* [for-next 01/15] net/mlx5: Vport, Use 'kvfree()' for memory allocated by 'kvzalloc()'
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Christophe JAILLET, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
When 'kvzalloc()' is used to allocate memory, 'kvfree()' must be used to
free it.
Fixes: 9efa75254593d ("net/mlx5_core: Introduce access functions to query vport RoCE fields")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/vport.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/vport.c b/drivers/net/ethernet/mellanox/mlx5/core/vport.c
index 177e076b8d17..719cecb182c6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/vport.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/vport.c
@@ -511,7 +511,7 @@ int mlx5_query_nic_vport_system_image_guid(struct mlx5_core_dev *mdev,
*system_image_guid = MLX5_GET64(query_nic_vport_context_out, out,
nic_vport_context.system_image_guid);
- kfree(out);
+ kvfree(out);
return 0;
}
@@ -531,7 +531,7 @@ int mlx5_query_nic_vport_node_guid(struct mlx5_core_dev *mdev, u64 *node_guid)
*node_guid = MLX5_GET64(query_nic_vport_context_out, out,
nic_vport_context.node_guid);
- kfree(out);
+ kvfree(out);
return 0;
}
@@ -587,7 +587,7 @@ int mlx5_query_nic_vport_qkey_viol_cntr(struct mlx5_core_dev *mdev,
*qkey_viol_cntr = MLX5_GET(query_nic_vport_context_out, out,
nic_vport_context.qkey_violation_counter);
- kfree(out);
+ kvfree(out);
return 0;
}
--
2.17.0
^ permalink raw reply related
* [for-next 02/15] net/mlx5: Eswitch, Use 'kvfree()' for memory allocated by 'kvzalloc()'
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Christophe JAILLET, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
When 'kvzalloc()' is used to allocate memory, 'kvfree()' must be used to
free it.
Fixes: fed9ce22bf8ae ("net/mlx5: E-Switch, Add API to create vport rx rules")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 2 +-
1 file changed, 1 insertion(+), 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 35e256eb2f6e..b123f8a52ad8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -663,7 +663,7 @@ static int esw_create_vport_rx_group(struct mlx5_eswitch *esw)
esw->offloads.vport_rx_group = g;
out:
- kfree(flow_group_in);
+ kvfree(flow_group_in);
return err;
}
--
2.17.0
^ permalink raw reply related
* [for-next 03/15] IB/mlx5: Use 'kvfree()' for memory allocated by 'kvzalloc()'
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Christophe JAILLET, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
When 'kvzalloc()' is used to allocate memory, 'kvfree()' must be used to
free it.
Fixes: 1cbe6fc86ccfe ("IB/mlx5: Add support for CQE compressing")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Acked-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/infiniband/hw/mlx5/cq.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 77d257ec899b..6d52ea03574e 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -849,7 +849,7 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata,
return 0;
err_cqb:
- kfree(*cqb);
+ kvfree(*cqb);
err_db:
mlx5_ib_db_unmap_user(to_mucontext(context), &cq->db);
--
2.17.0
^ permalink raw reply related
* [for-next 05/15] net/mlx5: Properly handle a vport destination when setting FTE
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Shahar Klein, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Shahar Klein <shahark@mellanox.com>
When creating FTE, properly distinguish between destination being vport
or tir. The previous code just worked accidentally b/c of both dest being
in the same offset within a union.
Signed-off-by: Shahar Klein <shahark@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index ef5afd7c9325..0bfce6a82c91 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -372,6 +372,9 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
if (dst->dest_attr.type ==
MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE) {
id = dst->dest_attr.ft->id;
+ } else if (dst->dest_attr.type ==
+ MLX5_FLOW_DESTINATION_TYPE_VPORT) {
+ id = dst->dest_attr.vport_num;
} else {
id = dst->dest_attr.tir_num;
}
--
2.17.0
^ permalink raw reply related
* [for-next 04/15] net/mlx5: Add merged e-switch cap
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Roi Dayan, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Roi Dayan <roid@mellanox.com>
When merged e-switch is supported, the per-port e-switch is logically
merged into one e-switch that spans both physical ports and all the VFs.
Under merged eswitch, both the matching on source vport and setting
destination vport can have a 2nd attribute which is the vhca id of the
eswitch owner.
For example:
esw0: {match: <src vport=1 owner=0> action: fwd to <dst vport=7, owner=1>}
is a flow set on eswitch0 matching on source vport=1 from his eswitch
and the action being fwd to dest vport=7 of eswitch1.
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Shahar Klein <shahark@mellanox.com>
Reviewed-by: Or Gerlitz Klein <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
include/linux/mlx5/mlx5_ifc.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 1aad455538f4..ef15f751a984 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -557,7 +557,8 @@ struct mlx5_ifc_e_switch_cap_bits {
u8 vport_svlan_insert[0x1];
u8 vport_cvlan_insert_if_not_exist[0x1];
u8 vport_cvlan_insert_overwrite[0x1];
- u8 reserved_at_5[0x19];
+ u8 reserved_at_5[0x18];
+ u8 merged_eswitch[0x1];
u8 nic_vport_node_guid_modify[0x1];
u8 nic_vport_port_guid_modify[0x1];
--
2.17.0
^ permalink raw reply related
* [for-next 06/15] net/mlx5: Add destination e-switch owner
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Shahar Klein, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Shahar Klein <shahark@mellanox.com>
The destination e-switch owner allows a rule in namespace of one e-switch
owner to point to a vport that is natively associated with another
e-switch owner.
Signed-off-by: Shahar Klein <shahark@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 2 +-
.../net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 6 +++---
drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c | 8 +++++++-
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 2 +-
include/linux/mlx5/fs.h | 6 +++++-
include/linux/mlx5/mlx5_ifc.h | 5 +++--
7 files changed, 21 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c
index d93ff567b40d..b3820a34e773 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c
@@ -235,7 +235,7 @@ const char *parse_fs_dst(struct trace_seq *p,
switch (dst->type) {
case MLX5_FLOW_DESTINATION_TYPE_VPORT:
- trace_seq_printf(p, "vport=%u\n", dst->vport_num);
+ trace_seq_printf(p, "vport=%u\n", dst->vport.num);
break;
case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE:
trace_seq_printf(p, "ft=%p\n", dst->ft);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index 332bc56306bf..9a24314b817a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -192,7 +192,7 @@ __esw_fdb_set_vport_rule(struct mlx5_eswitch *esw, u32 vport, bool rx_rule,
}
dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
- dest.vport_num = vport;
+ dest.vport.num = vport;
esw_debug(esw->dev,
"\tFDB add rule dmac_v(%pM) dmac_c(%pM) -> vport(%d)\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index b123f8a52ad8..90c8cb31e633 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -71,7 +71,7 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw,
if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) {
dest[i].type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
- dest[i].vport_num = attr->out_rep->vport;
+ dest[i].vport.num = attr->out_rep->vport;
i++;
}
if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_COUNT) {
@@ -343,7 +343,7 @@ mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *esw, int vport, u32 sqn
spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS;
dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
- dest.vport_num = vport;
+ dest.vport.num = vport;
flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.fdb, spec,
@@ -387,7 +387,7 @@ static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw)
dmac_c[0] = 0x01;
dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
- dest.vport_num = 0;
+ dest.vport.num = 0;
flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.fdb, spec,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index 0bfce6a82c91..5a00deff5457 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -374,7 +374,13 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
id = dst->dest_attr.ft->id;
} else if (dst->dest_attr.type ==
MLX5_FLOW_DESTINATION_TYPE_VPORT) {
- id = dst->dest_attr.vport_num;
+ id = dst->dest_attr.vport.num;
+ MLX5_SET(dest_format_struct, in_dests,
+ destination_eswitch_owner_vhca_id_valid,
+ dst->dest_attr.vport.vhca_id_valid);
+ MLX5_SET(dest_format_struct, in_dests,
+ destination_eswitch_owner_vhca_id,
+ dst->dest_attr.vport.vhca_id);
} else {
id = dst->dest_attr.tir_num;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index de51e7c39bc8..5a80279b052a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -1404,7 +1404,7 @@ static bool mlx5_flow_dests_cmp(struct mlx5_flow_destination *d1,
{
if (d1->type == d2->type) {
if ((d1->type == MLX5_FLOW_DESTINATION_TYPE_VPORT &&
- d1->vport_num == d2->vport_num) ||
+ d1->vport.num == d2->vport.num) ||
(d1->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE &&
d1->ft == d2->ft) ||
(d1->type == MLX5_FLOW_DESTINATION_TYPE_TIR &&
diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h
index 47aecc4fa8c2..9f4d32e41c06 100644
--- a/include/linux/mlx5/fs.h
+++ b/include/linux/mlx5/fs.h
@@ -90,8 +90,12 @@ struct mlx5_flow_destination {
union {
u32 tir_num;
struct mlx5_flow_table *ft;
- u32 vport_num;
struct mlx5_fc *counter;
+ struct {
+ u16 num;
+ u16 vhca_id;
+ bool vhca_id_valid;
+ } vport;
};
};
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index ef15f751a984..3d17709bc30c 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -1148,8 +1148,9 @@ enum mlx5_flow_destination_type {
struct mlx5_ifc_dest_format_struct_bits {
u8 destination_type[0x8];
u8 destination_id[0x18];
-
- u8 reserved_at_20[0x20];
+ u8 destination_eswitch_owner_vhca_id_valid[0x1];
+ u8 reserved_at_21[0xf];
+ u8 destination_eswitch_owner_vhca_id[0x10];
};
struct mlx5_ifc_flow_counter_list_bits {
--
2.17.0
^ permalink raw reply related
* [for-next 08/15] net/mlx5: Add source e-switch owner
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Shahar Klein, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Shahar Klein <shahark@mellanox.com>
The source e-switch owner allows a vport on one e-switch port be associated
with a rule defined on the second port e-switch.
The role of the source eswitch owner valid bit in the flow group is to
allow the firmware fail driver attempts to wild card the source eswitch
match field. If this bit is not set, the firmware ignores the source
eswitch owner field totally.
Signed-off-by: Shahar Klein <shahark@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 10 ++++++++++
include/linux/mlx5/mlx5_ifc.h | 6 ++++--
2 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 5a80279b052a..b1a2ca0ff320 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -1372,6 +1372,8 @@ static int create_auto_flow_group(struct mlx5_flow_table *ft,
struct mlx5_core_dev *dev = get_dev(&ft->node);
int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
void *match_criteria_addr;
+ u8 src_esw_owner_mask_on;
+ void *misc;
int err;
u32 *in;
@@ -1384,6 +1386,14 @@ static int create_auto_flow_group(struct mlx5_flow_table *ft,
MLX5_SET(create_flow_group_in, in, start_flow_index, fg->start_index);
MLX5_SET(create_flow_group_in, in, end_flow_index, fg->start_index +
fg->max_ftes - 1);
+
+ misc = MLX5_ADDR_OF(fte_match_param, fg->mask.match_criteria,
+ misc_parameters);
+ src_esw_owner_mask_on = !!MLX5_GET(fte_match_set_misc, misc,
+ source_eswitch_owner_vhca_id);
+ MLX5_SET(create_flow_group_in, in,
+ source_eswitch_owner_vhca_id_valid, src_esw_owner_mask_on);
+
match_criteria_addr = MLX5_ADDR_OF(create_flow_group_in,
in, match_criteria);
memcpy(match_criteria_addr, fg->mask.match_criteria,
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 3d17709bc30c..9c3538f1b8b9 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -412,7 +412,7 @@ struct mlx5_ifc_fte_match_set_misc_bits {
u8 reserved_at_0[0x8];
u8 source_sqn[0x18];
- u8 reserved_at_20[0x10];
+ u8 source_eswitch_owner_vhca_id[0x10];
u8 source_port[0x10];
u8 outer_second_prio[0x3];
@@ -6995,7 +6995,9 @@ struct mlx5_ifc_create_flow_group_in_bits {
u8 reserved_at_a0[0x8];
u8 table_id[0x18];
- u8 reserved_at_c0[0x20];
+ u8 source_eswitch_owner_vhca_id_valid[0x1];
+
+ u8 reserved_at_c1[0x1f];
u8 start_flow_index[0x20];
--
2.17.0
^ permalink raw reply related
* [for-next 07/15] net/mlx5e: Explicitly set destination e-switch in FDB rules
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Rabie Loulou, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Rabie Loulou <rabiel@mellanox.com>
Set a specific destination e-switch when setting a destination vport.
