use_ipv)
+{
+ int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
+ int ix = 0;
+ u32 *in;
+ int err;
+ u8 *mc;
+
+ ttc->g = kcalloc(MLX5_TTC_NUM_GROUPS, sizeof(*ttc->g), GFP_KERNEL);
+ if (!ttc->g)
+ return -ENOMEM;
+ in = kvzalloc(inlen, GFP_KERNEL);
+ if (!in) {
+ kfree(ttc->g);
+ ttc->g = NULL;
+ return -ENOMEM;
+ }
+
+ /* L4 Group */
+ mc = MLX5_ADDR_OF(create_flow_group_in, in, match_criteria);
+ MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ip_protocol);
+ if (use_ipv)
+ MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ip_version);
+ else
+ MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ethertype);
+ MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS);
+ MLX5_SET_CFG(in, start_flow_index, ix);
+ ix += MLX5_TTC_GROUP1_SIZE;
+ MLX5_SET_CFG(in, end_flow_index, ix - 1);
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
+ goto err;
+ ttc->num_groups++;
+
+ /* L3 Group */
+ MLX5_SET(fte_match_param, mc, outer_headers.ip_protocol, 0);
+ MLX5_SET_CFG(in, start_flow_index, ix);
+ ix += MLX5_TTC_GROUP2_SIZE;
+ MLX5_SET_CFG(in, end_flow_index, ix - 1);
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
+ goto err;
+ ttc->num_groups++;
+
+ /* Any Group */
+ memset(in, 0, inlen);
+ MLX5_SET_CFG(in, start_flow_index, ix);
+ ix += MLX5_TTC_GROUP3_SIZE;
+ MLX5_SET_CFG(in, end_flow_index, ix - 1);
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
+ goto err;
+ ttc->num_groups++;
+
+ kvfree(in);
+ return 0;
+
+err:
+ err = PTR_ERR(ttc->g[ttc->num_groups]);
+ ttc->g[ttc->num_groups] = NULL;
+ kvfree(in);
+
+ return err;
+}
+
+static struct mlx5_flow_handle *
+mlx5_generate_inner_ttc_rule(struct mlx5_core_dev *dev,
+ struct mlx5_flow_table *ft,
+ struct mlx5_flow_destination *dest,
+ u16 etype, u8 proto)
+{
+ MLX5_DECLARE_FLOW_ACT(flow_act);
+ struct mlx5_flow_handle *rule;
+ struct mlx5_flow_spec *spec;
+ int err = 0;
+ u8 ipv;
+
+ spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
+ if (!spec)
+ return ERR_PTR(-ENOMEM);
+
+ ipv = mlx5_etype_to_ipv(etype);
+ if (etype && ipv) {
+ spec->match_criteria_enable = MLX5_MATCH_INNER_HEADERS;
+ MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, inner_headers.ip_version);
+ MLX5_SET(fte_match_param, spec->match_value, inner_headers.ip_version, ipv);
+ }
+
+ if (proto) {
+ spec->match_criteria_enable = MLX5_MATCH_INNER_HEADERS;
+ MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, inner_headers.ip_protocol);
+ MLX5_SET(fte_match_param, spec->match_value, inner_headers.ip_protocol, proto);
+ }
+
+ rule = mlx5_add_flow_rules(ft, spec, &flow_act, dest, 1);
+ if (IS_ERR(rule)) {
+ err = PTR_ERR(rule);
+ mlx5_core_err(dev, "%s: add inner TTC rule failed\n", __func__);
+ }
+
+ kvfree(spec);
+ return err ? ERR_PTR(err) : rule;
+}
+
+static int mlx5_generate_inner_ttc_table_rules(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
+{
+ struct mlx5_ttc_rule *rules;
+ struct mlx5_flow_table *ft;
+ int err;
+ int tt;
+
+ ft = ttc->t;
+ rules = ttc->rules;
+
+ for (tt = 0; tt < MLX5_NUM_TT; tt++) {
+ struct mlx5_ttc_rule *rule = &rules[tt];
+
+ rule->rule = mlx5_generate_inner_ttc_rule(dev, ft,
+ ¶ms->dests[tt],
+ ttc_rules[tt].etype,
+ ttc_rules[tt].proto);
+ if (IS_ERR(rule->rule)) {
+ err = PTR_ERR(rule->rule);
+ rule->rule = NULL;
+ goto del_rules;
+ }
+ rule->default_dest = params->dests[tt];
+ }
+
+ return 0;
+
+del_rules:
+
+ mlx5_cleanup_ttc_rules(ttc);
+ return err;
+}
+
+static int mlx5_create_inner_ttc_table_groups(struct mlx5_ttc_table *ttc)
+{
+ int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
+ int ix = 0;
+ u32 *in;
+ int err;
+ u8 *mc;
+
+ ttc->g = kcalloc(MLX5_INNER_TTC_NUM_GROUPS, sizeof(*ttc->g),
+ GFP_KERNEL);
+ if (!ttc->g)
+ return -ENOMEM;
+ in = kvzalloc(inlen, GFP_KERNEL);
+ if (!in) {
+ kfree(ttc->g);
+ ttc->g = NULL;
+ return -ENOMEM;
+ }
+
+ /* L4 Group */
+ mc = MLX5_ADDR_OF(create_flow_group_in, in, match_criteria);
+ MLX5_SET_TO_ONES(fte_match_param, mc, inner_headers.ip_protocol);
+ MLX5_SET_TO_ONES(fte_match_param, mc, inner_headers.ip_version);
+ MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_INNER_HEADERS);
+ MLX5_SET_CFG(in, start_flow_index, ix);
+ ix += MLX5_INNER_TTC_GROUP1_SIZE;
+ MLX5_SET_CFG(in, end_flow_index, ix - 1);
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
+ goto err;
+ ttc->num_groups++;
+
+ /* L3 Group */
+ MLX5_SET(fte_match_param, mc, inner_headers.ip_protocol, 0);
+ MLX5_SET_CFG(in, start_flow_index, ix);
+ ix += MLX5_INNER_TTC_GROUP2_SIZE;
+ MLX5_SET_CFG(in, end_flow_index, ix - 1);
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
+ goto err;
+ ttc->num_groups++;
+
+ /* Any Group */
+ memset(in, 0, inlen);
+ MLX5_SET_CFG(in, start_flow_index, ix);
+ ix += MLX5_INNER_TTC_GROUP3_SIZE;
+ MLX5_SET_CFG(in, end_flow_index, ix - 1);
+ ttc->g[ttc->num_groups] = mlx5_create_flow_group(ttc->t, in);
+ if (IS_ERR(ttc->g[ttc->num_groups]))
+ goto err;
+ ttc->num_groups++;
+
+ kvfree(in);
+ return 0;
+
+err:
+ err = PTR_ERR(ttc->g[ttc->num_groups]);
+ ttc->g[ttc->num_groups] = NULL;
+ kvfree(in);
+
+ return err;
+}
+
+int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
+{
+ int err;