Signed-off-by: Rabie Loulou <rabiel@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Shahar Klein <shahark@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 2 ++
drivers/net/ethernet/mellanox/mlx5/core/eswitch.h | 1 +
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 5 +++++
3 files changed, 8 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 4197001f9801..880adc810ccc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -836,6 +836,7 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv,
out_priv = netdev_priv(encap_dev);
rpriv = out_priv->ppriv;
attr->out_rep = rpriv->rep;
+ attr->out_mdev = out_priv->mdev;
}
err = mlx5_eswitch_add_vlan_action(esw, attr);
@@ -2501,6 +2502,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
out_priv = netdev_priv(out_dev);
rpriv = out_priv->ppriv;
attr->out_rep = rpriv->rep;
+ attr->out_mdev = out_priv->mdev;
} else if (encap) {
parse_attr->mirred_ifindex = out_dev->ifindex;
parse_attr->tun_info = *info;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index 4cd773fa55e3..ac5db54823a1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -230,6 +230,7 @@ enum {
struct mlx5_esw_flow_attr {
struct mlx5_eswitch_rep *in_rep;
struct mlx5_eswitch_rep *out_rep;
+ struct mlx5_core_dev *out_mdev;
int action;
__be16 vlan_proto;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 90c8cb31e633..ea93867d1ab4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -72,6 +72,11 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw,
if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) {
dest[i].type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
dest[i].vport.num = attr->out_rep->vport;
+ if (MLX5_CAP_ESW(esw->dev, merged_eswitch)) {
+ dest[i].vport.vhca_id =
+ MLX5_CAP_GEN(attr->out_mdev, vhca_id);
+ dest[i].vport.vhca_id_valid = 1;
+ }
i++;
}
if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_COUNT) {
--
2.17.0
^ permalink raw reply related
* [for-next 09/15] net/mlx5e: Explicitly set source e-switch in offloaded TC rules
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Shahar Klein, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Shahar Klein <shahark@mellanox.com>
Set a specific source e-switch when setting a rule that matches on the
ingress port.
Signed-off-by: Shahar Klein <shahark@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 1 +
drivers/net/ethernet/mellanox/mlx5/core/eswitch.h | 1 +
.../net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 8 ++++++++
3 files changed, 10 insertions(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 880adc810ccc..1d2ba687b902 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -2462,6 +2462,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
memset(attr, 0, sizeof(*attr));
attr->in_rep = rpriv->rep;
+ attr->in_mdev = priv->mdev;
tcf_exts_to_list(exts, &actions);
list_for_each_entry(a, &actions, list) {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index ac5db54823a1..98a306e02640 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -231,6 +231,7 @@ struct mlx5_esw_flow_attr {
struct mlx5_eswitch_rep *in_rep;
struct mlx5_eswitch_rep *out_rep;
struct mlx5_core_dev *out_mdev;
+ struct mlx5_core_dev *in_mdev;
int action;
__be16 vlan_proto;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index ea93867d1ab4..6c83eef5141a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -93,8 +93,16 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw,
misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters);
MLX5_SET(fte_match_set_misc, misc, source_port, attr->in_rep->vport);
+ if (MLX5_CAP_ESW(esw->dev, merged_eswitch))
+ MLX5_SET(fte_match_set_misc, misc,
+ source_eswitch_owner_vhca_id,
+ MLX5_CAP_GEN(attr->in_mdev, vhca_id));
+
misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters);
MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
+ if (MLX5_CAP_ESW(esw->dev, merged_eswitch))
+ MLX5_SET_TO_ONES(fte_match_set_misc, misc,
+ source_eswitch_owner_vhca_id);
spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS |
MLX5_MATCH_MISC_PARAMETERS;
--
2.17.0
^ permalink raw reply related
* [for-next 10/15] net/mlx5e: Offload TC eswitch rules for VFs belonging to different PFs
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Rabie Loulou, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Rabie Loulou <rabiel@mellanox.com>
When the merged eswitch capability is supported, allow offloading rules
between VFs which belong to different PFs (and hence have different
eswitch affinity).
Signed-off-by: Rabie Loulou <rabiel@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Shahar Klein <shahark@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 17 ++++++++++++++++-
1 file changed, 16 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 630dd6dcabb9..77c3f8b8ae96 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -2077,6 +2077,20 @@ static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv,
return 0;
}
+static bool is_merged_eswitch_dev(struct mlx5e_priv *priv,
+ struct net_device *peer_netdev)
+{
+ struct mlx5e_priv *peer_priv;
+
+ peer_priv = netdev_priv(peer_netdev);
+
+ return (MLX5_CAP_ESW(priv->mdev, merged_eswitch) &&
+ (priv->netdev->netdev_ops == peer_netdev->netdev_ops) &&
+ same_hw_devs(priv, peer_priv) &&
+ MLX5_VPORT_MANAGER(peer_priv->mdev) &&
+ (peer_priv->mdev->priv.eswitch->mode == SRIOV_OFFLOADS));
+}
+
static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv,
struct net_device *mirred_dev,
struct net_device **out_dev,
@@ -2535,7 +2549,8 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
out_dev = tcf_mirred_dev(a);
if (switchdev_port_same_parent_id(priv->netdev,
- out_dev)) {
+ out_dev) ||
+ is_merged_eswitch_dev(priv, out_dev)) {
action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
MLX5_FLOW_CONTEXT_ACTION_COUNT;
out_priv = netdev_priv(out_dev);
--
2.17.0
^ permalink raw reply related
* [for-next 11/15] net/mlx5e: Add ingress/egress indication for offloaded TC flows
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Jiri Pirko, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Or Gerlitz <ogerlitz@mellanox.com>
When an e-switch TC rule is offloaded through the egdev (egress
device) mechanism, we treat this as egress, all other cases (NIC
and e-switch) are considred ingress.
This is preparation step that will allow us to identify "wrong"
stat/del offload calls made by the TC core on egdev based flows and
ignore them.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 3 --
.../net/ethernet/mellanox/mlx5/core/en_main.c | 15 ++++----
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 32 ++++++++++++----
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 38 ++++++++++++++-----
.../net/ethernet/mellanox/mlx5/core/en_tc.h | 13 +++++--
5 files changed, 70 insertions(+), 31 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 7c930088e96e..51a1d36a56c5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -1118,9 +1118,6 @@ int mlx5e_ethtool_get_ts_info(struct mlx5e_priv *priv,
int mlx5e_ethtool_flash_device(struct mlx5e_priv *priv,
struct ethtool_flash *flash);
-int mlx5e_setup_tc_block_cb(enum tc_setup_type type, void *type_data,
- void *cb_priv);
-
/* mlx5e generic netdev management API */
struct net_device*
mlx5e_create_netdev(struct mlx5_core_dev *mdev, const struct mlx5e_profile *profile,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 417bf2e8ab85..27e8375a476b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -3136,22 +3136,23 @@ static int mlx5e_setup_tc_mqprio(struct net_device *netdev,
#ifdef CONFIG_MLX5_ESWITCH
static int mlx5e_setup_tc_cls_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *cls_flower)
+ struct tc_cls_flower_offload *cls_flower,
+ int flags)
{
switch (cls_flower->command) {
case TC_CLSFLOWER_REPLACE:
- return mlx5e_configure_flower(priv, cls_flower);
+ return mlx5e_configure_flower(priv, cls_flower, flags);
case TC_CLSFLOWER_DESTROY:
- return mlx5e_delete_flower(priv, cls_flower);
+ return mlx5e_delete_flower(priv, cls_flower, flags);
case TC_CLSFLOWER_STATS:
- return mlx5e_stats_flower(priv, cls_flower);
+ return mlx5e_stats_flower(priv, cls_flower, flags);
default:
return -EOPNOTSUPP;
}
}
-int mlx5e_setup_tc_block_cb(enum tc_setup_type type, void *type_data,
- void *cb_priv)
+static int mlx5e_setup_tc_block_cb(enum tc_setup_type type, void *type_data,
+ void *cb_priv)
{
struct mlx5e_priv *priv = cb_priv;
@@ -3160,7 +3161,7 @@ int mlx5e_setup_tc_block_cb(enum tc_setup_type type, void *type_data,
switch (type) {
case TC_SETUP_CLSFLOWER:
- return mlx5e_setup_tc_cls_flower(priv, type_data);
+ return mlx5e_setup_tc_cls_flower(priv, type_data, MLX5E_TC_INGRESS);
default:
return -EOPNOTSUPP;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index a689f4c90fe3..182b636552a6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -723,15 +723,31 @@ static int mlx5e_rep_get_phys_port_name(struct net_device *dev,
static int
mlx5e_rep_setup_tc_cls_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *cls_flower)
+ struct tc_cls_flower_offload *cls_flower, int flags)
{
switch (cls_flower->command) {
case TC_CLSFLOWER_REPLACE:
- return mlx5e_configure_flower(priv, cls_flower);
+ return mlx5e_configure_flower(priv, cls_flower, flags);
case TC_CLSFLOWER_DESTROY:
- return mlx5e_delete_flower(priv, cls_flower);
+ return mlx5e_delete_flower(priv, cls_flower, flags);
case TC_CLSFLOWER_STATS:
- return mlx5e_stats_flower(priv, cls_flower);
+ return mlx5e_stats_flower(priv, cls_flower, flags);
+ default:
+ return -EOPNOTSUPP;
+ }
+}
+
+static int mlx5e_rep_setup_tc_cb_egdev(enum tc_setup_type type, void *type_data,
+ void *cb_priv)
+{
+ struct mlx5e_priv *priv = cb_priv;
+
+ if (!tc_cls_can_offload_and_chain0(priv->netdev, type_data))
+ return -EOPNOTSUPP;
+
+ switch (type) {
+ case TC_SETUP_CLSFLOWER:
+ return mlx5e_rep_setup_tc_cls_flower(priv, type_data, MLX5E_TC_EGRESS);
default:
return -EOPNOTSUPP;
}
@@ -747,7 +763,7 @@ static int mlx5e_rep_setup_tc_cb(enum tc_setup_type type, void *type_data,
switch (type) {
case TC_SETUP_CLSFLOWER:
- return mlx5e_rep_setup_tc_cls_flower(priv, type_data);
+ return mlx5e_rep_setup_tc_cls_flower(priv, type_data, MLX5E_TC_INGRESS);
default:
return -EOPNOTSUPP;
}
@@ -1107,7 +1123,7 @@ mlx5e_vport_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
uplink_rpriv = mlx5_eswitch_get_uplink_priv(dev->priv.eswitch, REP_ETH);
upriv = netdev_priv(uplink_rpriv->netdev);
- err = tc_setup_cb_egdev_register(netdev, mlx5e_setup_tc_block_cb,
+ err = tc_setup_cb_egdev_register(netdev, mlx5e_rep_setup_tc_cb_egdev,
upriv);
if (err)
goto err_neigh_cleanup;
@@ -1122,7 +1138,7 @@ mlx5e_vport_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
return 0;
err_egdev_cleanup:
- tc_setup_cb_egdev_unregister(netdev, mlx5e_setup_tc_block_cb,
+ tc_setup_cb_egdev_unregister(netdev, mlx5e_rep_setup_tc_cb_egdev,
upriv);
err_neigh_cleanup:
@@ -1151,7 +1167,7 @@ mlx5e_vport_rep_unload(struct mlx5_eswitch_rep *rep)
uplink_rpriv = mlx5_eswitch_get_uplink_priv(priv->mdev->priv.eswitch,
REP_ETH);
upriv = netdev_priv(uplink_rpriv->netdev);
- tc_setup_cb_egdev_unregister(netdev, mlx5e_setup_tc_block_cb,
+ tc_setup_cb_egdev_unregister(netdev, mlx5e_rep_setup_tc_cb_egdev,
upriv);
mlx5e_rep_neigh_cleanup(rpriv);
mlx5e_detach_netdev(priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 77c3f8b8ae96..26a1312ec9f8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -62,12 +62,16 @@ struct mlx5_nic_flow_attr {
struct mlx5_flow_table *hairpin_ft;
};
+#define MLX5E_TC_FLOW_BASE (MLX5E_TC_LAST_EXPORTED_BIT + 1)
+
enum {
- MLX5E_TC_FLOW_ESWITCH = BIT(0),
- MLX5E_TC_FLOW_NIC = BIT(1),
- MLX5E_TC_FLOW_OFFLOADED = BIT(2),
- MLX5E_TC_FLOW_HAIRPIN = BIT(3),
- MLX5E_TC_FLOW_HAIRPIN_RSS = BIT(4),
+ MLX5E_TC_FLOW_INGRESS = MLX5E_TC_INGRESS,
+ MLX5E_TC_FLOW_EGRESS = MLX5E_TC_EGRESS,
+ MLX5E_TC_FLOW_ESWITCH = BIT(MLX5E_TC_FLOW_BASE),
+ MLX5E_TC_FLOW_NIC = BIT(MLX5E_TC_FLOW_BASE + 1),
+ MLX5E_TC_FLOW_OFFLOADED = BIT(MLX5E_TC_FLOW_BASE + 2),
+ MLX5E_TC_FLOW_HAIRPIN = BIT(MLX5E_TC_FLOW_BASE + 3),
+ MLX5E_TC_FLOW_HAIRPIN_RSS = BIT(MLX5E_TC_FLOW_BASE + 4),
};
struct mlx5e_tc_flow {
@@ -2618,8 +2622,20 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
return 0;
}
+static void get_flags(int flags, u8 *flow_flags)
+{
+ u8 __flow_flags = 0;
+
+ if (flags & MLX5E_TC_INGRESS)
+ __flow_flags |= MLX5E_TC_FLOW_INGRESS;
+ if (flags & MLX5E_TC_EGRESS)
+ __flow_flags |= MLX5E_TC_FLOW_EGRESS;
+
+ *flow_flags = __flow_flags;
+}
+
int mlx5e_configure_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *f)
+ struct tc_cls_flower_offload *f, int flags)
{
struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
struct mlx5e_tc_flow_parse_attr *parse_attr;
@@ -2628,11 +2644,13 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
int attr_size, err = 0;
u8 flow_flags = 0;
+ get_flags(flags, &flow_flags);
+
if (esw && esw->mode == SRIOV_OFFLOADS) {
- flow_flags = MLX5E_TC_FLOW_ESWITCH;
+ flow_flags |= MLX5E_TC_FLOW_ESWITCH;
attr_size = sizeof(struct mlx5_esw_flow_attr);
} else {
- flow_flags = MLX5E_TC_FLOW_NIC;
+ flow_flags |= MLX5E_TC_FLOW_NIC;
attr_size = sizeof(struct mlx5_nic_flow_attr);
}
@@ -2691,7 +2709,7 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
}
int mlx5e_delete_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *f)
+ struct tc_cls_flower_offload *f, int flags)
{
struct mlx5e_tc_flow *flow;
struct mlx5e_tc_table *tc = &priv->fs.tc;
@@ -2711,7 +2729,7 @@ int mlx5e_delete_flower(struct mlx5e_priv *priv,
}
int mlx5e_stats_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *f)
+ struct tc_cls_flower_offload *f, int flags)
{
struct mlx5e_tc_table *tc = &priv->fs.tc;
struct mlx5e_tc_flow *flow;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
index c14c263a739b..2255345c2e18 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
@@ -38,16 +38,23 @@
#define MLX5E_TC_FLOW_ID_MASK 0x0000ffff
#ifdef CONFIG_MLX5_ESWITCH
+
+enum {
+ MLX5E_TC_INGRESS = BIT(0),
+ MLX5E_TC_EGRESS = BIT(1),
+ MLX5E_TC_LAST_EXPORTED_BIT = 1,
+};
+
int mlx5e_tc_init(struct mlx5e_priv *priv);
void mlx5e_tc_cleanup(struct mlx5e_priv *priv);
int mlx5e_configure_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *f);
+ struct tc_cls_flower_offload *f, int flags);
int mlx5e_delete_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *f);
+ struct tc_cls_flower_offload *f, int flags);
int mlx5e_stats_flower(struct mlx5e_priv *priv,
- struct tc_cls_flower_offload *f);
+ struct tc_cls_flower_offload *f, int flags);
struct mlx5e_encap_entry;
void mlx5e_tc_encap_flows_add(struct mlx5e_priv *priv,
--
2.17.0
^ permalink raw reply related
* [for-next 12/15] net/mlx5e: Prepare for shared table to keep TC eswitch flows
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Jiri Pirko, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Or Gerlitz <ogerlitz@mellanox.com>
This is a refactoring step to be able and store the hash table which
keeps track of offloaded TC flows in a different location for NIC
vs e-switch rules.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 -
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 39 ++++++++++---------
2 files changed, 20 insertions(+), 20 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 51a1d36a56c5..bc91a7335c93 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -634,7 +634,6 @@ struct mlx5e_flow_table {
struct mlx5e_tc_table {
struct mlx5_flow_table *t;
- struct rhashtable_params ht_params;
struct rhashtable ht;
DECLARE_HASHTABLE(mod_hdr_tbl, 8);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 26a1312ec9f8..1c90586d7f58 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -2634,12 +2634,24 @@ static void get_flags(int flags, u8 *flow_flags)
*flow_flags = __flow_flags;
}
+static const struct rhashtable_params tc_ht_params = {
+ .head_offset = offsetof(struct mlx5e_tc_flow, node),
+ .key_offset = offsetof(struct mlx5e_tc_flow, cookie),
+ .key_len = sizeof(((struct mlx5e_tc_flow *)0)->cookie),
+ .automatic_shrinking = true,
+};
+
+static struct rhashtable *get_tc_ht(struct mlx5e_priv *priv)
+{
+ return &priv->fs.tc.ht;
+}
+
int mlx5e_configure_flower(struct mlx5e_priv *priv,
struct tc_cls_flower_offload *f, int flags)
{
struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
struct mlx5e_tc_flow_parse_attr *parse_attr;
- struct mlx5e_tc_table *tc = &priv->fs.tc;
+ struct rhashtable *tc_ht = get_tc_ht(priv);
struct mlx5e_tc_flow *flow;
int attr_size, err = 0;
u8 flow_flags = 0;
@@ -2693,8 +2705,7 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
!(flow->esw_attr->action & MLX5_FLOW_CONTEXT_ACTION_ENCAP))
kvfree(parse_attr);
- err = rhashtable_insert_fast(&tc->ht, &flow->node,
- tc->ht_params);
+ err = rhashtable_insert_fast(tc_ht, &flow->node, tc_ht_params);
if (err) {
mlx5e_tc_del_flow(priv, flow);
kfree(flow);
@@ -2711,15 +2722,14 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
int mlx5e_delete_flower(struct mlx5e_priv *priv,
struct tc_cls_flower_offload *f, int flags)
{
+ struct rhashtable *tc_ht = get_tc_ht(priv);
struct mlx5e_tc_flow *flow;
- struct mlx5e_tc_table *tc = &priv->fs.tc;
- flow = rhashtable_lookup_fast(&tc->ht, &f->cookie,
- tc->ht_params);
+ flow = rhashtable_lookup_fast(tc_ht, &f->cookie, tc_ht_params);
if (!flow)
return -EINVAL;
- rhashtable_remove_fast(&tc->ht, &flow->node, tc->ht_params);
+ rhashtable_remove_fast(tc_ht, &flow->node, tc_ht_params);
mlx5e_tc_del_flow(priv, flow);
@@ -2731,15 +2741,14 @@ int mlx5e_delete_flower(struct mlx5e_priv *priv,
int mlx5e_stats_flower(struct mlx5e_priv *priv,
struct tc_cls_flower_offload *f, int flags)
{
- struct mlx5e_tc_table *tc = &priv->fs.tc;
+ struct rhashtable *tc_ht = get_tc_ht(priv);
struct mlx5e_tc_flow *flow;
struct mlx5_fc *counter;
u64 bytes;
u64 packets;
u64 lastuse;
- flow = rhashtable_lookup_fast(&tc->ht, &f->cookie,
- tc->ht_params);
+ flow = rhashtable_lookup_fast(tc_ht, &f->cookie, tc_ht_params);
if (!flow)
return -EINVAL;
@@ -2757,13 +2766,6 @@ int mlx5e_stats_flower(struct mlx5e_priv *priv,
return 0;
}
-static const struct rhashtable_params mlx5e_tc_flow_ht_params = {
- .head_offset = offsetof(struct mlx5e_tc_flow, node),
- .key_offset = offsetof(struct mlx5e_tc_flow, cookie),
- .key_len = sizeof(((struct mlx5e_tc_flow *)0)->cookie),
- .automatic_shrinking = true,
-};
-
int mlx5e_tc_init(struct mlx5e_priv *priv)
{
struct mlx5e_tc_table *tc = &priv->fs.tc;
@@ -2771,8 +2773,7 @@ int mlx5e_tc_init(struct mlx5e_priv *priv)
hash_init(tc->mod_hdr_tbl);
hash_init(tc->hairpin_tbl);
- tc->ht_params = mlx5e_tc_flow_ht_params;
- return rhashtable_init(&tc->ht, &tc->ht_params);
+ return rhashtable_init(&tc->ht, &tc_ht_params);
}
static void _mlx5e_tc_del_flow(void *ptr, void *arg)
--
2.17.0
^ permalink raw reply related
* [for-next 13/15] net/mlx5e: Use shared table for offloaded TC eswitch flows
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Jiri Pirko, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Or Gerlitz <ogerlitz@mellanox.com>
Currently, each representor netdev use their own hash table to keep
the mapping from TC flow (f->cookie) to the driver offloaded instance.
The table is the one which originally was added for offloading TC NIC
(not eswitch) rules.
This scheme breaks when the core TC code calls us to add the same flow
twice, (e.g under egdev use case) since we don't spot that and offload
a 2nd flow into the HW with the wrong source vport.
As a pre-step to solve that, we move to use a single table which keeps
all offloaded TC eswitch flows. The table is located at the eswitch
uplink representor object.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en_main.c | 4 +--
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 19 ++++++------
.../net/ethernet/mellanox/mlx5/core/en_rep.h | 1 +
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 29 +++++++++++++++----
.../net/ethernet/mellanox/mlx5/core/en_tc.h | 11 ++++---
5 files changed, 43 insertions(+), 21 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 27e8375a476b..b5a7580b12fe 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -4462,7 +4462,7 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv)
goto err_destroy_direct_tirs;
}
- err = mlx5e_tc_init(priv);
+ err = mlx5e_tc_nic_init(priv);
if (err)
goto err_destroy_flow_steering;
@@ -4483,7 +4483,7 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv)
static void mlx5e_cleanup_nic_rx(struct mlx5e_priv *priv)
{
- mlx5e_tc_cleanup(priv);
+ mlx5e_tc_nic_cleanup(priv);
mlx5e_destroy_flow_steering(priv);
mlx5e_destroy_direct_tirs(priv);
mlx5e_destroy_indirect_tirs(priv);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 182b636552a6..aa32592a54cb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -981,14 +981,8 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv)
}
rpriv->vport_rx_rule = flow_rule;
- err = mlx5e_tc_init(priv);
- if (err)
- goto err_del_flow_rule;
-
return 0;
-err_del_flow_rule:
- mlx5_del_flow_rules(rpriv->vport_rx_rule);
err_destroy_direct_tirs:
mlx5e_destroy_direct_tirs(priv);
err_destroy_direct_rqts:
@@ -1000,7 +994,6 @@ static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv)
{
struct mlx5e_rep_priv *rpriv = priv->ppriv;
- mlx5e_tc_cleanup(priv);
mlx5_del_flow_rules(rpriv->vport_rx_rule);
mlx5e_destroy_direct_tirs(priv);
mlx5e_destroy_direct_rqts(priv);
@@ -1058,8 +1051,15 @@ mlx5e_nic_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
if (err)
goto err_remove_sqs;
+ /* init shared tc flow table */
+ err = mlx5e_tc_esw_init(&rpriv->tc_ht);
+ if (err)
+ goto err_neigh_cleanup;
+
return 0;
+err_neigh_cleanup:
+ mlx5e_rep_neigh_cleanup(rpriv);
err_remove_sqs:
mlx5e_remove_sqs_fwd_rules(priv);
return err;
@@ -1074,9 +1074,8 @@ mlx5e_nic_rep_unload(struct mlx5_eswitch_rep *rep)
if (test_bit(MLX5E_STATE_OPENED, &priv->state))
mlx5e_remove_sqs_fwd_rules(priv);
- /* clean (and re-init) existing uplink offloaded TC rules */
- mlx5e_tc_cleanup(priv);
- mlx5e_tc_init(priv);
+ /* clean uplink offloaded TC rules, delete shared tc flow table */
+ mlx5e_tc_esw_cleanup(&rpriv->tc_ht);
mlx5e_rep_neigh_cleanup(rpriv);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
index b9b481f2833a..844d32d5c29f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
@@ -59,6 +59,7 @@ struct mlx5e_rep_priv {
struct net_device *netdev;
struct mlx5_flow_handle *vport_rx_rule;
struct list_head vport_sqs_list;
+ struct rhashtable tc_ht; /* valid for uplink rep */
};
static inline
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 1c90586d7f58..05c90b4f8a31 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -76,6 +76,7 @@ enum {
struct mlx5e_tc_flow {
struct rhash_head node;
+ struct mlx5e_priv *priv;
u64 cookie;
u8 flags;
struct mlx5_flow_handle *rule;
@@ -2643,7 +2644,14 @@ static const struct rhashtable_params tc_ht_params = {
static struct rhashtable *get_tc_ht(struct mlx5e_priv *priv)
{
- return &priv->fs.tc.ht;
+ struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
+ struct mlx5e_rep_priv *uplink_rpriv;
+
+ if (MLX5_VPORT_MANAGER(priv->mdev) && esw->mode == SRIOV_OFFLOADS) {
+ uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH);
+ return &uplink_rpriv->tc_ht;
+ } else
+ return &priv->fs.tc.ht;
}
int mlx5e_configure_flower(struct mlx5e_priv *priv,
@@ -2675,6 +2683,7 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
flow->cookie = f->cookie;
flow->flags = flow_flags;
+ flow->priv = priv;
err = parse_cls_flower(priv, flow, &parse_attr->spec, f);
if (err < 0)
@@ -2766,7 +2775,7 @@ int mlx5e_stats_flower(struct mlx5e_priv *priv,
return 0;
}
-int mlx5e_tc_init(struct mlx5e_priv *priv)
+int mlx5e_tc_nic_init(struct mlx5e_priv *priv)
{
struct mlx5e_tc_table *tc = &priv->fs.tc;
@@ -2779,20 +2788,30 @@ int mlx5e_tc_init(struct mlx5e_priv *priv)
static void _mlx5e_tc_del_flow(void *ptr, void *arg)
{
struct mlx5e_tc_flow *flow = ptr;
- struct mlx5e_priv *priv = arg;
+ struct mlx5e_priv *priv = flow->priv;
mlx5e_tc_del_flow(priv, flow);
kfree(flow);
}
-void mlx5e_tc_cleanup(struct mlx5e_priv *priv)
+void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv)
{
struct mlx5e_tc_table *tc = &priv->fs.tc;
- rhashtable_free_and_destroy(&tc->ht, _mlx5e_tc_del_flow, priv);
+ rhashtable_free_and_destroy(&tc->ht, _mlx5e_tc_del_flow, NULL);
if (!IS_ERR_OR_NULL(tc->t)) {
mlx5_destroy_flow_table(tc->t);
tc->t = NULL;
}
}
+
+int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
+{
+ return rhashtable_init(tc_ht, &tc_ht_params);
+}
+
+void mlx5e_tc_esw_cleanup(struct rhashtable *tc_ht)
+{
+ rhashtable_free_and_destroy(tc_ht, _mlx5e_tc_del_flow, NULL);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
index 2255345c2e18..59e52b845beb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
@@ -45,8 +45,11 @@ enum {
MLX5E_TC_LAST_EXPORTED_BIT = 1,
};
-int mlx5e_tc_init(struct mlx5e_priv *priv);
-void mlx5e_tc_cleanup(struct mlx5e_priv *priv);
+int mlx5e_tc_nic_init(struct mlx5e_priv *priv);
+void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv);
+
+int mlx5e_tc_esw_init(struct rhashtable *tc_ht);
+void mlx5e_tc_esw_cleanup(struct rhashtable *tc_ht);
int mlx5e_configure_flower(struct mlx5e_priv *priv,
struct tc_cls_flower_offload *f, int flags);
@@ -71,8 +74,8 @@ static inline int mlx5e_tc_num_filters(struct mlx5e_priv *priv)
}
#else /* CONFIG_MLX5_ESWITCH */
-static inline int mlx5e_tc_init(struct mlx5e_priv *priv) { return 0; }
-static inline void mlx5e_tc_cleanup(struct mlx5e_priv *priv) {}
+static inline int mlx5e_tc_nic_init(struct mlx5e_priv *priv) { return 0; }
+static inline void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv) {}
static inline int mlx5e_tc_num_filters(struct mlx5e_priv *priv) { return 0; }
#endif
--
2.17.0
^ permalink raw reply related
* [for-next 14/15] net/mlx5e: Ignore attempts to offload multiple times a TC flow
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Jiri Pirko, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Or Gerlitz <ogerlitz@mellanox.com>
For VF->VF and uplink->VF rules, the TC core (cls_api) attempts
to offload the same flow multiple times into the driver, b/c we
registered to the egdev callback.