+
+ WARN_ON_ONCE(params->ft_attr.max_fte);
+ params->ft_attr.max_fte = MLX5_INNER_TTC_TABLE_SIZE;
+ ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
+ if (IS_ERR(ttc->t)) {
+ err = PTR_ERR(ttc->t);
+ ttc->t = NULL;
+ return err;
+ }
+
+ err = mlx5_create_inner_ttc_table_groups(ttc);
+ if (err)
+ goto destroy_ft;
+
+ err = mlx5_generate_inner_ttc_table_rules(dev, params, ttc);
+ if (err)
+ goto destroy_ft;
+
+ return 0;
+
+destroy_ft:
+ mlx5_destroy_ttc_table(ttc);
+ return err;
+}
+
+void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc)
+{
+ int i;
+
+ mlx5_cleanup_ttc_rules(ttc);
+ for (i = ttc->num_groups - 1; i >= 0; i--) {
+ if (!IS_ERR_OR_NULL(ttc->g[i]))
+ mlx5_destroy_flow_group(ttc->g[i]);
+ ttc->g[i] = NULL;
+ }
+
+ ttc->num_groups = 0;
+ kfree(ttc->g);
+ mlx5_destroy_flow_table(ttc->t);
+ ttc->t = NULL;
+}
+
+void mlx5_destroy_inner_ttc_table(struct mlx5_ttc_table *ttc)
+{
+ mlx5_destroy_ttc_table(ttc);
+}
+
+int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
+ struct mlx5_ttc_table *ttc)
+{
+ bool match_ipv_outer =
+ MLX5_CAP_FLOWTABLE_NIC_RX(dev,
+ ft_field_support.outer_ip_version);
+ int err;
+
+ WARN_ON_ONCE(params->ft_attr.max_fte);
+ params->ft_attr.max_fte = MLX5_TTC_TABLE_SIZE;
+ ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
+ if (IS_ERR(ttc->t)) {
+ err = PTR_ERR(ttc->t);
+ ttc->t = NULL;
+ return err;
+ }
+
+ err = mlx5_create_ttc_table_groups(ttc, match_ipv_outer);
+ if (err)
+ goto destroy_ft;
+
+ err = mlx5_generate_ttc_table_rules(dev, params, ttc);
+ if (err)
+ goto destroy_ft;
+
+ return 0;
+destroy_ft:
+ mlx5_destroy_ttc_table(ttc);
+ return err;
+}
+
+int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
+ struct mlx5_flow_destination *new_dest)
+{
+ return mlx5_modify_rule_destination(ttc->rules[type].rule, new_dest,
+ NULL);
+}
+
+struct mlx5_flow_destination
+mlx5_ttc_get_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type)
+{
+ struct mlx5_flow_destination *dest = &ttc->rules[type].default_dest;
+
+ WARN_ONCE(dest->type != MLX5_FLOW_DESTINATION_TYPE_TIR,
+ "TTC[%d] default dest is not setup yet", type);
+
+ return *dest;
+}
+
+int mlx5_ttc_fwd_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type)
+{
+ struct mlx5_flow_destination dest = mlx5_ttc_get_default_dest(ttc, type);
+
+ return mlx5_ttc_fwd_dest(ttc, type, &dest);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
new file mode 100644
index 000000000000..1010e00c10bd
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
@@ -0,0 +1,77 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/* Copyright (c) 2020 Mellanox Technologies. */
+
+#ifndef __ML5_FS_TTC_H__
+#define __ML5_FS_TTC_H__
+
+#include <linux/mlx5/fs.h>
+
+enum mlx5_traffic_types {
+ MLX5_TT_IPV4_TCP,
+ MLX5_TT_IPV6_TCP,
+ MLX5_TT_IPV4_UDP,
+ MLX5_TT_IPV6_UDP,
+ MLX5_TT_IPV4_IPSEC_AH,
+ MLX5_TT_IPV6_IPSEC_AH,
+ MLX5_TT_IPV4_IPSEC_ESP,
+ MLX5_TT_IPV6_IPSEC_ESP,
+ MLX5_TT_IPV4,
+ MLX5_TT_IPV6,
+ MLX5_TT_ANY,
+ MLX5_NUM_TT,
+ MLX5_NUM_INDIR_TIRS = MLX5_TT_ANY,
+};
+
+enum mlx5_tunnel_types {
+ MLX5_TT_IPV4_GRE,
+ MLX5_TT_IPV6_GRE,
+ MLX5_TT_IPV4_IPIP,
+ MLX5_TT_IPV6_IPIP,
+ MLX5_TT_IPV4_IPV6,
+ MLX5_TT_IPV6_IPV6,
+ MLX5_NUM_TUNNEL_TT,
+};
+
+struct mlx5_ttc_rule {
+ struct mlx5_flow_handle *rule;
+ struct mlx5_flow_destination default_dest;
+};
+
+/* L3/L4 traffic type classifier */
+struct mlx5_ttc_table {
+ int num_groups;
+ struct mlx5_flow_table *t;
+ struct mlx5_flow_group **g;
+ struct mlx5_ttc_rule rules[MLX5_NUM_TT];
+ struct mlx5_flow_handle *tunnel_rules[MLX5_NUM_TUNNEL_TT];
+};
+
+struct ttc_params {
+ struct mlx5_flow_namespace *ns;
+ struct mlx5_flow_table_attr ft_attr;
+ struct mlx5_flow_destination dests[MLX5_NUM_TT];
+ bool inner_ttc;
+ struct mlx5_flow_destination tunnel_dests[MLX5_NUM_TUNNEL_TT];
+};
+
+int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
+ struct mlx5_ttc_table *ttc);
+void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc);
+
+int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params,
+ struct mlx5_ttc_table *ttc);
+void mlx5_destroy_inner_ttc_table(struct mlx5_ttc_table *ttc);
+
+int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
+ struct mlx5_flow_destination *new_dest);
+struct mlx5_flow_destination
+mlx5_ttc_get_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type);
+int mlx5_ttc_fwd_default_dest(struct mlx5_ttc_table *ttc,
+ enum mlx5_traffic_types type);
+
+bool mlx5_tunnel_inner_ft_supported(struct mlx5_core_dev *mdev);
+u8 mlx5_get_proto_by_tunnel_type(enum mlx5_tunnel_types tt);
+
+#endif /* __MLX5_FS_TTC_H__ */
diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h
index 77746f7e35b8..0106c67e8ccb 100644
--- a/include/linux/mlx5/fs.h
+++ b/include/linux/mlx5/fs.h
@@ -38,6 +38,8 @@
#define MLX5_FS_DEFAULT_FLOW_TAG 0x0
+#define MLX5_SET_CFG(p, f, v) MLX5_SET(create_flow_group_in, p, f, v)
+
enum {
MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO = 1 << 16,
MLX5_FLOW_CONTEXT_ACTION_ENCRYPT = 1 << 17,
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 09/16] net/mlx5: Embed mlx5_ttc_table
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (7 preceding siblings ...)