Use the flow cookie to ignore attempts to add such flows, we can't
reject them (return error), b/c this will fail the offload attempt,
so we ignore that. We indentify wrong stat/del calls using the flow
ingress/egress flags, here we do return error to the core.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 21 +++++++++++++++++--
1 file changed, 19 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 05c90b4f8a31..674f1d7d2737 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -2666,6 +2666,12 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
get_flags(flags, &flow_flags);
+ flow = rhashtable_lookup_fast(tc_ht, &f->cookie, tc_ht_params);
+ if (flow) {
+ netdev_warn_once(priv->netdev, "flow cookie %lx already exists, ignoring\n", f->cookie);
+ return 0;
+ }
+
if (esw && esw->mode == SRIOV_OFFLOADS) {
flow_flags |= MLX5E_TC_FLOW_ESWITCH;
attr_size = sizeof(struct mlx5_esw_flow_attr);
@@ -2728,6 +2734,17 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
return err;
}
+#define DIRECTION_MASK (MLX5E_TC_INGRESS | MLX5E_TC_EGRESS)
+#define FLOW_DIRECTION_MASK (MLX5E_TC_FLOW_INGRESS | MLX5E_TC_FLOW_EGRESS)
+
+static bool same_flow_direction(struct mlx5e_tc_flow *flow, int flags)
+{
+ if ((flow->flags & FLOW_DIRECTION_MASK) == (flags & DIRECTION_MASK))
+ return true;
+
+ return false;
+}
+
int mlx5e_delete_flower(struct mlx5e_priv *priv,
struct tc_cls_flower_offload *f, int flags)
{
@@ -2735,7 +2752,7 @@ int mlx5e_delete_flower(struct mlx5e_priv *priv,
struct mlx5e_tc_flow *flow;
flow = rhashtable_lookup_fast(tc_ht, &f->cookie, tc_ht_params);
- if (!flow)
+ if (!flow || !same_flow_direction(flow, flags))
return -EINVAL;
rhashtable_remove_fast(tc_ht, &flow->node, tc_ht_params);
@@ -2758,7 +2775,7 @@ int mlx5e_stats_flower(struct mlx5e_priv *priv,
u64 lastuse;
flow = rhashtable_lookup_fast(tc_ht, &f->cookie, tc_ht_params);
- if (!flow)
+ if (!flow || !same_flow_direction(flow, flags))
return -EINVAL;
if (!(flow->flags & MLX5E_TC_FLOW_OFFLOADED))
--
2.17.0
^ permalink raw reply related
* [for-next 15/15] net/mlx5e: Add HW vport counters to representor ethtool stats
From: Saeed Mahameed @ 2018-05-18 1:22 UTC (permalink / raw)
To: David S. Miller, Doug Ledford
Cc: Jason Gunthorpe, Leon Romanovsky, Or Gerlitz, netdev, linux-rdma,
Adi Nissim, Saeed Mahameed
In-Reply-To: <20180518012258.26968-1-saeedm@mellanox.com>
From: Or Gerlitz <ogerlitz@mellanox.com>
Currently the representor only report the SW (slow-path) traffic
counters.
Add packet/bytes reporting of the HW counters, which account for the
total amount of traffic that was handled by the vport, both slow and
fast (offloaded) paths. The newly exposed counters are named
vport_rx/tx_packets/bytes.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Adi Nissim <adin@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 35 +++++++++++++++----
1 file changed, 29 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index aa32592a54cb..c3034f58aa33 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -66,18 +66,36 @@ static const struct counter_desc sw_rep_stats_desc[] = {
{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_bytes) },
};
-#define NUM_VPORT_REP_COUNTERS ARRAY_SIZE(sw_rep_stats_desc)
+struct vport_stats {
+ u64 vport_rx_packets;
+ u64 vport_tx_packets;
+ u64 vport_rx_bytes;
+ u64 vport_tx_bytes;
+};
+
+static const struct counter_desc vport_rep_stats_desc[] = {
+ { MLX5E_DECLARE_STAT(struct vport_stats, vport_rx_packets) },
+ { MLX5E_DECLARE_STAT(struct vport_stats, vport_rx_bytes) },
+ { MLX5E_DECLARE_STAT(struct vport_stats, vport_tx_packets) },
+ { MLX5E_DECLARE_STAT(struct vport_stats, vport_tx_bytes) },
+};
+
+#define NUM_VPORT_REP_SW_COUNTERS ARRAY_SIZE(sw_rep_stats_desc)
+#define NUM_VPORT_REP_HW_COUNTERS ARRAY_SIZE(vport_rep_stats_desc)
static void mlx5e_rep_get_strings(struct net_device *dev,
u32 stringset, uint8_t *data)
{
- int i;
+ int i, j;
switch (stringset) {
case ETH_SS_STATS:
- for (i = 0; i < NUM_VPORT_REP_COUNTERS; i++)
+ for (i = 0; i < NUM_VPORT_REP_SW_COUNTERS; i++)
strcpy(data + (i * ETH_GSTRING_LEN),
sw_rep_stats_desc[i].format);
+ for (j = 0; j < NUM_VPORT_REP_HW_COUNTERS; j++, i++)
+ strcpy(data + (i * ETH_GSTRING_LEN),
+ vport_rep_stats_desc[j].format);
break;
}
}
@@ -140,7 +158,7 @@ static void mlx5e_rep_get_ethtool_stats(struct net_device *dev,
struct ethtool_stats *stats, u64 *data)
{
struct mlx5e_priv *priv = netdev_priv(dev);
- int i;
+ int i, j;
if (!data)
return;
@@ -148,18 +166,23 @@ static void mlx5e_rep_get_ethtool_stats(struct net_device *dev,
mutex_lock(&priv->state_lock);
if (test_bit(MLX5E_STATE_OPENED, &priv->state))
mlx5e_rep_update_sw_counters(priv);
+ mlx5e_rep_update_hw_counters(priv);
mutex_unlock(&priv->state_lock);
- for (i = 0; i < NUM_VPORT_REP_COUNTERS; i++)
+ for (i = 0; i < NUM_VPORT_REP_SW_COUNTERS; i++)
data[i] = MLX5E_READ_CTR64_CPU(&priv->stats.sw,
sw_rep_stats_desc, i);
+
+ for (j = 0; j < NUM_VPORT_REP_HW_COUNTERS; j++, i++)
+ data[i] = MLX5E_READ_CTR64_CPU(&priv->stats.vf_vport,
+ vport_rep_stats_desc, j);
}
static int mlx5e_rep_get_sset_count(struct net_device *dev, int sset)
{
switch (sset) {
case ETH_SS_STATS:
- return NUM_VPORT_REP_COUNTERS;
+ return NUM_VPORT_REP_SW_COUNTERS + NUM_VPORT_REP_HW_COUNTERS;
default:
return -EOPNOTSUPP;
}
--
2.17.0
^ permalink raw reply related
* Re: [patch net-next RFC 04/12] dsa: set devlink port attrs for dsa ports
From: Andrew Lunn @ 2018-05-18 1:41 UTC (permalink / raw)
To: Florian Fainelli
Cc: Jiri Pirko, netdev, davem, idosch, jakub.kicinski, mlxsw,
vivien.didelot, michael.chan, ganeshgr, saeedm, simon.horman,
pieter.jansenvanvuuren, john.hurley, dirk.vandermerwe,
alexander.h.duyck, ogerlitz, dsahern, vijaya.guvva,
satananda.burla, raghu.vatsavayi, felix.manlunas, gospo,
sathya.perla, vasundhara-v.volam, tariqt, eranbe,
jeffrey.t.kirsher
In-Reply-To: <57d602af-2996-dde5-72b1-2d4f58a47032@gmail.com>
> >>> ds = dsa_switch_alloc(&mdiodev->dev, DSA_MAX_PORTS);
> >>>
> >>> It is allocating a switch with 12 ports. However only 4 of them have
> >>> names. So the core only creates slave devices for those 4.
> >>>
> >>> This is a useful test. Real hardware often has unused ports. A WiFi AP
> >>> with a 7 port switch which only uses 6 ports is often seen.
> >>
> >> The following patch should fix this:
> >>
> >>
> >> diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
> >> index adf50fbc4c13..a06c29ec91f0 100644
> >> --- a/net/dsa/dsa2.c
> >> +++ b/net/dsa/dsa2.c
> >> @@ -262,13 +262,14 @@ static int dsa_port_setup(struct dsa_port *dp)
> >>
> >> memset(&dp->devlink_port, 0, sizeof(dp->devlink_port));
> >>
> >> + if (dp->type == DSA_PORT_TYPE_UNUSED)
> >> + return 0;
> >> +
> >> err = devlink_port_register(ds->devlink, &dp->devlink_port,
> >> dp->index);
> >
> > Hi Florian, Jiri
> >
> > Maybe it is better to add a devlink port type unused?
>
> The port does not exist on the switch, so it should not even be
> registered IMHO.
Hi Florian
The ports do exist, when you called dsa_switch_alloc() you said the
switch has 12 ports.
Andrew
^ permalink raw reply
* Re: [Cake] [PATCH net-next v12 3/7] sch_cake: Add optional ACK filter
From: Ryan Mounce @ 2018-05-18 2:36 UTC (permalink / raw)
To: Toke Høiland-Jørgensen; +Cc: Eric Dumazet, netdev, Cake List
In-Reply-To: <877eo2xw8s.fsf@toke.dk>
On 17 May 2018 at 22:41, Toke Høiland-Jørgensen <toke@toke.dk> wrote:
> Eric Dumazet <eric.dumazet@gmail.com> writes:
>
>> On 05/17/2018 04:23 AM, Toke Høiland-Jørgensen wrote:
>>
>>>
>>> We don't do full parsing of SACKs, no; we were trying to keep things
>>> simple... We do detect the presence of SACK options, though, and the
>>> presence of SACK options on an ACK will make previous ACKs be considered
>>> redundant.
>>>
>>
>> But they are not redundant in some cases, particularly when reorders
>> happen in the network.
>
> Huh. I was under the impression that SACKs were basically cumulative
> until cleared.
>
> I.e., in packet sequence ABCDE where B and D are lost, C would have
> SACK(B) and E would have SACK(B,D). Are you saying that E would only
> have SACK(D)?
SACK works by acknowledging additional ranges above those that have
been ACKed, rather than ACKing up to the largest seen sequence number
and reporting missing ranges before that.
A - ACK(A)
B - lost
C - ACK(A) + SACK(C)
D - lost
E - ACK(A) + SACK(C, E)
Cake does check that the ACK sequence number is greater, or if it is
equal and the 'newer' ACK has the SACK option present. It doesn't
compare the sequence numbers inside two SACKs. If the two SACKs in the
above example had been reordered before reaching cake's ACK filter in
aggressive mode, the wrong one will be filtered.
This is a limitation of my naive SACK handling in cake. The default
'conservative' mode happens to mitigate the problem in the above
scenario, but the issue could still present itself in more
pathological cases. It's fixable, however I'm not sure this corner
case is sufficiently common or severe to warrant the extra complexity.
Ryan.