2021-08-03 2:28 ` [net-next 08/16] net/mlx5: Move TTC logic to fs_ttc Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 10/16] net/mlx5e: Remove redundant tc act includes Saeed Mahameed
` (6 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Maor Gottlieb, Tariq Toukan, Saeed Mahameed, Mark Bloch
From: Maor Gottlieb <maorg@nvidia.com>
mlx5_ttc_table struct shouldn't be exposed to the users so
this patch make it internal to ttc.
In addition add a getter function to get the TTC flow table for users
that need to add a rule which points on it.
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/en/fs.h | 7 +-
.../mellanox/mlx5/core/en/fs_tt_redirect.c | 13 ++--
.../mellanox/mlx5/core/en_accel/fs_tcp.c | 6 +-
.../mellanox/mlx5/core/en_accel/ipsec_fs.c | 7 +-
.../net/ethernet/mellanox/mlx5/core/en_arfs.c | 4 +-
.../net/ethernet/mellanox/mlx5/core/en_fs.c | 74 +++++++++++++------
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 14 ++--
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 16 ++--
.../ethernet/mellanox/mlx5/core/ipoib/ipoib.c | 6 +-
.../ethernet/mellanox/mlx5/core/lib/fs_ttc.c | 58 ++++++++++-----
.../ethernet/mellanox/mlx5/core/lib/fs_ttc.h | 21 ++----
11 files changed, 137 insertions(+), 89 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
index 8e7794c3d330..e348c276eaa1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h
@@ -169,8 +169,8 @@ struct mlx5e_flow_steering {
struct mlx5e_promisc_table promisc;
struct mlx5e_vlan_table *vlan;
struct mlx5e_l2_table l2;
- struct mlx5_ttc_table ttc;
- struct mlx5_ttc_table inner_ttc;
+ struct mlx5_ttc_table *ttc;
+ struct mlx5_ttc_table *inner_ttc;
#ifdef CONFIG_MLX5_EN_ARFS
struct mlx5e_arfs_tables *arfs;
#endif
@@ -185,6 +185,9 @@ struct mlx5e_flow_steering {
void mlx5e_set_ttc_params(struct mlx5e_priv *priv,
struct ttc_params *ttc_params, bool tunnel);
+void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv);
+int mlx5e_create_ttc_table(struct mlx5e_priv *priv);
+
void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft);
void mlx5e_enable_cvlan_filter(struct mlx5e_priv *priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
index 68cc3a8fd6b7..7aa25a5e29d7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c
@@ -124,7 +124,7 @@ static int fs_udp_add_default_rule(struct mlx5e_priv *priv, enum fs_udp_type typ
fs_udp = priv->fs.udp;
fs_udp_t = &fs_udp->tables[type];
- dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, fs_udp2tt(type));
+ dest = mlx5_ttc_get_default_dest(priv->fs.ttc, fs_udp2tt(type));
rule = mlx5_add_flow_rules(fs_udp_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -259,7 +259,7 @@ static int fs_udp_disable(struct mlx5e_priv *priv)
for (i = 0; i < FS_UDP_NUM_TYPES; i++) {
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, fs_udp2tt(i));
+ err = mlx5_ttc_fwd_default_dest(priv->fs.ttc, fs_udp2tt(i));
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -281,8 +281,7 @@ static int fs_udp_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.udp->tables[i].t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5_ttc_fwd_dest(&priv->fs.ttc, fs_udp2tt(i),
- &dest);
+ err = mlx5_ttc_fwd_dest(priv->fs.ttc, fs_udp2tt(i), &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
@@ -402,7 +401,7 @@ static int fs_any_add_default_rule(struct mlx5e_priv *priv)
fs_any = priv->fs.any;
fs_any_t = &fs_any->table;
- dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, MLX5_TT_ANY);
+ dest = mlx5_ttc_get_default_dest(priv->fs.ttc, MLX5_TT_ANY);
rule = mlx5_add_flow_rules(fs_any_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -515,7 +514,7 @@ static int fs_any_disable(struct mlx5e_priv *priv)
int err;
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, MLX5_TT_ANY);
+ err = mlx5_ttc_fwd_default_dest(priv->fs.ttc, MLX5_TT_ANY);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -534,7 +533,7 @@ static int fs_any_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.any->table.t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5_ttc_fwd_dest(&priv->fs.ttc, MLX5_TT_ANY, &dest);
+ err = mlx5_ttc_fwd_dest(priv->fs.ttc, MLX5_TT_ANY, &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
index a82be377e9f7..4c4ee524176c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c
@@ -161,7 +161,7 @@ static int accel_fs_tcp_add_default_rule(struct mlx5e_priv *priv,
fs_tcp = priv->fs.accel_tcp;
accel_fs_t = &fs_tcp->tables[type];
- dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, fs_accel2tt(type));
+ dest = mlx5_ttc_get_default_dest(priv->fs.ttc, fs_accel2tt(type));
rule = mlx5_add_flow_rules(accel_fs_t->t, NULL, &flow_act, &dest, 1);
if (IS_ERR(rule)) {
err = PTR_ERR(rule);
@@ -307,7 +307,7 @@ static int accel_fs_tcp_disable(struct mlx5e_priv *priv)
for (i = 0; i < ACCEL_FS_TCP_NUM_TYPES; i++) {
/* Modify ttc rules destination to point back to the indir TIRs */
- err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, fs_accel2tt(i));
+ err = mlx5_ttc_fwd_default_dest(priv->fs.ttc, fs_accel2tt(i));
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -329,7 +329,7 @@ static int accel_fs_tcp_enable(struct mlx5e_priv *priv)
dest.ft = priv->fs.accel_tcp->tables[i].t;
/* Modify ttc rules destination to point on the accel_fs FTs */
- err = mlx5_ttc_fwd_dest(&priv->fs.ttc, fs_accel2tt(i), &dest);
+ err = mlx5_ttc_fwd_dest(priv->fs.ttc, fs_accel2tt(i), &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] destination to accel failed, err(%d)\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
index ff177bb74bb4..17da23dff0ed 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
@@ -265,7 +265,8 @@ static int rx_create(struct mlx5e_priv *priv, enum accel_fs_esp_type type)
accel_esp = priv->ipsec->rx_fs;
fs_prot = &accel_esp->fs_prot[type];
- fs_prot->default_dest = mlx5_ttc_get_default_dest(&priv->fs.ttc, fs_esp2tt(type));
+ fs_prot->default_dest =
+ mlx5_ttc_get_default_dest(priv->fs.ttc, fs_esp2tt(type));
err = rx_err_create_ft(priv, fs_prot, &fs_prot->rx_err);
if (err)