^ permalink raw reply
* [net-next:master 1230/1233] arch/sparc/include/asm/io_64.h:177:20: note: in expansion of macro 'writel'
From: kbuild test robot @ 2018-05-18 2:56 UTC (permalink / raw)
To: Florian Fainelli; +Cc: kbuild-all, netdev
[-- Attachment #1: Type: text/plain, Size: 8651 bytes --]
tree: https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git master
head: 538e2de104cfb4ef1acb35af42427bff42adbe4d
commit: 2652113ff043ca2ce1cb3be529b5ca9270c421d4 [1230/1233] net: ethernet: ti: Allow most drivers with COMPILE_TEST
config: sparc64-allyesconfig (attached as .config)
compiler: sparc64-linux-gnu-gcc (Debian 7.2.0-11) 7.2.0
reproduce:
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
git checkout 2652113ff043ca2ce1cb3be529b5ca9270c421d4
# save the attached .config to linux build tree
make.cross ARCH=sparc64
All warnings (new ones prefixed by >>):
In file included from arch/sparc/include/asm/bug.h:25:0,
from include/linux/bug.h:5,
from include/linux/thread_info.h:12,
from include/asm-generic/preempt.h:5,
from ./arch/sparc/include/generated/asm/preempt.h:1,
from include/linux/preempt.h:81,
from include/linux/spinlock.h:51,
from drivers/net/ethernet/ti/davinci_cpdma.c:16:
drivers/net/ethernet/ti/davinci_cpdma.c: In function 'cpdma_desc_pool_destroy':
drivers/net/ethernet/ti/davinci_cpdma.c:194:7: warning: format '%d' expects argument of type 'int', but argument 4 has type 'size_t {aka long unsigned int}' [-Wformat=]
"cpdma_desc_pool size %d != avail %d",
^
gen_pool_size(pool->gen_pool),
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/asm-generic/bug.h:92:69: note: in definition of macro '__WARN_printf'
#define __WARN_printf(arg...) warn_slowpath_fmt(__FILE__, __LINE__, arg)
^~~
drivers/net/ethernet/ti/davinci_cpdma.c:193:2: note: in expansion of macro 'WARN'
WARN(gen_pool_size(pool->gen_pool) != gen_pool_avail(pool->gen_pool),
^~~~
drivers/net/ethernet/ti/davinci_cpdma.c:194:7: warning: format '%d' expects argument of type 'int', but argument 5 has type 'size_t {aka long unsigned int}' [-Wformat=]
"cpdma_desc_pool size %d != avail %d",
^
drivers/net/ethernet/ti/davinci_cpdma.c:196:7:
gen_pool_avail(pool->gen_pool));
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/asm-generic/bug.h:92:69: note: in definition of macro '__WARN_printf'
#define __WARN_printf(arg...) warn_slowpath_fmt(__FILE__, __LINE__, arg)
^~~
drivers/net/ethernet/ti/davinci_cpdma.c:193:2: note: in expansion of macro 'WARN'
WARN(gen_pool_size(pool->gen_pool) != gen_pool_avail(pool->gen_pool),
^~~~
drivers/net/ethernet/ti/davinci_cpdma.c: In function 'cpdma_chan_submit':
drivers/net/ethernet/ti/davinci_cpdma.c:1083:17: warning: passing argument 1 of 'writel' makes integer from pointer without a cast [-Wint-conversion]
writel_relaxed(token, &desc->sw_token);
^~~~~
In file included from arch/sparc/include/asm/io.h:5:0,
from include/linux/scatterlist.h:9,
from include/linux/dma-mapping.h:11,
from drivers/net/ethernet/ti/davinci_cpdma.c:21:
arch/sparc/include/asm/io_64.h:175:16: note: expected 'u32 {aka unsigned int}' but argument is of type 'void *'
#define writel writel
^
>> arch/sparc/include/asm/io_64.h:177:20: note: in expansion of macro 'writel'
static inline void writel(u32 l, volatile void __iomem *addr)
^~~~~~
drivers/net/ethernet/ti/davinci_cpdma.c: In function '__cpdma_chan_free':
drivers/net/ethernet/ti/davinci_cpdma.c:1126:15: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
token = (void *)desc_read(desc, sw_token);
^
--
In file included from arch/sparc/include/asm/bug.h:25:0,
from include/linux/bug.h:5,
from include/linux/thread_info.h:12,
from include/asm-generic/preempt.h:5,
from ./arch/sparc/include/generated/asm/preempt.h:1,
from include/linux/preempt.h:81,
from include/linux/spinlock.h:51,
from drivers/net//ethernet/ti/davinci_cpdma.c:16:
drivers/net//ethernet/ti/davinci_cpdma.c: In function 'cpdma_desc_pool_destroy':
drivers/net//ethernet/ti/davinci_cpdma.c:194:7: warning: format '%d' expects argument of type 'int', but argument 4 has type 'size_t {aka long unsigned int}' [-Wformat=]
"cpdma_desc_pool size %d != avail %d",
^
gen_pool_size(pool->gen_pool),
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/asm-generic/bug.h:92:69: note: in definition of macro '__WARN_printf'
#define __WARN_printf(arg...) warn_slowpath_fmt(__FILE__, __LINE__, arg)
^~~
drivers/net//ethernet/ti/davinci_cpdma.c:193:2: note: in expansion of macro 'WARN'
WARN(gen_pool_size(pool->gen_pool) != gen_pool_avail(pool->gen_pool),
^~~~
drivers/net//ethernet/ti/davinci_cpdma.c:194:7: warning: format '%d' expects argument of type 'int', but argument 5 has type 'size_t {aka long unsigned int}' [-Wformat=]
"cpdma_desc_pool size %d != avail %d",
^
drivers/net//ethernet/ti/davinci_cpdma.c:196:7:
gen_pool_avail(pool->gen_pool));
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/asm-generic/bug.h:92:69: note: in definition of macro '__WARN_printf'
#define __WARN_printf(arg...) warn_slowpath_fmt(__FILE__, __LINE__, arg)
^~~
drivers/net//ethernet/ti/davinci_cpdma.c:193:2: note: in expansion of macro 'WARN'
WARN(gen_pool_size(pool->gen_pool) != gen_pool_avail(pool->gen_pool),
^~~~
drivers/net//ethernet/ti/davinci_cpdma.c: In function 'cpdma_chan_submit':
drivers/net//ethernet/ti/davinci_cpdma.c:1083:17: warning: passing argument 1 of 'writel' makes integer from pointer without a cast [-Wint-conversion]
writel_relaxed(token, &desc->sw_token);
^~~~~
In file included from arch/sparc/include/asm/io.h:5:0,
from include/linux/scatterlist.h:9,
from include/linux/dma-mapping.h:11,
from drivers/net//ethernet/ti/davinci_cpdma.c:21:
arch/sparc/include/asm/io_64.h:175:16: note: expected 'u32 {aka unsigned int}' but argument is of type 'void *'
#define writel writel
^
>> arch/sparc/include/asm/io_64.h:177:20: note: in expansion of macro 'writel'
static inline void writel(u32 l, volatile void __iomem *addr)
^~~~~~
drivers/net//ethernet/ti/davinci_cpdma.c: In function '__cpdma_chan_free':
drivers/net//ethernet/ti/davinci_cpdma.c:1126:15: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
token = (void *)desc_read(desc, sw_token);
^
vim +/writel +177 arch/sparc/include/asm/io_64.h
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 174
79294d7e arch/sparc/include/asm/io_64.h Sam Ravnborg 2014-07-20 @175 #define writel writel
7c3969c3 arch/sparc/include/asm/io_64.h Arnd Bergmann 2014-11-19 176 #define writel_relaxed writel
79294d7e arch/sparc/include/asm/io_64.h Sam Ravnborg 2014-07-20 @177 static inline void writel(u32 l, volatile void __iomem *addr)
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 178 {
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 179 __asm__ __volatile__("stwa\t%r0, [%1] %2\t/* pci_writel */"
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 180 : /* no outputs */
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 181 : "Jr" (l), "r" (addr), "i" (ASI_PHYS_BYPASS_EC_E_L)
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 182 : "memory");
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 183 }
f5e706ad include/asm-sparc/io_64.h Sam Ravnborg 2008-07-17 184
:::::: The code at line 177 was first introduced by commit
:::::: 79294d7eff89f65d2e58021bca81437e88fa61e3 sparc64: remove macro indirection in io_64.h
:::::: TO: Sam Ravnborg <sam@ravnborg.org>
:::::: CC: David S. Miller <davem@davemloft.net>
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 53378 bytes --]
^ permalink raw reply
* [PATCH v2] net: qcom/emac: Allocate buffers from local node
From: Hemanth Puranik @ 2018-05-18 3:29 UTC (permalink / raw)
To: netdev, linux-kernel; +Cc: Timur Tabi, Hemanth Puranik
Currently we use non-NUMA aware allocation for TPD and RRD buffers,
this patch modifies to use NUMA friendly allocation.
Signed-off-by: Hemanth Puranik <hpuranik@codeaurora.org>
---
Change since v1:
- Addressed comments related to ordering
drivers/net/ethernet/qualcomm/emac/emac-mac.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/qualcomm/emac/emac-mac.c b/drivers/net/ethernet/qualcomm/emac/emac-mac.c
index 092718a..031f6e6 100644
--- a/drivers/net/ethernet/qualcomm/emac/emac-mac.c
+++ b/drivers/net/ethernet/qualcomm/emac/emac-mac.c
@@ -683,10 +683,11 @@ static int emac_tx_q_desc_alloc(struct emac_adapter *adpt,
struct emac_tx_queue *tx_q)
{
struct emac_ring_header *ring_header = &adpt->ring_header;
+ int node = dev_to_node(adpt->netdev->dev.parent);
size_t size;
size = sizeof(struct emac_buffer) * tx_q->tpd.count;
- tx_q->tpd.tpbuff = kzalloc(size, GFP_KERNEL);
+ tx_q->tpd.tpbuff = kzalloc_node(size, GFP_KERNEL, node);
if (!tx_q->tpd.tpbuff)
return -ENOMEM;
@@ -723,11 +724,12 @@ static void emac_rx_q_bufs_free(struct emac_adapter *adpt)
static int emac_rx_descs_alloc(struct emac_adapter *adpt)
{
struct emac_ring_header *ring_header = &adpt->ring_header;
+ int node = dev_to_node(adpt->netdev->dev.parent);
struct emac_rx_queue *rx_q = &adpt->rx_q;
size_t size;
size = sizeof(struct emac_buffer) * rx_q->rfd.count;
- rx_q->rfd.rfbuff = kzalloc(size, GFP_KERNEL);
+ rx_q->rfd.rfbuff = kzalloc_node(size, GFP_KERNEL, node);
if (!rx_q->rfd.rfbuff)
return -ENOMEM;
--
Qualcomm Datacenter Technologies as an affiliate of Qualcomm Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the
Code Aurora Forum, a Linux Foundation Collaborative Project.
^ permalink raw reply related
* Re: pull-request: bpf 2018-05-18
From: David Miller @ 2018-05-18 3:34 UTC (permalink / raw)
To: daniel; +Cc: ast, netdev
In-Reply-To: <20180518002617.15231-1-daniel@iogearbox.net>
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Fri, 18 May 2018 02:26:17 +0200
> The following pull-request contains BPF updates for your *net* tree.
>
> The main changes are:
...
> Please consider pulling these changes from:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf.git
Pulled.
> When this gets later merged into net-next there are a two trivial
> BPF conflicts to resolve:
...
Thanks a lot for the conflict guidance.
^ permalink raw reply
* Re: [PATCH bpf-next v3 00/15] Introducing AF_XDP support
From: Alexei Starovoitov @ 2018-05-18 3:38 UTC (permalink / raw)
To: Björn Töpel, Alexei Starovoitov
Cc: Daniel Borkmann, Karlsson, Magnus, Duyck, Alexander H,
Alexander Duyck, John Fastabend, Jesper Dangaard Brouer,
Willem de Bruijn, Michael S. Tsirkin, Netdev,
Björn Töpel, michael.lundkvist, Brandeburg, Jesse,
Singhai, Anjali, Zhang, Qi Z
In-Reply-To: <CAJ+HfNin-ZFMqbVCVJLSvdiHEEUr8Q4gDLxY9EzGjr__rMW2Lg@mail.gmail.com>
On 5/16/18 11:46 PM, Björn Töpel wrote:
> 2018-05-04 1:38 GMT+02:00 Alexei Starovoitov <alexei.starovoitov@gmail.com>:
>> On Fri, May 04, 2018 at 12:49:09AM +0200, Daniel Borkmann wrote:
>>> On 05/02/2018 01:01 PM, Björn Töpel wrote:
>>>> From: Björn Töpel <bjorn.topel@intel.com>
>>>>
>>>> This patch set introduces a new address family called AF_XDP that is
>>>> optimized for high performance packet processing and, in upcoming
>>>> patch sets, zero-copy semantics. In this patch set, we have removed
>>>> all zero-copy related code in order to make it smaller, simpler and
>>>> hopefully more review friendly. This patch set only supports copy-mode
>>>> for the generic XDP path (XDP_SKB) for both RX and TX and copy-mode
>>>> for RX using the XDP_DRV path. Zero-copy support requires XDP and
>>>> driver changes that Jesper Dangaard Brouer is working on. Some of his
>>>> work has already been accepted. We will publish our zero-copy support
>>>> for RX and TX on top of his patch sets at a later point in time.
>>>
>>> +1, would be great to see it land this cycle. Saw few minor nits here
>>> and there but nothing to hold it up, for the series:
>>>
>>> Acked-by: Daniel Borkmann <daniel@iogearbox.net>
>>>
>>> Thanks everyone!
>>
>> Great stuff!
>>
>> Applied to bpf-next, with one condition.
>> Upcoming zero-copy patches for both RX and TX need to be posted
>> and reviewed within this release window.
>> If netdev community as a whole won't be able to agree on the zero-copy
>> bits we'd need to revert this feature before the next merge window.