@@ -301,7 +302,7 @@ static int rx_ft_get(struct mlx5e_priv *priv, enum accel_fs_esp_type type)
/* connect */
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
dest.ft = fs_prot->ft;
- mlx5_ttc_fwd_dest(&priv->fs.ttc, fs_esp2tt(type), &dest);
+ mlx5_ttc_fwd_dest(priv->fs.ttc, fs_esp2tt(type), &dest);
out:
mutex_unlock(&fs_prot->prot_mutex);
@@ -320,7 +321,7 @@ static void rx_ft_put(struct mlx5e_priv *priv, enum accel_fs_esp_type type)
goto out;
/* disconnect */
- mlx5_ttc_fwd_default_dest(&priv->fs.ttc, fs_esp2tt(type));
+ mlx5_ttc_fwd_default_dest(priv->fs.ttc, fs_esp2tt(type));
/* remove FT */
rx_destroy(priv, type);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
index 374e262d9917..fe5d82fa6e92 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
@@ -120,7 +120,7 @@ static int arfs_disable(struct mlx5e_priv *priv)
for (i = 0; i < ARFS_NUM_TYPES; i++) {
/* Modify ttc rules destination back to their default */
- err = mlx5_ttc_fwd_default_dest(&priv->fs.ttc, arfs_get_tt(i));
+ err = mlx5_ttc_fwd_default_dest(priv->fs.ttc, arfs_get_tt(i));
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] default destination failed, err(%d)\n",
@@ -149,7 +149,7 @@ int mlx5e_arfs_enable(struct mlx5e_priv *priv)
for (i = 0; i < ARFS_NUM_TYPES; i++) {
dest.ft = priv->fs.arfs->arfs_tables[i].ft.t;
/* Modify ttc rules destination to point on the aRFS FTs */
- err = mlx5_ttc_fwd_dest(&priv->fs.ttc, arfs_get_tt(i), &dest);
+ err = mlx5_ttc_fwd_dest(priv->fs.ttc, arfs_get_tt(i), &dest);
if (err) {
netdev_err(priv->netdev,
"%s: modify ttc[%d] dest to arfs, failed err(%d)\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index cbad05760551..5c754e9af669 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -718,7 +718,7 @@ static int mlx5e_add_promisc_rule(struct mlx5e_priv *priv)
if (!spec)
return -ENOMEM;
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- dest.ft = priv->fs.ttc.t;
+ dest.ft = mlx5_get_ttc_flow_table(priv->fs.ttc);
rule_p = &priv->fs.promisc.rule;
*rule_p = mlx5_add_flow_rules(ft, spec, &flow_act, &dest, 1);
@@ -904,7 +904,8 @@ void mlx5e_set_ttc_params(struct mlx5e_priv *priv,
for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) {
ttc_params->tunnel_dests[tt].type =
MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- ttc_params->tunnel_dests[tt].ft = priv->fs.inner_ttc.t;
+ ttc_params->tunnel_dests[tt].ft =
+ mlx5_get_ttc_flow_table(priv->fs.inner_ttc);
}
}
@@ -938,7 +939,7 @@ static int mlx5e_add_l2_flow_rule(struct mlx5e_priv *priv,
outer_headers.dmac_47_16);
dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- dest.ft = priv->fs.ttc.t;
+ dest.ft = mlx5_get_ttc_flow_table(priv->fs.ttc);
switch (type) {
case MLX5E_FULLMATCH:
@@ -1234,9 +1235,45 @@ static void mlx5e_destroy_vlan_table(struct mlx5e_priv *priv)
kvfree(priv->fs.vlan);
}
-int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
+static void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv)
+{
+ if (!mlx5_tunnel_inner_ft_supported(priv->mdev))
+ return;
+ mlx5_destroy_ttc_table(priv->fs.inner_ttc);
+}
+
+void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv)
+{
+ mlx5_destroy_ttc_table(priv->fs.ttc);
+}
+
+static int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv)
+{
+ struct ttc_params ttc_params = {};
+
+ if (!mlx5_tunnel_inner_ft_supported(priv->mdev))
+ return 0;
+
+ mlx5e_set_inner_ttc_params(priv, &ttc_params);
+ priv->fs.inner_ttc = mlx5_create_ttc_table(priv->mdev, &ttc_params);
+ if (IS_ERR(priv->fs.inner_ttc))
+ return PTR_ERR(priv->fs.inner_ttc);
+ return 0;
+}
+
+int mlx5e_create_ttc_table(struct mlx5e_priv *priv)
{
struct ttc_params ttc_params = {};
+
+ mlx5e_set_ttc_params(priv, &ttc_params, true);
+ priv->fs.ttc = mlx5_create_ttc_table(priv->mdev, &ttc_params);
+ if (IS_ERR(priv->fs.ttc))
+ return PTR_ERR(priv->fs.ttc);
+ return 0;
+}
+
+int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
+{
int err;
priv->fs.ns = mlx5_get_flow_namespace(priv->mdev,
@@ -1252,20 +1289,15 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
priv->netdev->hw_features &= ~NETIF_F_NTUPLE;
}
- if (mlx5_tunnel_inner_ft_supported(priv->mdev)) {
- mlx5e_set_inner_ttc_params(priv, &ttc_params);
- err = mlx5_create_inner_ttc_table(priv->mdev, &ttc_params,
- &priv->fs.inner_ttc);
- if (err) {
- netdev_err(priv->netdev,
- "Failed to create inner ttc table, err=%d\n",
- err);
- goto err_destroy_arfs_tables;
- }
+ err = mlx5e_create_inner_ttc_table(priv);
+ if (err) {
+ netdev_err(priv->netdev,
+ "Failed to create inner ttc table, err=%d\n",
+ err);
+ goto err_destroy_arfs_tables;
}
- mlx5e_set_ttc_params(priv, &ttc_params, true);
- err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &priv->fs.ttc);
+ err = mlx5e_create_ttc_table(priv);
if (err) {
netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n",
err);
@@ -1299,10 +1331,9 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
err_destroy_l2_table:
mlx5e_destroy_l2_table(priv);
err_destroy_ttc_table:
- mlx5_destroy_ttc_table(&priv->fs.ttc);
+ mlx5e_destroy_ttc_table(priv);
err_destroy_inner_ttc_table:
- if (mlx5_tunnel_inner_ft_supported(priv->mdev))
- mlx5_destroy_inner_ttc_table(&priv->fs.inner_ttc);
+ mlx5e_destroy_inner_ttc_table(priv);
err_destroy_arfs_tables:
mlx5e_arfs_destroy_tables(priv);
@@ -1314,9 +1345,8 @@ void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv)
mlx5e_ptp_free_rx_fs(priv);
mlx5e_destroy_vlan_table(priv);
mlx5e_destroy_l2_table(priv);
- mlx5_destroy_ttc_table(&priv->fs.ttc);
- if (mlx5_tunnel_inner_ft_supported(priv->mdev))
- mlx5_destroy_inner_ttc_table(&priv->fs.inner_ttc);
+ mlx5e_destroy_ttc_table(priv);
+ mlx5e_destroy_inner_ttc_table(priv);
mlx5e_arfs_destroy_tables(priv);
mlx5e_ethtool_cleanup_steering(priv);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 9817a176916a..1e520640f7e0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -660,9 +660,11 @@ static int mlx5e_create_rep_ttc_table(struct mlx5e_priv *priv)
/* To give uplik rep TTC a lower level for chaining from root ft */
ttc_params.ft_attr.level = MLX5E_TTC_FT_LEVEL + 1;