>>
>> Few other minor nits:
>> patch 3:
>> +struct xdp_ring {
>> + __u32 producer __attribute__((aligned(64)));
>> + __u32 consumer __attribute__((aligned(64)));
>> +};
>> It kinda begs for ____cacheline_aligned_in_smp to be introduced for uapi headers.
>>
>
> Hmm, I need some guidance on what a sane uapi variant would be. We
> can't have the uapi depend on the kernel build. ARM64, e.g., can have
> both 64B and 128B according to the specs. Contemporary IA processors
> have 64B.
>
> The simplest, and maybe most future-proof, would be 128B aligned for
> all. Another is having 128B for ARM and 64B for all IA. A third option
> is having a hand-shaking API (I think virtio has that) for determine
> the cache line size, but I'd rather not go down that route.
>
> Thoughts/ideas on how a uapi ____cacheline_aligned_in_smp version
> would look like?
I suspect i40e+arm combination wasn't tested anyway.
The api may have endianness issues too on something like sparc.
I think the way to be backwards compatible in this area
is to make the api usable on x86 only by adding
to include/uapi/linux/if_xdp.h
#if defined(__x86_64__)
#define AF_XDP_CACHE_BYTES 64
#else
#error "AF_XDP support is not yet available for this architecture"
#endif
and doing:
__u32 producer __attribute__((aligned(AF_XDP_CACHE_BYTES)));
__u32 consumer __attribute__((aligned(AF_XDP_CACHE_BYTES)));
And progressively add to this for arm64 and few other archs.
Eventually removing #error and adding some generic define
that's good enough for long tail of architectures that
we really cannot test.
^ permalink raw reply
* Re: [net-next PATCH v2 2/4] net: Enable Tx queue selection based on Rx queues
From: Tom Herbert @ 2018-05-18 4:03 UTC (permalink / raw)
To: Amritha Nambiar
Cc: Linux Kernel Network Developers, David S. Miller, Alexander Duyck,
Sridhar Samudrala, Eric Dumazet, Hannes Frederic Sowa
In-Reply-To: <152643400925.4991.5029989601625953592.stgit@anamdev.jf.intel.com>
On Tue, May 15, 2018 at 6:26 PM, Amritha Nambiar
<amritha.nambiar@intel.com> wrote:
> This patch adds support to pick Tx queue based on the Rx queue map
> configuration set by the admin through the sysfs attribute
> for each Tx queue. If the user configuration for receive
> queue map does not apply, then the Tx queue selection falls back
> to CPU map based selection and finally to hashing.
>
> Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
> Signed-off-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
> ---
> include/net/sock.h | 18 ++++++++++++++++++
> net/core/dev.c | 36 +++++++++++++++++++++++++++++-------
> net/core/sock.c | 5 +++++
> net/ipv4/tcp_input.c | 7 +++++++
> net/ipv4/tcp_ipv4.c | 1 +
> net/ipv4/tcp_minisocks.c | 1 +
> 6 files changed, 61 insertions(+), 7 deletions(-)
>
> diff --git a/include/net/sock.h b/include/net/sock.h
> index 4f7c584..0613f63 100644
> --- a/include/net/sock.h
> +++ b/include/net/sock.h
> @@ -139,6 +139,8 @@ typedef __u64 __bitwise __addrpair;
> * @skc_node: main hash linkage for various protocol lookup tables
> * @skc_nulls_node: main hash linkage for TCP/UDP/UDP-Lite protocol
> * @skc_tx_queue_mapping: tx queue number for this connection
> + * @skc_rx_queue_mapping: rx queue number for this connection
> + * @skc_rx_ifindex: rx ifindex for this connection
> * @skc_flags: place holder for sk_flags
> * %SO_LINGER (l_onoff), %SO_BROADCAST, %SO_KEEPALIVE,
> * %SO_OOBINLINE settings, %SO_TIMESTAMPING settings
> @@ -215,6 +217,10 @@ struct sock_common {
> struct hlist_nulls_node skc_nulls_node;
> };
> int skc_tx_queue_mapping;
> +#ifdef CONFIG_XPS
> + int skc_rx_queue_mapping;
> + int skc_rx_ifindex;
Isn't this increasing size of sock_common for a narrow use case functionality?
> +#endif
> union {
> int skc_incoming_cpu;
> u32 skc_rcv_wnd;
> @@ -326,6 +332,10 @@ struct sock {
> #define sk_nulls_node __sk_common.skc_nulls_node
> #define sk_refcnt __sk_common.skc_refcnt
> #define sk_tx_queue_mapping __sk_common.skc_tx_queue_mapping
> +#ifdef CONFIG_XPS
> +#define sk_rx_queue_mapping __sk_common.skc_rx_queue_mapping
> +#define sk_rx_ifindex __sk_common.skc_rx_ifindex
> +#endif
>
> #define sk_dontcopy_begin __sk_common.skc_dontcopy_begin
> #define sk_dontcopy_end __sk_common.skc_dontcopy_end
> @@ -1696,6 +1706,14 @@ static inline int sk_tx_queue_get(const struct sock *sk)
> return sk ? sk->sk_tx_queue_mapping : -1;
> }
>
> +static inline void sk_mark_rx_queue(struct sock *sk, struct sk_buff *skb)
> +{
> +#ifdef CONFIG_XPS
> + sk->sk_rx_ifindex = skb->skb_iif;
> + sk->sk_rx_queue_mapping = skb_get_rx_queue(skb);
> +#endif
> +}
> +
> static inline void sk_set_socket(struct sock *sk, struct socket *sock)
> {
> sk_tx_queue_clear(sk);
> diff --git a/net/core/dev.c b/net/core/dev.c
> index 7e5dfdb..4030368 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -3458,18 +3458,14 @@ sch_handle_egress(struct sk_buff *skb, int *ret, struct net_device *dev)
> }
> #endif /* CONFIG_NET_EGRESS */
>
> -static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
> -{
> #ifdef CONFIG_XPS
> - struct xps_dev_maps *dev_maps;
> +static int __get_xps_queue_idx(struct net_device *dev, struct sk_buff *skb,
> + struct xps_dev_maps *dev_maps, unsigned int tci)
> +{
> struct xps_map *map;
> int queue_index = -1;
>
> - rcu_read_lock();
> - dev_maps = rcu_dereference(dev->xps_maps[XPS_MAP_CPUS]);
> if (dev_maps) {
> - unsigned int tci = skb->sender_cpu - 1;
> -
> if (dev->num_tc) {
> tci *= dev->num_tc;
> tci += netdev_get_prio_tc_map(dev, skb->priority);
> @@ -3486,6 +3482,32 @@ static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
> queue_index = -1;
> }
> }
> + return queue_index;
> +}
> +#endif
> +
> +static int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
> +{
> +#ifdef CONFIG_XPS
> + enum xps_map_type i = XPS_MAP_RXQS;
> + struct xps_dev_maps *dev_maps;
> + struct sock *sk = skb->sk;
> + int queue_index = -1;
> + unsigned int tci = 0;
> +
> + if (sk && sk->sk_rx_queue_mapping <= dev->real_num_rx_queues &&
> + dev->ifindex == sk->sk_rx_ifindex)
> + tci = sk->sk_rx_queue_mapping;
> +
> + rcu_read_lock();
> + while (queue_index < 0 && i < __XPS_MAP_MAX) {
> + if (i == XPS_MAP_CPUS)
This while loop typifies exactly why I don't think the XPS maps should
be an array. There's only two and we really don't want this to be an
open ended invitation for people to create new mapping methods. The
code is much simpler and potentially more efficient if the two maps
are just separate pointers. It should look something like this:
dev_maps = rcu_dereference(dev->xps_rxqs_map);
if (dev_maps) {
queue_index = __get_xps_queue_idx(dev, skb, dev_maps, tci);
if (queue_index < 0) {
dev_maps = rcu_dereference(dev->xps_rxqs_map);
if (dev_maps) {
queue_index = __get_xps_queue_idx(dev, skb,
dev_rxqs_maps, tci);
...
Also, the rxqs is a pretty narrow use case and it's likely to be
rarely configured (relative to a CPU map). A static_key could be used
to eliminate the cost of the extra map check (the static_key could
also be used on for CPU maps, this is analogous to how there are
static keys for RPS and RFS).
> + tci = skb->sender_cpu - 1;
> + dev_maps = rcu_dereference(dev->xps_maps[i]);
> + queue_index = __get_xps_queue_idx(dev, skb, dev_maps, tci);
> + i++;
> + }
> +
> rcu_read_unlock();
>
> return queue_index;
> diff --git a/net/core/sock.c b/net/core/sock.c
> index 042cfc6..73d7fa8 100644
> --- a/net/core/sock.c
> +++ b/net/core/sock.c
> @@ -2824,6 +2824,11 @@ void sock_init_data(struct socket *sock, struct sock *sk)
> sk->sk_pacing_rate = ~0U;
> sk->sk_pacing_shift = 10;
> sk->sk_incoming_cpu = -1;
> +
> +#ifdef CONFIG_XPS
> + sk->sk_rx_ifindex = -1;
> + sk->sk_rx_queue_mapping = -1;
> +#endif
> /*
> * Before updating sk_refcnt, we must commit prior changes to memory
> * (Documentation/RCU/rculist_nulls.txt for details)
> diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
> index b188e0d..d33911c 100644
> --- a/net/ipv4/tcp_input.c
> +++ b/net/ipv4/tcp_input.c
> @@ -78,6 +78,7 @@
> #include <linux/errqueue.h>
> #include <trace/events/tcp.h>
> #include <linux/static_key.h>
> +#include <net/busy_poll.h>
>
> int sysctl_tcp_max_orphans __read_mostly = NR_FILE;
>
> @@ -5559,6 +5560,11 @@ void tcp_finish_connect(struct sock *sk, struct sk_buff *skb)
> __tcp_fast_path_on(tp, tp->snd_wnd);
> else
> tp->pred_flags = 0;
> +
> + if (skb) {
> + sk_mark_napi_id(sk, skb);
> + sk_mark_rx_queue(sk, skb);
> + }
> }
>
> static bool tcp_rcv_fastopen_synack(struct sock *sk, struct sk_buff *synack,
> @@ -6371,6 +6377,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
> tcp_rsk(req)->snt_isn = isn;
> tcp_rsk(req)->txhash = net_tx_rndhash();
> tcp_openreq_init_rwin(req, sk, dst);
> + sk_mark_rx_queue(req_to_sk(req), skb);
> if (!want_cookie) {
> tcp_reqsk_record_syn(sk, req, skb);
> fastopen_sk = tcp_try_fastopen(sk, skb, req, &foc, dst);
> diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
> index caf23de..abdf02e 100644
> --- a/net/ipv4/tcp_ipv4.c
> +++ b/net/ipv4/tcp_ipv4.c
> @@ -1479,6 +1479,7 @@ int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb)
>
> sock_rps_save_rxhash(sk, skb);
> sk_mark_napi_id(sk, skb);
> + sk_mark_rx_queue(sk, skb);
Can this be done in sock_rps_save_rxhash?
> if (dst) {
> if (inet_sk(sk)->rx_dst_ifindex != skb->skb_iif ||
> !dst->ops->check(dst, 0)) {
> diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
> index f867658..4939c28 100644
> --- a/net/ipv4/tcp_minisocks.c
> +++ b/net/ipv4/tcp_minisocks.c
> @@ -836,6 +836,7 @@ int tcp_child_process(struct sock *parent, struct sock *child,
>
> /* record NAPI ID of child */
> sk_mark_napi_id(child, skb);
> + sk_mark_rx_queue(child, skb);
>
> tcp_segs_in(tcp_sk(child), skb);
> if (!sock_owned_by_user(child)) {
>
^ permalink raw reply
* Re: [net-next PATCH v2 1/4] net: Refactor XPS for CPUs and Rx queues
From: Tom Herbert @ 2018-05-18 4:08 UTC (permalink / raw)
To: Amritha Nambiar
Cc: Linux Kernel Network Developers, David S. Miller, Alexander Duyck,
Sridhar Samudrala, Eric Dumazet, Hannes Frederic Sowa
In-Reply-To: <152643400370.4991.2044471541271189575.stgit@anamdev.jf.intel.com>
On Tue, May 15, 2018 at 6:26 PM, Amritha Nambiar
<amritha.nambiar@intel.com> wrote:
> Refactor XPS code to support Tx queue selection based on
> CPU map or Rx queue map.