- err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &priv->fs.ttc);
- if (err) {
- netdev_err(priv->netdev, "Failed to create rep ttc table, err=%d\n", err);
+ priv->fs.ttc = mlx5_create_ttc_table(priv->mdev, &ttc_params);
+ if (IS_ERR(priv->fs.ttc)) {
+ err = PTR_ERR(priv->fs.ttc);
+ netdev_err(priv->netdev, "Failed to create rep ttc table, err=%d\n",
+ err);
return err;
}
return 0;
@@ -680,7 +682,7 @@ static int mlx5e_create_rep_root_ft(struct mlx5e_priv *priv)
/* non uplik reps will skip any bypass tables and go directly to
* their own ttc
*/
- rpriv->root_ft = priv->fs.ttc.t;
+ rpriv->root_ft = mlx5_get_ttc_flow_table(priv->fs.ttc);
return 0;
}
@@ -794,7 +796,7 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv)
err_destroy_root_ft:
mlx5e_destroy_rep_root_ft(priv);
err_destroy_ttc_table:
- mlx5_destroy_ttc_table(&priv->fs.ttc);
+ mlx5_destroy_ttc_table(priv->fs.ttc);
err_destroy_rx_res:
mlx5e_rx_res_destroy(priv->rx_res);
err_close_drop_rq:
@@ -809,7 +811,7 @@ static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv)
mlx5e_ethtool_cleanup_steering(priv);
rep_vport_rx_rule_destroy(priv);
mlx5e_destroy_rep_root_ft(priv);
- mlx5_destroy_ttc_table(&priv->fs.ttc);
+ mlx5_destroy_ttc_table(priv->fs.ttc);
mlx5e_rx_res_destroy(priv->rx_res);
mlx5e_close_drop_rq(&priv->drop_rq);
mlx5e_rx_res_free(priv->rx_res);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index afbd0caf31ae..1a606dc8bed5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -345,7 +345,7 @@ struct mlx5e_hairpin {
int num_channels;
struct mlx5e_rqt indir_rqt;
struct mlx5e_tir indir_tir[MLX5E_NUM_INDIR_TIRS];
- struct mlx5_ttc_table ttc;
+ struct mlx5_ttc_table *ttc;
};
struct mlx5e_hairpin_entry {
@@ -624,12 +624,15 @@ static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp)
goto err_create_indirect_tirs;
mlx5e_hairpin_set_ttc_params(hp, &ttc_params);
- err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &hp->ttc);
- if (err)
+ hp->ttc = mlx5_create_ttc_table(priv->mdev, &ttc_params);
+ if (IS_ERR(hp->ttc)) {
+ err = PTR_ERR(hp->ttc);
goto err_create_ttc_table;
+ }
netdev_dbg(priv->netdev, "add hairpin: using %d channels rss ttc table id %x\n",
- hp->num_channels, hp->ttc.t->id);
+ hp->num_channels,
+ mlx5_get_ttc_flow_table(priv->fs.ttc)->id);
return 0;
@@ -643,7 +646,7 @@ static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp)
static void mlx5e_hairpin_rss_cleanup(struct mlx5e_hairpin *hp)
{
- mlx5_destroy_ttc_table(&hp->ttc);
+ mlx5_destroy_ttc_table(hp->ttc);
mlx5e_hairpin_destroy_indirect_tirs(hp);
mlx5e_rqt_destroy(&hp->indir_rqt);
}
@@ -887,7 +890,8 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv,
attach_flow:
if (hpe->hp->num_channels > 1) {
flow_flag_set(flow, HAIRPIN_RSS);
- flow->attr->nic_attr->hairpin_ft = hpe->hp->ttc.t;
+ flow->attr->nic_attr->hairpin_ft =
+ mlx5_get_ttc_flow_table(hpe->hp->ttc);
} else {
flow->attr->nic_attr->hairpin_tirn = mlx5e_tir_get_tirn(&hpe->hp->direct_tir);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
index e04b758f20e3..67571e5040d6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
@@ -314,7 +314,6 @@ static void mlx5i_cleanup_tx(struct mlx5e_priv *priv)
static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
{
- struct ttc_params ttc_params = {};
int err;
priv->fs.ns = mlx5_get_flow_namespace(priv->mdev,
@@ -330,8 +329,7 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
priv->netdev->hw_features &= ~NETIF_F_NTUPLE;
}
- mlx5e_set_ttc_params(priv, &ttc_params, true);
- err = mlx5_create_ttc_table(priv->mdev, &ttc_params, &priv->fs.ttc);
+ err = mlx5e_create_ttc_table(priv);
if (err) {
netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n",
err);
@@ -348,7 +346,7 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv)
{
- mlx5_destroy_ttc_table(&priv->fs.ttc);
+ mlx5e_destroy_ttc_table(priv);
mlx5e_arfs_destroy_tables(priv);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
index 4b54b4127d33..749d17c0057d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.c
@@ -25,6 +25,20 @@
MLX5_INNER_TTC_GROUP2_SIZE +\
MLX5_INNER_TTC_GROUP3_SIZE)
+/* L3/L4 traffic type classifier */
+struct mlx5_ttc_table {
+ int num_groups;
+ struct mlx5_flow_table *t;
+ struct mlx5_flow_group **g;
+ struct mlx5_ttc_rule rules[MLX5_NUM_TT];
+ struct mlx5_flow_handle *tunnel_rules[MLX5_NUM_TUNNEL_TT];
+};
+
+struct mlx5_flow_table *mlx5_get_ttc_flow_table(struct mlx5_ttc_table *ttc)
+{
+ return ttc->t;
+}
+
static void mlx5_cleanup_ttc_rules(struct mlx5_ttc_table *ttc)
{
int i;
@@ -473,19 +487,23 @@ static int mlx5_create_inner_ttc_table_groups(struct mlx5_ttc_table *ttc)
return err;
}
-int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
- struct ttc_params *params,
- struct mlx5_ttc_table *ttc)
+struct mlx5_ttc_table *mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params)
{
+ struct mlx5_ttc_table *ttc;
int err;
+ ttc = kvzalloc(sizeof(*ttc), GFP_KERNEL);
+ if (!ttc)
+ return ERR_PTR(-ENOMEM);
+
WARN_ON_ONCE(params->ft_attr.max_fte);
params->ft_attr.max_fte = MLX5_INNER_TTC_TABLE_SIZE;
ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
if (IS_ERR(ttc->t)) {
err = PTR_ERR(ttc->t);
- ttc->t = NULL;
- return err;
+ kvfree(ttc);
+ return ERR_PTR(err);
}
err = mlx5_create_inner_ttc_table_groups(ttc);
@@ -496,11 +514,11 @@ int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
if (err)
goto destroy_ft;
- return 0;
+ return ttc;
destroy_ft:
mlx5_destroy_ttc_table(ttc);
- return err;
+ return ERR_PTR(err);
}
void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc)
@@ -514,32 +532,31 @@ void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc)
ttc->g[i] = NULL;
}
- ttc->num_groups = 0;
kfree(ttc->g);
mlx5_destroy_flow_table(ttc->t);
- ttc->t = NULL;
+ kvfree(ttc);
}
-void mlx5_destroy_inner_ttc_table(struct mlx5_ttc_table *ttc)
-{
- mlx5_destroy_ttc_table(ttc);
-}
-
-int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
- struct mlx5_ttc_table *ttc)
+struct mlx5_ttc_table *mlx5_create_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params)
{
bool match_ipv_outer =
MLX5_CAP_FLOWTABLE_NIC_RX(dev,
ft_field_support.outer_ip_version);
+ struct mlx5_ttc_table *ttc;
int err;
+ ttc = kvzalloc(sizeof(*ttc), GFP_KERNEL);
+ if (!ttc)
+ return ERR_PTR(-ENOMEM);
+
WARN_ON_ONCE(params->ft_attr.max_fte);
params->ft_attr.max_fte = MLX5_TTC_TABLE_SIZE;
ttc->t = mlx5_create_flow_table(params->ns, ¶ms->ft_attr);
if (IS_ERR(ttc->t)) {
err = PTR_ERR(ttc->t);
- ttc->t = NULL;
- return err;
+ kvfree(ttc);
+ return ERR_PTR(err);
}
err = mlx5_create_ttc_table_groups(ttc, match_ipv_outer);
@@ -550,10 +567,11 @@ int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
if (err)
goto destroy_ft;
- return 0;
+ return ttc;
+
destroy_ft:
mlx5_destroy_ttc_table(ttc);
- return err;
+ return ERR_PTR(err);
}
int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
index 1010e00c10bd..ce95be8f8382 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_ttc.h
@@ -37,14 +37,7 @@ struct mlx5_ttc_rule {
struct mlx5_flow_destination default_dest;
};
-/* L3/L4 traffic type classifier */
-struct mlx5_ttc_table {
- int num_groups;
- struct mlx5_flow_table *t;
- struct mlx5_flow_group **g;
- struct mlx5_ttc_rule rules[MLX5_NUM_TT];
- struct mlx5_flow_handle *tunnel_rules[MLX5_NUM_TUNNEL_TT];
-};
+struct mlx5_ttc_table;
struct ttc_params {
struct mlx5_flow_namespace *ns;
@@ -54,14 +47,14 @@ struct ttc_params {
struct mlx5_flow_destination tunnel_dests[MLX5_NUM_TUNNEL_TT];
};
-int mlx5_create_ttc_table(struct mlx5_core_dev *dev, struct ttc_params *params,
- struct mlx5_ttc_table *ttc);
+struct mlx5_flow_table *mlx5_get_ttc_flow_table(struct mlx5_ttc_table *ttc);
+
+struct mlx5_ttc_table *mlx5_create_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params);
void mlx5_destroy_ttc_table(struct mlx5_ttc_table *ttc);
-int mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
- struct ttc_params *params,
- struct mlx5_ttc_table *ttc);
-void mlx5_destroy_inner_ttc_table(struct mlx5_ttc_table *ttc);
+struct mlx5_ttc_table *mlx5_create_inner_ttc_table(struct mlx5_core_dev *dev,
+ struct ttc_params *params);
int mlx5_ttc_fwd_dest(struct mlx5_ttc_table *ttc, enum mlx5_traffic_types type,
struct mlx5_flow_destination *new_dest);
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 10/16] net/mlx5e: Remove redundant tc act includes
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (8 preceding siblings ...)
2021-08-03 2:28 ` [net-next 09/16] net/mlx5: Embed mlx5_ttc_table Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 11/16] net/mlx5e: Remove redundant filter_dev arg from parse_tc_fdb_actions() Saeed Mahameed
` (5 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Roi Dayan, Maor Dickman, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Since the code changed to use the flow action infra
there is no usage of tcf values from those includes.
Remove those.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 6 ------
1 file changed, 6 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 1a606dc8bed5..9671fb0e1432 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -34,19 +34,13 @@
#include <net/flow_offload.h>
#include <net/sch_generic.h>
#include <net/pkt_cls.h>
-#include <net/tc_act/tc_gact.h>
-#include <net/tc_act/tc_skbedit.h>
#include <linux/mlx5/fs.h>
#include <linux/mlx5/device.h>
#include <linux/rhashtable.h>
#include <linux/refcount.h>
#include <linux/completion.h>
-#include <net/tc_act/tc_mirred.h>
-#include <net/tc_act/tc_vlan.h>
-#include <net/tc_act/tc_tunnel_key.h>
#include <net/tc_act/tc_pedit.h>
#include <net/tc_act/tc_csum.h>
-#include <net/tc_act/tc_mpls.h>
#include <net/psample.h>
#include <net/arp.h>
#include <net/ipv6_stubs.h>
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 11/16] net/mlx5e: Remove redundant filter_dev arg from parse_tc_fdb_actions()
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (9 preceding siblings ...)
2021-08-03 2:28 ` [net-next 10/16] net/mlx5e: Remove redundant tc act includes Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 12/16] net/mlx5e: Remove redundant cap check for flow counter Saeed Mahameed
` (4 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Roi Dayan, Maor Dickman, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
filter_dev is saved in parse_attr. and being used in other cases from
there. use it also for the leftover case.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 9671fb0e1432..f4f8a6728e95 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -3707,8 +3707,7 @@ static int verify_uplink_forwarding(struct mlx5e_priv *priv,
static int parse_tc_fdb_actions(struct mlx5e_priv *priv,
struct flow_action *flow_action,
struct mlx5e_tc_flow *flow,
- struct netlink_ext_ack *extack,
- struct net_device *filter_dev)
+ struct netlink_ext_ack *extack)
{
struct pedit_headers_action hdrs[2] = {};
struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
@@ -3773,7 +3772,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv,
"mpls pop supported only as first action");
return -EOPNOTSUPP;
}
- if (!netif_is_bareudp(filter_dev)) {
+ if (!netif_is_bareudp(parse_attr->filter_dev)) {
NL_SET_ERR_MSG_MOD(extack,
"mpls pop supported only on bareudp devices");
return -EOPNOTSUPP;
@@ -4275,7 +4274,7 @@ __mlx5e_add_fdb_flow(struct mlx5e_priv *priv,
if (err)
goto err_free;
- err = parse_tc_fdb_actions(priv, &rule->action, flow, extack, filter_dev);
+ err = parse_tc_fdb_actions(priv, &rule->action, flow, extack);
if (err)
goto err_free;
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 12/16] net/mlx5e: Remove redundant cap check for flow counter
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (10 preceding siblings ...)
2021-08-03 2:28 ` [net-next 11/16] net/mlx5e: Remove redundant filter_dev arg from parse_tc_fdb_actions() Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 13/16] net/mlx5e: Remove redundant parse_attr arg Saeed Mahameed
` (3 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Roi Dayan, Maor Dickman, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
The cap is very old and today will always exists.
The cap is not being checked anywhere else. Remove the check from
drop action when parsing tc rules in nic mode.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index f4f8a6728e95..3453f37a0741 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -3356,10 +3356,8 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv,
MLX5_FLOW_CONTEXT_ACTION_COUNT;
break;
case FLOW_ACTION_DROP:
- action |= MLX5_FLOW_CONTEXT_ACTION_DROP;
- if (MLX5_CAP_FLOWTABLE(priv->mdev,
- flow_table_properties_nic_receive.flow_counter))
- action |= MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ action |= MLX5_FLOW_CONTEXT_ACTION_DROP |
+ MLX5_FLOW_CONTEXT_ACTION_COUNT;
break;
case FLOW_ACTION_MANGLE:
case FLOW_ACTION_ADD:
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 13/16] net/mlx5e: Remove redundant parse_attr arg
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (11 preceding siblings ...)
2021-08-03 2:28 ` [net-next 12/16] net/mlx5e: Remove redundant cap check for flow counter Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 14/16] net/mlx5e: Remove redundant assignment of counter to null Saeed Mahameed
` (2 subsequent siblings)
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Roi Dayan, Maor Dickman, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Passing parse_attr is redundant in parse_tc_nic_actions() and
mlx5e_tc_add_nic_flow() as we can get it from flow.
This is the same as with parse_tc_fdb_actions() and mlx5e_tc_add_fdb_flow().
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 12 +++++++-----
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 3453f37a0741..472c0c756a69 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1031,15 +1031,17 @@ mlx5e_add_offloaded_nic_rule(struct mlx5e_priv *priv,
static int
mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv,
- struct mlx5e_tc_flow_parse_attr *parse_attr,
struct mlx5e_tc_flow *flow,
struct netlink_ext_ack *extack)
{
+ struct mlx5e_tc_flow_parse_attr *parse_attr;
struct mlx5_flow_attr *attr = flow->attr;
struct mlx5_core_dev *dev = priv->mdev;
struct mlx5_fc *counter = NULL;
int err;
+ parse_attr = attr->parse_attr;
+
if (flow_flag_test(flow, HAIRPIN)) {
err = mlx5e_hairpin_flow_add(priv, flow, parse_attr, extack);
if (err)
@@ -3327,10 +3329,10 @@ static int validate_goto_chain(struct mlx5e_priv *priv,
static int parse_tc_nic_actions(struct mlx5e_priv *priv,
struct flow_action *flow_action,
- struct mlx5e_tc_flow_parse_attr *parse_attr,
struct mlx5e_tc_flow *flow,
struct netlink_ext_ack *extack)
{
+ struct mlx5e_tc_flow_parse_attr *parse_attr;
struct mlx5_flow_attr *attr = flow->attr;
struct pedit_headers_action hdrs[2] = {};
const struct flow_action_entry *act;
@@ -3346,8 +3348,8 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv,
return -EOPNOTSUPP;
nic_attr = attr->nic_attr;
-
nic_attr->flow_tag = MLX5_FS_DEFAULT_FLOW_TAG;
+ parse_attr = attr->parse_attr;
flow_action_for_each(i, act, flow_action) {
switch (act->id) {
@@ -4418,11 +4420,11 @@ mlx5e_add_nic_flow(struct mlx5e_priv *priv,
if (err)
goto err_free;
- err = parse_tc_nic_actions(priv, &rule->action, parse_attr, flow, extack);
+ err = parse_tc_nic_actions(priv, &rule->action, flow, extack);
if (err)
goto err_free;
- err = mlx5e_tc_add_nic_flow(priv, parse_attr, flow, extack);
+ err = mlx5e_tc_add_nic_flow(priv, flow, extack);
if (err)
goto err_free;
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 14/16] net/mlx5e: Remove redundant assignment of counter to null
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (12 preceding siblings ...)
2021-08-03 2:28 ` [net-next 13/16] net/mlx5e: Remove redundant parse_attr arg Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 15/16] net/mlx5e: Return -EOPNOTSUPP if more relevant when parsing tc actions Saeed Mahameed
2021-08-03 2:28 ` [net-next 16/16] net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set() Saeed Mahameed
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Roi Dayan, Maor Dickman, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
counter is being initialized before being used.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 472c0c756a69..980a668bbc3c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1037,7 +1037,7 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv,
struct mlx5e_tc_flow_parse_attr *parse_attr;
struct mlx5_flow_attr *attr = flow->attr;
struct mlx5_core_dev *dev = priv->mdev;
- struct mlx5_fc *counter = NULL;
+ struct mlx5_fc *counter;
int err;
parse_attr = attr->parse_attr;
@@ -1361,9 +1361,9 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv,
bool vf_tun = false, encap_valid = true;
struct net_device *encap_dev = NULL;
struct mlx5_esw_flow_attr *esw_attr;
- struct mlx5_fc *counter = NULL;
struct mlx5e_rep_priv *rpriv;
struct mlx5e_priv *out_priv;
+ struct mlx5_fc *counter;
u32 max_prio, max_chain;
int err = 0;
int out_index;
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 15/16] net/mlx5e: Return -EOPNOTSUPP if more relevant when parsing tc actions
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (13 preceding siblings ...)
2021-08-03 2:28 ` [net-next 14/16] net/mlx5e: Remove redundant assignment of counter to null Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
2021-08-03 2:28 ` [net-next 16/16] net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set() Saeed Mahameed
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Roi Dayan, Maor Dickman, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Instead of returning -EINVAL.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 980a668bbc3c..349a93e0213d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -3400,7 +3400,7 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv,
"device is not on same HW, can't offload");
netdev_warn(priv->netdev, "device %s not on same HW, can't offload\n",
peer_dev->name);
- return -EINVAL;
+ return -EOPNOTSUPP;
}
}
break;
@@ -3410,7 +3410,7 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv,
if (mark & ~MLX5E_TC_FLOW_ID_MASK) {
NL_SET_ERR_MSG_MOD(extack,
"Bad flow mark - only 16 bit is supported");
- return -EINVAL;
+ return -EOPNOTSUPP;
}
nic_attr->flow_tag = mark;
@@ -3921,7 +3921,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv,
"devices %s %s not on same switch HW, can't offload forwarding\n",
priv->netdev->name,
out_dev->name);
- return -EINVAL;
+ return -EOPNOTSUPP;
}
}
break;
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [net-next 16/16] net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set()
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
` (14 preceding siblings ...)
2021-08-03 2:28 ` [net-next 15/16] net/mlx5e: Return -EOPNOTSUPP if more relevant when parsing tc actions Saeed Mahameed
@ 2021-08-03 2:28 ` Saeed Mahameed
15 siblings, 0 replies; 18+ messages in thread
From: Saeed Mahameed @ 2021-08-03 2:28 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Jiapeng Chong, Abaci Robot, Leon Romanovsky, Saeed Mahameed
From: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
The return value is missing in this code scenario, add the return value
'0' to the return value 'err'.
Eliminate the follow smatch warning:
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c:3083
mlx5_devlink_eswitch_inline_mode_set() warn: missing error code 'err'.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Fixes: 8e0aa4bc959c ("net/mlx5: E-switch, Protect eswitch mode changes")
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 011e766e4f67..feecf44994a9 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -3083,8 +3083,11 @@ int mlx5_devlink_eswitch_inline_mode_set(struct devlink *devlink, u8 mode,
switch (MLX5_CAP_ETH(dev, wqe_inline_mode)) {
case MLX5_CAP_INLINE_MODE_NOT_REQUIRED:
- if (mode == DEVLINK_ESWITCH_INLINE_MODE_NONE)
+ if (mode == DEVLINK_ESWITCH_INLINE_MODE_NONE) {
+ err = 0;
goto out;
+ }
+
fallthrough;
case MLX5_CAP_INLINE_MODE_L2:
NL_SET_ERR_MSG_MOD(extack, "Inline mode can't be set");
--
2.31.1
^ permalink raw reply related [flat|nested] 18+ messages in thread
* Re: [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table
2021-08-03 2:28 ` [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table Saeed Mahameed
@ 2021-08-03 10:30 ` patchwork-bot+netdevbpf
0 siblings, 0 replies; 18+ messages in thread
From: patchwork-bot+netdevbpf @ 2021-08-03 10:30 UTC (permalink / raw)
To: Saeed Mahameed; +Cc: davem, kuba, netdev, maximmi, tariqt, saeedm
Hello:
This series was applied to netdev/net-next.git (refs/heads/master):
On Mon, 2 Aug 2021 19:28:38 -0700 you wrote:
> From: Maxim Mikityanskiy <maximmi@nvidia.com>
>
> Replace mlx5e_build_default_indir_rqt with a new initializer of struct
> mlx5e_rss_params_indir that works directly with the struct, rather than
> its internals.
>
> The new initializer is called mlx5e_rss_params_indir_init_uniform, which
> also reflects the purpose (uniform spreading) better.
>
> [...]
Here is the summary with links:
- [net-next,01/16] net/mlx5e: Use a new initializer to build uniform indir table
https://git.kernel.org/netdev/net-next/c/43befe99bc62
- [net-next,02/16] net/mlx5e: Introduce mlx5e_channels API to get RQNs
https://git.kernel.org/netdev/net-next/c/e6e01b5fdc28
- [net-next,03/16] net/mlx5e: Hide all implementation details of mlx5e_rx_res
https://git.kernel.org/netdev/net-next/c/43ec0f41fa73
- [net-next,04/16] net/mlx5e: Allocate the array of channels according to the real max_nch
https://git.kernel.org/netdev/net-next/c/3ac90dec3a01
- [net-next,05/16] net/mlx5e: Rename traffic type enums
https://git.kernel.org/netdev/net-next/c/d443c6f684d3
- [net-next,06/16] net/mlx5e: Rename some related TTC args and functions
https://git.kernel.org/netdev/net-next/c/5fba089e960c
- [net-next,07/16] net/mlx5e: Decouple TTC logic from mlx5e
https://git.kernel.org/netdev/net-next/c/bc29764ed9a2
- [net-next,08/16] net/mlx5: Move TTC logic to fs_ttc
https://git.kernel.org/netdev/net-next/c/371cf74e78f3
- [net-next,09/16] net/mlx5: Embed mlx5_ttc_table
https://git.kernel.org/netdev/net-next/c/f4b45940e9b9
- [net-next,10/16] net/mlx5e: Remove redundant tc act includes
https://git.kernel.org/netdev/net-next/c/696ceeb203c7
- [net-next,11/16] net/mlx5e: Remove redundant filter_dev arg from parse_tc_fdb_actions()
https://git.kernel.org/netdev/net-next/c/70f8019e7b56
- [net-next,12/16] net/mlx5e: Remove redundant cap check for flow counter
https://git.kernel.org/netdev/net-next/c/950b4df9fba9
- [net-next,13/16] net/mlx5e: Remove redundant parse_attr arg
https://git.kernel.org/netdev/net-next/c/c6cfe1137f88
- [net-next,14/16] net/mlx5e: Remove redundant assignment of counter to null
https://git.kernel.org/netdev/net-next/c/97a8d29ae9d2
- [net-next,15/16] net/mlx5e: Return -EOPNOTSUPP if more relevant when parsing tc actions
https://git.kernel.org/netdev/net-next/c/25f150f4bbe9
- [net-next,16/16] net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set()
https://git.kernel.org/netdev/net-next/c/bcd68c04c769
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 18+ messages in thread
end of thread, other threads:[~2021-08-03 10:30 UTC | newest]
Thread overview: 18+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-08-03 2:28 [pull request][net-next 00/16] mlx5 updates 2021-08-02 Saeed Mahameed
2021-08-03 2:28 ` [net-next 01/16] net/mlx5e: Use a new initializer to build uniform indir table Saeed Mahameed
2021-08-03 10:30 ` patchwork-bot+netdevbpf
2021-08-03 2:28 ` [net-next 02/16] net/mlx5e: Introduce mlx5e_channels API to get RQNs Saeed Mahameed
2021-08-03 2:28 ` [net-next 03/16] net/mlx5e: Hide all implementation details of mlx5e_rx_res Saeed Mahameed
2021-08-03 2:28 ` [net-next 04/16] net/mlx5e: Allocate the array of channels according to the real max_nch Saeed Mahameed
2021-08-03 2:28 ` [net-next 05/16] net/mlx5e: Rename traffic type enums Saeed Mahameed
2021-08-03 2:28 ` [net-next 06/16] net/mlx5e: Rename some related TTC args and functions Saeed Mahameed
2021-08-03 2:28 ` [net-next 07/16] net/mlx5e: Decouple TTC logic from mlx5e Saeed Mahameed
2021-08-03 2:28 ` [net-next 08/16] net/mlx5: Move TTC logic to fs_ttc Saeed Mahameed
2021-08-03 2:28 ` [net-next 09/16] net/mlx5: Embed mlx5_ttc_table Saeed Mahameed
2021-08-03 2:28 ` [net-next 10/16] net/mlx5e: Remove redundant tc act includes Saeed Mahameed
2021-08-03 2:28 ` [net-next 11/16] net/mlx5e: Remove redundant filter_dev arg from parse_tc_fdb_actions() Saeed Mahameed
2021-08-03 2:28 ` [net-next 12/16] net/mlx5e: Remove redundant cap check for flow counter Saeed Mahameed
2021-08-03 2:28 ` [net-next 13/16] net/mlx5e: Remove redundant parse_attr arg Saeed Mahameed
2021-08-03 2:28 ` [net-next 14/16] net/mlx5e: Remove redundant assignment of counter to null Saeed Mahameed
2021-08-03 2:28 ` [net-next 15/16] net/mlx5e: Return -EOPNOTSUPP if more relevant when parsing tc actions Saeed Mahameed
2021-08-03 2:28 ` [net-next 16/16] net/mlx5: Fix missing return value in mlx5_devlink_eswitch_inline_mode_set() Saeed Mahameed
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).