>
> Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
> ---
> include/linux/cpumask.h | 11 ++
> include/linux/netdevice.h | 72 +++++++++++++++-
> net/core/dev.c | 208 +++++++++++++++++++++++++++++----------------
> net/core/net-sysfs.c | 4 -
> 4 files changed, 215 insertions(+), 80 deletions(-)
>
> diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
> index bf53d89..57f20a0 100644
> --- a/include/linux/cpumask.h
> +++ b/include/linux/cpumask.h
> @@ -115,12 +115,17 @@ extern struct cpumask __cpu_active_mask;
> #define cpu_active(cpu) ((cpu) == 0)
> #endif
>
> -/* verify cpu argument to cpumask_* operators */
> -static inline unsigned int cpumask_check(unsigned int cpu)
> +static inline void cpu_max_bits_warn(unsigned int cpu, unsigned int bits)
> {
> #ifdef CONFIG_DEBUG_PER_CPU_MAPS
> - WARN_ON_ONCE(cpu >= nr_cpumask_bits);
> + WARN_ON_ONCE(cpu >= bits);
> #endif /* CONFIG_DEBUG_PER_CPU_MAPS */
> +}
> +
> +/* verify cpu argument to cpumask_* operators */
> +static inline unsigned int cpumask_check(unsigned int cpu)
> +{
> + cpu_max_bits_warn(cpu, nr_cpumask_bits);
> return cpu;
> }
>
> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> index 03ed492..c2eeb36 100644
> --- a/include/linux/netdevice.h
> +++ b/include/linux/netdevice.h
> @@ -730,10 +730,21 @@ struct xps_map {
> */
> struct xps_dev_maps {
> struct rcu_head rcu;
> - struct xps_map __rcu *cpu_map[0];
> + struct xps_map __rcu *attr_map[0];
> };
> -#define XPS_DEV_MAPS_SIZE(_tcs) (sizeof(struct xps_dev_maps) + \
> +
> +#define XPS_CPU_DEV_MAPS_SIZE(_tcs) (sizeof(struct xps_dev_maps) + \
> (nr_cpu_ids * (_tcs) * sizeof(struct xps_map *)))
> +
> +#define XPS_RXQ_DEV_MAPS_SIZE(_tcs, _rxqs) (sizeof(struct xps_dev_maps) +\
> + (_rxqs * (_tcs) * sizeof(struct xps_map *)))
> +
> +enum xps_map_type {
> + XPS_MAP_RXQS,
> + XPS_MAP_CPUS,
> + __XPS_MAP_MAX
> +};
> +
> #endif /* CONFIG_XPS */
>
> #define TC_MAX_QUEUE 16
> @@ -1891,7 +1902,7 @@ struct net_device {
> int watchdog_timeo;
>
> #ifdef CONFIG_XPS
> - struct xps_dev_maps __rcu *xps_maps;
> + struct xps_dev_maps __rcu *xps_maps[__XPS_MAP_MAX];
> #endif
> #ifdef CONFIG_NET_CLS_ACT
> struct mini_Qdisc __rcu *miniq_egress;
> @@ -3229,6 +3240,61 @@ static inline void netif_wake_subqueue(struct net_device *dev, u16 queue_index)
> #ifdef CONFIG_XPS
> int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> u16 index);
> +int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask,
> + u16 index, enum xps_map_type type);
> +
> +static inline bool attr_test_mask(unsigned long j, const unsigned long *mask,
> + unsigned int nr_bits)
> +{
> + cpu_max_bits_warn(j, nr_bits);
> + return test_bit(j, mask);
> +}
> +
> +static inline bool attr_test_online(unsigned long j,
> + const unsigned long *online_mask,
> + unsigned int nr_bits)
> +{
> + cpu_max_bits_warn(j, nr_bits);
> +
> + if (online_mask)
> + return test_bit(j, online_mask);
> +
> + if (j >= 0 && j < nr_bits)
> + return true;
> +
> + return false;
> +}
> +
> +static inline unsigned int attrmask_next(int n, const unsigned long *srcp,
> + unsigned int nr_bits)
> +{
> + /* -1 is a legal arg here. */
> + if (n != -1)
> + cpu_max_bits_warn(n, nr_bits);
> +
> + if (srcp)
> + return find_next_bit(srcp, nr_bits, n + 1);
> +
> + return n + 1;
> +}
> +
> +static inline int attrmask_next_and(int n, const unsigned long *src1p,
> + const unsigned long *src2p,
> + unsigned int nr_bits)
> +{
> + /* -1 is a legal arg here. */
> + if (n != -1)
> + cpu_max_bits_warn(n, nr_bits);
> +
> + if (src1p && src2p)
> + return find_next_and_bit(src1p, src2p, nr_bits, n + 1);
> + else if (src1p)
> + return find_next_bit(src1p, nr_bits, n + 1);
> + else if (src2p)
> + return find_next_bit(src2p, nr_bits, n + 1);
> +
> + return n + 1;
> +}
> #else
> static inline int netif_set_xps_queue(struct net_device *dev,
> const struct cpumask *mask,
> diff --git a/net/core/dev.c b/net/core/dev.c
> index 9f43901..7e5dfdb 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -2092,7 +2092,7 @@ static bool remove_xps_queue(struct xps_dev_maps *dev_maps,
> int pos;
>
> if (dev_maps)
> - map = xmap_dereference(dev_maps->cpu_map[tci]);
> + map = xmap_dereference(dev_maps->attr_map[tci]);
> if (!map)
> return false;
>
> @@ -2105,7 +2105,7 @@ static bool remove_xps_queue(struct xps_dev_maps *dev_maps,
> break;
> }
>
> - RCU_INIT_POINTER(dev_maps->cpu_map[tci], NULL);
> + RCU_INIT_POINTER(dev_maps->attr_map[tci], NULL);
> kfree_rcu(map, rcu);
> return false;
> }
> @@ -2125,7 +2125,7 @@ static bool remove_xps_queue_cpu(struct net_device *dev,
> int i, j;
>
> for (i = count, j = offset; i--; j++) {
> - if (!remove_xps_queue(dev_maps, cpu, j))
> + if (!remove_xps_queue(dev_maps, tci, j))
> break;
> }
>
> @@ -2138,30 +2138,47 @@ static bool remove_xps_queue_cpu(struct net_device *dev,
> static void netif_reset_xps_queues(struct net_device *dev, u16 offset,
> u16 count)
> {
> + const unsigned long *possible_mask = NULL;
> + enum xps_map_type type = XPS_MAP_RXQS;
> struct xps_dev_maps *dev_maps;
> - int cpu, i;
> bool active = false;
> + unsigned int nr_ids;
> + int i, j;
>
> mutex_lock(&xps_map_mutex);
> - dev_maps = xmap_dereference(dev->xps_maps);
>
> - if (!dev_maps)
> - goto out_no_maps;
> + while (type < __XPS_MAP_MAX) {
> + dev_maps = xmap_dereference(dev->xps_maps[type]);
> + if (!dev_maps)
> + goto out_no_maps;
> +
> + if (type == XPS_MAP_CPUS) {
> + if (num_possible_cpus() > 1)
> + possible_mask = cpumask_bits(cpu_possible_mask);
> + nr_ids = nr_cpu_ids;
> + } else if (type == XPS_MAP_RXQS) {
> + nr_ids = dev->num_rx_queues;
> + }
type is an enum so this should be a switch
>
> - for_each_possible_cpu(cpu)
> - active |= remove_xps_queue_cpu(dev, dev_maps, cpu,
> - offset, count);
> + for (j = -1; j = attrmask_next(j, possible_mask, nr_ids),
> + j < nr_ids;)
> + active |= remove_xps_queue_cpu(dev, dev_maps, j, offset,
> + count);
> + if (!active) {
> + RCU_INIT_POINTER(dev->xps_maps[type], NULL);
> + kfree_rcu(dev_maps, rcu);
> + }
>
> - if (!active) {
> - RCU_INIT_POINTER(dev->xps_maps, NULL);
> - kfree_rcu(dev_maps, rcu);
> + if (type == XPS_MAP_CPUS) {
> + for (i = offset + (count - 1); count--; i--)
> + netdev_queue_numa_node_write(
> + netdev_get_tx_queue(dev, i),
> + NUMA_NO_NODE);
> + }
> +out_no_maps:
> + type++;
> }
>
> - for (i = offset + (count - 1); count--; i--)
> - netdev_queue_numa_node_write(netdev_get_tx_queue(dev, i),
> - NUMA_NO_NODE);
> -
> -out_no_maps:
> mutex_unlock(&xps_map_mutex);
> }
>
> @@ -2170,11 +2187,11 @@ static void netif_reset_xps_queues_gt(struct net_device *dev, u16 index)
> netif_reset_xps_queues(dev, index, dev->num_tx_queues - index);
> }
>
> -static struct xps_map *expand_xps_map(struct xps_map *map,
> - int cpu, u16 index)
> +static struct xps_map *expand_xps_map(struct xps_map *map, int attr_index,
> + u16 index, enum xps_map_type type)
> {
> - struct xps_map *new_map;
> int alloc_len = XPS_MIN_MAP_ALLOC;
> + struct xps_map *new_map = NULL;
> int i, pos;
>
> for (pos = 0; map && pos < map->len; pos++) {
> @@ -2183,7 +2200,7 @@ static struct xps_map *expand_xps_map(struct xps_map *map,
> return map;
> }
>
> - /* Need to add queue to this CPU's existing map */
> + /* Need to add tx-queue to this CPU's/rx-queue's existing map */
> if (map) {
> if (pos < map->alloc_len)
> return map;
> @@ -2191,9 +2208,14 @@ static struct xps_map *expand_xps_map(struct xps_map *map,
> alloc_len = map->alloc_len * 2;
> }
>
> - /* Need to allocate new map to store queue on this CPU's map */
> - new_map = kzalloc_node(XPS_MAP_SIZE(alloc_len), GFP_KERNEL,
> - cpu_to_node(cpu));
> + /* Need to allocate new map to store tx-queue on this CPU's/rx-queue's
> + * map
> + */
> + if (type == XPS_MAP_RXQS)
> + new_map = kzalloc(XPS_MAP_SIZE(alloc_len), GFP_KERNEL);
> + else if (type == XPS_MAP_CPUS)
> + new_map = kzalloc_node(XPS_MAP_SIZE(alloc_len), GFP_KERNEL,
> + cpu_to_node(attr_index));
switch here also
> if (!new_map)
> return NULL;
>
> @@ -2205,14 +2227,16 @@ static struct xps_map *expand_xps_map(struct xps_map *map,
> return new_map;
> }
>
> -int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> - u16 index)
> +int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask,
> + u16 index, enum xps_map_type type)
> {
> + const unsigned long *online_mask = NULL, *possible_mask = NULL;
> struct xps_dev_maps *dev_maps, *new_dev_maps = NULL;
> - int i, cpu, tci, numa_node_id = -2;
> + int i, j, tci, numa_node_id = -2;
> int maps_sz, num_tc = 1, tc = 0;
> struct xps_map *map, *new_map;
> bool active = false;
> + unsigned int nr_ids;
>
> if (dev->num_tc) {
> num_tc = dev->num_tc;
> @@ -2221,16 +2245,33 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> return -EINVAL;
> }
>
> - maps_sz = XPS_DEV_MAPS_SIZE(num_tc);
> + switch (type) {
> + case XPS_MAP_RXQS:
> + maps_sz = XPS_RXQ_DEV_MAPS_SIZE(num_tc, dev->num_rx_queues);
> + dev_maps = xmap_dereference(dev->xps_maps[XPS_MAP_RXQS]);
> + nr_ids = dev->num_rx_queues;
> + break;
> + case XPS_MAP_CPUS:
> + maps_sz = XPS_CPU_DEV_MAPS_SIZE(num_tc);
> + if (num_possible_cpus() > 1) {
> + online_mask = cpumask_bits(cpu_online_mask);
> + possible_mask = cpumask_bits(cpu_possible_mask);
> + }
> + dev_maps = xmap_dereference(dev->xps_maps[XPS_MAP_CPUS]);
> + nr_ids = nr_cpu_ids;
> + break;
> + default:
> + return -EINVAL;
> + }
> +
> if (maps_sz < L1_CACHE_BYTES)
> maps_sz = L1_CACHE_BYTES;
>
> mutex_lock(&xps_map_mutex);
>
> - dev_maps = xmap_dereference(dev->xps_maps);
> -
> /* allocate memory for queue storage */
> - for_each_cpu_and(cpu, cpu_online_mask, mask) {
> + for (j = -1; j = attrmask_next_and(j, online_mask, mask, nr_ids),
> + j < nr_ids;) {
> if (!new_dev_maps)
> new_dev_maps = kzalloc(maps_sz, GFP_KERNEL);
> if (!new_dev_maps) {
> @@ -2238,73 +2279,81 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> return -ENOMEM;
> }
>
> - tci = cpu * num_tc + tc;
> - map = dev_maps ? xmap_dereference(dev_maps->cpu_map[tci]) :
> + tci = j * num_tc + tc;
> + map = dev_maps ? xmap_dereference(dev_maps->attr_map[tci]) :
> NULL;
>
> - map = expand_xps_map(map, cpu, index);
> + map = expand_xps_map(map, j, index, type);
> if (!map)
> goto error;
>
> - RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
> + RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
> }
>
> if (!new_dev_maps)
> goto out_no_new_maps;
>
> - for_each_possible_cpu(cpu) {
> + for (j = -1; j = attrmask_next(j, possible_mask, nr_ids),
> + j < nr_ids;) {
> /* copy maps belonging to foreign traffic classes */
> - for (i = tc, tci = cpu * num_tc; dev_maps && i--; tci++) {
> + for (i = tc, tci = j * num_tc; dev_maps && i--; tci++) {
> /* fill in the new device map from the old device map */
> - map = xmap_dereference(dev_maps->cpu_map[tci]);
> - RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
> + map = xmap_dereference(dev_maps->attr_map[tci]);
> + RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
> }
>
> /* We need to explicitly update tci as prevous loop
> * could break out early if dev_maps is NULL.
> */
> - tci = cpu * num_tc + tc;
> + tci = j * num_tc + tc;
>
> - if (cpumask_test_cpu(cpu, mask) && cpu_online(cpu)) {
> - /* add queue to CPU maps */
> + if (attr_test_mask(j, mask, nr_ids) &&
> + attr_test_online(j, online_mask, nr_ids)) {
> + /* add tx-queue to CPU/rx-queue maps */
> int pos = 0;
>
> - map = xmap_dereference(new_dev_maps->cpu_map[tci]);
> + map = xmap_dereference(new_dev_maps->attr_map[tci]);
> while ((pos < map->len) && (map->queues[pos] != index))
> pos++;
>
> if (pos == map->len)
> map->queues[map->len++] = index;
> #ifdef CONFIG_NUMA
> - if (numa_node_id == -2)
> - numa_node_id = cpu_to_node(cpu);
> - else if (numa_node_id != cpu_to_node(cpu))
> - numa_node_id = -1;
> + if (type == XPS_MAP_CPUS) {
> + if (numa_node_id == -2)
> + numa_node_id = cpu_to_node(j);
> + else if (numa_node_id != cpu_to_node(j))
> + numa_node_id = -1;
> + }
> #endif
> } else if (dev_maps) {
> /* fill in the new device map from the old device map */
> - map = xmap_dereference(dev_maps->cpu_map[tci]);
> - RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
> + map = xmap_dereference(dev_maps->attr_map[tci]);
> + RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
> }
>
> /* copy maps belonging to foreign traffic classes */
> for (i = num_tc - tc, tci++; dev_maps && --i; tci++) {
> /* fill in the new device map from the old device map */
> - map = xmap_dereference(dev_maps->cpu_map[tci]);
> - RCU_INIT_POINTER(new_dev_maps->cpu_map[tci], map);
> + map = xmap_dereference(dev_maps->attr_map[tci]);
> + RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map);
> }
> }
>
> - rcu_assign_pointer(dev->xps_maps, new_dev_maps);
> + if (type == XPS_MAP_RXQS)
> + rcu_assign_pointer(dev->xps_maps[XPS_MAP_RXQS], new_dev_maps);
> + else if (type == XPS_MAP_CPUS)
> + rcu_assign_pointer(dev->xps_maps[XPS_MAP_CPUS], new_dev_maps);
>
> /* Cleanup old maps */
> if (!dev_maps)
> goto out_no_old_maps;
>
> - for_each_possible_cpu(cpu) {
> - for (i = num_tc, tci = cpu * num_tc; i--; tci++) {
> - new_map = xmap_dereference(new_dev_maps->cpu_map[tci]);
> - map = xmap_dereference(dev_maps->cpu_map[tci]);
> + for (j = -1; j = attrmask_next(j, possible_mask, nr_ids),
> + j < nr_ids;) {
> + for (i = num_tc, tci = j * num_tc; i--; tci++) {
> + new_map = xmap_dereference(new_dev_maps->attr_map[tci]);
> + map = xmap_dereference(dev_maps->attr_map[tci]);
> if (map && map != new_map)
> kfree_rcu(map, rcu);
> }
> @@ -2317,19 +2366,23 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> active = true;
>
> out_no_new_maps:
> - /* update Tx queue numa node */
> - netdev_queue_numa_node_write(netdev_get_tx_queue(dev, index),
> - (numa_node_id >= 0) ? numa_node_id :
> - NUMA_NO_NODE);
> + if (type == XPS_MAP_CPUS) {
> + /* update Tx queue numa node */
> + netdev_queue_numa_node_write(netdev_get_tx_queue(dev, index),
> + (numa_node_id >= 0) ?
> + numa_node_id : NUMA_NO_NODE);
> + }
>
> if (!dev_maps)
> goto out_no_maps;
>
> - /* removes queue from unused CPUs */
> - for_each_possible_cpu(cpu) {
> - for (i = tc, tci = cpu * num_tc; i--; tci++)
> + /* removes tx-queue from unused CPUs/rx-queues */
> + for (j = -1; j = attrmask_next(j, possible_mask, nr_ids),
> + j < nr_ids;) {
> + for (i = tc, tci = j * num_tc; i--; tci++)
> active |= remove_xps_queue(dev_maps, tci, index);
> - if (!cpumask_test_cpu(cpu, mask) || !cpu_online(cpu))
> + if (!attr_test_mask(j, mask, nr_ids) ||
> + !attr_test_online(j, online_mask, nr_ids))
> active |= remove_xps_queue(dev_maps, tci, index);
> for (i = num_tc - tc, tci++; --i; tci++)
> active |= remove_xps_queue(dev_maps, tci, index);
> @@ -2337,7 +2390,10 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
>
> /* free map if not active */
> if (!active) {
> - RCU_INIT_POINTER(dev->xps_maps, NULL);
> + if (type == XPS_MAP_RXQS)
> + RCU_INIT_POINTER(dev->xps_maps[XPS_MAP_RXQS], NULL);
> + else if (type == XPS_MAP_CPUS)
> + RCU_INIT_POINTER(dev->xps_maps[XPS_MAP_CPUS], NULL);
> kfree_rcu(dev_maps, rcu);
> }
>
> @@ -2347,11 +2403,12 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> return 0;
> error:
> /* remove any maps that we added */
> - for_each_possible_cpu(cpu) {
> - for (i = num_tc, tci = cpu * num_tc; i--; tci++) {
> - new_map = xmap_dereference(new_dev_maps->cpu_map[tci]);
> + for (j = -1; j = attrmask_next(j, possible_mask, nr_ids),
> + j < nr_ids;) {
> + for (i = num_tc, tci = j * num_tc; i--; tci++) {
> + new_map = xmap_dereference(new_dev_maps->attr_map[tci]);
> map = dev_maps ?
> - xmap_dereference(dev_maps->cpu_map[tci]) :
> + xmap_dereference(dev_maps->attr_map[tci]) :
> NULL;
> if (new_map && new_map != map)
> kfree(new_map);
> @@ -2363,6 +2420,13 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> kfree(new_dev_maps);
> return -ENOMEM;
> }
> +
> +int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
> + u16 index)
> +{
> + return __netif_set_xps_queue(dev, cpumask_bits(mask), index,
> + XPS_MAP_CPUS);
> +}
> EXPORT_SYMBOL(netif_set_xps_queue);
>
> #endif
> @@ -3402,7 +3466,7 @@ static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
> int queue_index = -1;
>
> rcu_read_lock();
> - dev_maps = rcu_dereference(dev->xps_maps);
> + dev_maps = rcu_dereference(dev->xps_maps[XPS_MAP_CPUS]);
> if (dev_maps) {
> unsigned int tci = skb->sender_cpu - 1;
>
> @@ -3411,7 +3475,7 @@ static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
> tci += netdev_get_prio_tc_map(dev, skb->priority);
> }
>
> - map = rcu_dereference(dev_maps->cpu_map[tci]);
> + map = rcu_dereference(dev_maps->attr_map[tci]);
> if (map) {
> if (map->len == 1)
> queue_index = map->queues[0];
> diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
> index c476f07..d7abd33 100644
> --- a/net/core/net-sysfs.c
> +++ b/net/core/net-sysfs.c
> @@ -1227,13 +1227,13 @@ static ssize_t xps_cpus_show(struct netdev_queue *queue,
> }
>
> rcu_read_lock();
> - dev_maps = rcu_dereference(dev->xps_maps);
> + dev_maps = rcu_dereference(dev->xps_maps[XPS_MAP_CPUS]);
> if (dev_maps) {
> for_each_possible_cpu(cpu) {
> int i, tci = cpu * num_tc + tc;
> struct xps_map *map;
>
> - map = rcu_dereference(dev_maps->cpu_map[tci]);
> + map = rcu_dereference(dev_maps->attr_map[tci]);
> if (!map)
> continue;
>
>
^ permalink raw reply
* Re: [Cake] [PATCH net-next v12 3/7] sch_cake: Add optional ACK filter
From: Eric Dumazet @ 2018-05-18 4:08 UTC (permalink / raw)
To: Ryan Mounce, Toke Høiland-Jørgensen
Cc: Eric Dumazet, netdev, Cake List
In-Reply-To: <CAN+fvRYP4oHQDyz=NGPnfR8+PZ04ammP-sgeB_9Gryu4xxyf0w@mail.gmail.com>
On 05/17/2018 07:36 PM, Ryan Mounce wrote:
> On 17 May 2018 at 22:41, Toke Høiland-Jørgensen <toke@toke.dk> wrote:
>> Eric Dumazet <eric.dumazet@gmail.com> writes:
>>
>>> On 05/17/2018 04:23 AM, Toke Høiland-Jørgensen wrote:
>>>
>>>>
>>>> We don't do full parsing of SACKs, no; we were trying to keep things
>>>> simple... We do detect the presence of SACK options, though, and the
>>>> presence of SACK options on an ACK will make previous ACKs be considered
>>>> redundant.
>>>>
>>>
>>> But they are not redundant in some cases, particularly when reorders
>>> happen in the network.
>>
>> Huh. I was under the impression that SACKs were basically cumulative
>> until cleared.
>>
>> I.e., in packet sequence ABCDE where B and D are lost, C would have
>> SACK(B) and E would have SACK(B,D). Are you saying that E would only
>> have SACK(D)?
>
> SACK works by acknowledging additional ranges above those that have
> been ACKed, rather than ACKing up to the largest seen sequence number
> and reporting missing ranges before that.
>
> A - ACK(A)
> B - lost
> C - ACK(A) + SACK(C)
> D - lost
> E - ACK(A) + SACK(C, E)
>
> Cake does check that the ACK sequence number is greater, or if it is
> equal and the 'newer' ACK has the SACK option present. It doesn't
> compare the sequence numbers inside two SACKs. If the two SACKs in the
> above example had been reordered before reaching cake's ACK filter in
> aggressive mode, the wrong one will be filtered.
>
> This is a limitation of my naive SACK handling in cake. The default
> 'conservative' mode happens to mitigate the problem in the above
> scenario, but the issue could still present itself in more
> pathological cases. It's fixable, however I'm not sure this corner
> case is sufficiently common or severe to warrant the extra complexity.
The extra complexity is absolutely requested for inclusion in upstream linux.
I recommend reading rfc 2018, whole section 4 (Generating Sack Options: Data Receiver Behavior
)
Proposed ACK filter in Cake is messing the protocol, since the first rule is not respected
* The first SACK block (i.e., the one immediately following the
kind and length fields in the option) MUST specify the contiguous
block of data containing the segment which triggered this ACK,
unless that segment advanced the Acknowledgment Number field in
the header. This assures that the ACK with the SACK option
reflects the most recent change in the data receiver's buffer
queue.
An ACK filter must either :
Not merge ACK if they contain different SACK blocks.
Or make a precise analysis of the SACK blocks to determine if the merge is allowed,
ie no useful information is lost.
The sender should get all the information as which segments were received correctly,
assuming no ACK are dropped because of congestion on return path.
^ permalink raw reply
* Re: [net-next PATCH v2 0/4] Symmetric queue selection using XPS for Rx queues
From: Tom Herbert @ 2018-05-18 4:11 UTC (permalink / raw)
To: Amritha Nambiar
Cc: Linux Kernel Network Developers, David S. Miller, Alexander Duyck,
Sridhar Samudrala, Eric Dumazet, Hannes Frederic Sowa
In-Reply-To: <152643356116.4991.7215767041139726872.stgit@anamdev.jf.intel.com>
On Tue, May 15, 2018 at 6:26 PM, Amritha Nambiar
<amritha.nambiar@intel.com> wrote:
> This patch series implements support for Tx queue selection based on
> Rx queue(s) map. This is done by configuring Rx queue(s) map per Tx-queue
> using sysfs attribute. If the user configuration for Rx queues does
> not apply, then the Tx queue selection falls back to XPS using CPUs and
> finally to hashing.
>
> XPS is refactored to support Tx queue selection based on either the
> CPUs map or the Rx-queues map. The config option CONFIG_XPS needs to be
> enabled. By default no receive queues are configured for the Tx queue.
>
> - /sys/class/net/<dev>/queues/tx-*/xps_rxqs
>
> This is to enable sending packets on the same Tx-Rx queue pair as this
If I'm reading the patch correctly, isn't this mapping rxq to a set of
txqs (in other words not strictly queue pair which has other
connotations in NIC HW). It is important to make it clear that this
feature is no HW dependent.
> is useful for busy polling multi-threaded workloads where it is not
> possible to pin the threads to a CPU. This is a rework of Sridhar's
> patch for symmetric queueing via socket option:
> https://www.spinics.net/lists/netdev/msg453106.html
>
Please add something about how this was tested and what the
performance gain is to justify the feature.
> v2:
> - Added documentation in networking/scaling.txt
> - Added a simple routine to replace multiple ifdef blocks.
>
> ---
>
> Amritha Nambiar (4):
> net: Refactor XPS for CPUs and Rx queues
> net: Enable Tx queue selection based on Rx queues
> net-sysfs: Add interface for Rx queue map per Tx queue
> Documentation: Add explanation for XPS using Rx-queue map
>
>
> Documentation/networking/scaling.txt | 58 +++++++-
> include/linux/cpumask.h | 11 +-
> include/linux/netdevice.h | 72 ++++++++++
> include/net/sock.h | 18 +++
> net/core/dev.c | 242 +++++++++++++++++++++++-----------
> net/core/net-sysfs.c | 85 ++++++++++++
> net/core/sock.c | 5 +
> net/ipv4/tcp_input.c | 7 +
> net/ipv4/tcp_ipv4.c | 1
> net/ipv4/tcp_minisocks.c | 1
> 10 files changed, 404 insertions(+), 96 deletions(-)
>
> --
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox