From 1543e317f1da31b75942316931e8f491a8920811 Mon Sep 17 00:00:00 2001
From: hc <hc@nodka.com>
Date: Thu, 04 Jan 2024 10:08:02 +0000
Subject: [PATCH] disable FB
---
kernel/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 1397 ++++++++++++++++++++++++++++++++++++++++++---------------
1 files changed, 1,019 insertions(+), 378 deletions(-)
diff --git a/kernel/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/kernel/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index f0aa7f0..4e8e379 100644
--- a/kernel/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/kernel/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -32,6 +32,7 @@
#include <linux/mutex.h>
#include <linux/mlx5/driver.h>
+#include <linux/mlx5/vport.h>
#include <linux/mlx5/eswitch.h>
#include "mlx5_core.h"
@@ -58,7 +59,8 @@
ADD_PRIO(num_prios_val, 0, num_levels_val, {},\
__VA_ARGS__)\
-#define ADD_NS(...) {.type = FS_TYPE_NAMESPACE,\
+#define ADD_NS(def_miss_act, ...) {.type = FS_TYPE_NAMESPACE, \
+ .def_miss_action = def_miss_act,\
.children = (struct init_tree_node[]) {__VA_ARGS__},\
.ar_size = INIT_TREE_NODE_ARRAY_SIZE(__VA_ARGS__) \
}
@@ -76,6 +78,23 @@
FS_CAP(flow_table_properties_nic_receive.identified_miss_table_mode), \
FS_CAP(flow_table_properties_nic_receive.flow_table_modify))
+#define FS_CHAINING_CAPS_EGRESS \
+ FS_REQUIRED_CAPS( \
+ FS_CAP(flow_table_properties_nic_transmit.flow_modify_en), \
+ FS_CAP(flow_table_properties_nic_transmit.modify_root), \
+ FS_CAP(flow_table_properties_nic_transmit \
+ .identified_miss_table_mode), \
+ FS_CAP(flow_table_properties_nic_transmit.flow_table_modify))
+
+#define FS_CHAINING_CAPS_RDMA_TX \
+ FS_REQUIRED_CAPS( \
+ FS_CAP(flow_table_properties_nic_transmit_rdma.flow_modify_en), \
+ FS_CAP(flow_table_properties_nic_transmit_rdma.modify_root), \
+ FS_CAP(flow_table_properties_nic_transmit_rdma \
+ .identified_miss_table_mode), \
+ FS_CAP(flow_table_properties_nic_transmit_rdma \
+ .flow_table_modify))
+
#define LEFTOVERS_NUM_LEVELS 1
#define LEFTOVERS_NUM_PRIOS 1
@@ -86,8 +105,8 @@
#define ETHTOOL_PRIO_NUM_LEVELS 1
#define ETHTOOL_NUM_PRIOS 11
#define ETHTOOL_MIN_LEVEL (KERNEL_MIN_LEVEL + ETHTOOL_NUM_PRIOS)
-/* Vlan, mac, ttc, inner ttc, aRFS */
-#define KERNEL_NIC_PRIO_NUM_LEVELS 5
+/* Vlan, mac, ttc, inner ttc, {aRFS/accel and esp/esp_err} */
+#define KERNEL_NIC_PRIO_NUM_LEVELS 6
#define KERNEL_NIC_NUM_PRIOS 1
/* One more level for tc */
#define KERNEL_MIN_LEVEL (KERNEL_NIC_PRIO_NUM_LEVELS + 1)
@@ -99,13 +118,17 @@
#define ANCHOR_NUM_PRIOS 1
#define ANCHOR_MIN_LEVEL (BY_PASS_MIN_LEVEL + 1)
-#define OFFLOADS_MAX_FT 1
-#define OFFLOADS_NUM_PRIOS 1
-#define OFFLOADS_MIN_LEVEL (ANCHOR_MIN_LEVEL + 1)
+#define OFFLOADS_MAX_FT 2
+#define OFFLOADS_NUM_PRIOS 2
+#define OFFLOADS_MIN_LEVEL (ANCHOR_MIN_LEVEL + OFFLOADS_NUM_PRIOS)
#define LAG_PRIO_NUM_LEVELS 1
#define LAG_NUM_PRIOS 1
#define LAG_MIN_LEVEL (OFFLOADS_MIN_LEVEL + 1)
+
+#define KERNEL_TX_IPSEC_NUM_PRIOS 1
+#define KERNEL_TX_IPSEC_NUM_LEVELS 1
+#define KERNEL_TX_MIN_LEVEL (KERNEL_TX_IPSEC_NUM_LEVELS)
struct node_caps {
size_t arr_sz;
@@ -121,33 +144,96 @@
int num_leaf_prios;
int prio;
int num_levels;
+ enum mlx5_flow_table_miss_action def_miss_action;
} root_fs = {
.type = FS_TYPE_NAMESPACE,
.ar_size = 7,
+ .children = (struct init_tree_node[]){
+ ADD_PRIO(0, BY_PASS_MIN_LEVEL, 0, FS_CHAINING_CAPS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_PRIOS,
+ BY_PASS_PRIO_NUM_LEVELS))),
+ ADD_PRIO(0, LAG_MIN_LEVEL, 0, FS_CHAINING_CAPS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(LAG_NUM_PRIOS,
+ LAG_PRIO_NUM_LEVELS))),
+ ADD_PRIO(0, OFFLOADS_MIN_LEVEL, 0, FS_CHAINING_CAPS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(OFFLOADS_NUM_PRIOS,
+ OFFLOADS_MAX_FT))),
+ ADD_PRIO(0, ETHTOOL_MIN_LEVEL, 0, FS_CHAINING_CAPS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(ETHTOOL_NUM_PRIOS,
+ ETHTOOL_PRIO_NUM_LEVELS))),
+ ADD_PRIO(0, KERNEL_MIN_LEVEL, 0, {},
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(KERNEL_NIC_TC_NUM_PRIOS,
+ KERNEL_NIC_TC_NUM_LEVELS),
+ ADD_MULTIPLE_PRIO(KERNEL_NIC_NUM_PRIOS,
+ KERNEL_NIC_PRIO_NUM_LEVELS))),
+ ADD_PRIO(0, BY_PASS_MIN_LEVEL, 0, FS_CHAINING_CAPS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(LEFTOVERS_NUM_PRIOS,
+ LEFTOVERS_NUM_LEVELS))),
+ ADD_PRIO(0, ANCHOR_MIN_LEVEL, 0, {},
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(ANCHOR_NUM_PRIOS,
+ ANCHOR_NUM_LEVELS))),
+ }
+};
+
+static struct init_tree_node egress_root_fs = {
+ .type = FS_TYPE_NAMESPACE,
+#ifdef CONFIG_MLX5_IPSEC
+ .ar_size = 2,
+#else
+ .ar_size = 1,
+#endif
.children = (struct init_tree_node[]) {
- ADD_PRIO(0, BY_PASS_MIN_LEVEL, 0,
- FS_CHAINING_CAPS,
- ADD_NS(ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_PRIOS,
+ ADD_PRIO(0, MLX5_BY_PASS_NUM_PRIOS, 0,
+ FS_CHAINING_CAPS_EGRESS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_PRIOS,
BY_PASS_PRIO_NUM_LEVELS))),
- ADD_PRIO(0, LAG_MIN_LEVEL, 0,
+#ifdef CONFIG_MLX5_IPSEC
+ ADD_PRIO(0, KERNEL_TX_MIN_LEVEL, 0,
+ FS_CHAINING_CAPS_EGRESS,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(KERNEL_TX_IPSEC_NUM_PRIOS,
+ KERNEL_TX_IPSEC_NUM_LEVELS))),
+#endif
+ }
+};
+
+#define RDMA_RX_BYPASS_PRIO 0
+#define RDMA_RX_KERNEL_PRIO 1
+static struct init_tree_node rdma_rx_root_fs = {
+ .type = FS_TYPE_NAMESPACE,
+ .ar_size = 2,
+ .children = (struct init_tree_node[]) {
+ [RDMA_RX_BYPASS_PRIO] =
+ ADD_PRIO(0, MLX5_BY_PASS_NUM_REGULAR_PRIOS, 0,
FS_CHAINING_CAPS,
- ADD_NS(ADD_MULTIPLE_PRIO(LAG_NUM_PRIOS,
- LAG_PRIO_NUM_LEVELS))),
- ADD_PRIO(0, OFFLOADS_MIN_LEVEL, 0, {},
- ADD_NS(ADD_MULTIPLE_PRIO(OFFLOADS_NUM_PRIOS, OFFLOADS_MAX_FT))),
- ADD_PRIO(0, ETHTOOL_MIN_LEVEL, 0,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_REGULAR_PRIOS,
+ BY_PASS_PRIO_NUM_LEVELS))),
+ [RDMA_RX_KERNEL_PRIO] =
+ ADD_PRIO(0, MLX5_BY_PASS_NUM_REGULAR_PRIOS + 1, 0,
FS_CHAINING_CAPS,
- ADD_NS(ADD_MULTIPLE_PRIO(ETHTOOL_NUM_PRIOS,
- ETHTOOL_PRIO_NUM_LEVELS))),
- ADD_PRIO(0, KERNEL_MIN_LEVEL, 0, {},
- ADD_NS(ADD_MULTIPLE_PRIO(KERNEL_NIC_TC_NUM_PRIOS, KERNEL_NIC_TC_NUM_LEVELS),
- ADD_MULTIPLE_PRIO(KERNEL_NIC_NUM_PRIOS,
- KERNEL_NIC_PRIO_NUM_LEVELS))),
- ADD_PRIO(0, BY_PASS_MIN_LEVEL, 0,
- FS_CHAINING_CAPS,
- ADD_NS(ADD_MULTIPLE_PRIO(LEFTOVERS_NUM_PRIOS, LEFTOVERS_NUM_LEVELS))),
- ADD_PRIO(0, ANCHOR_MIN_LEVEL, 0, {},
- ADD_NS(ADD_MULTIPLE_PRIO(ANCHOR_NUM_PRIOS, ANCHOR_NUM_LEVELS))),
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_SWITCH_DOMAIN,
+ ADD_MULTIPLE_PRIO(1, 1))),
+ }
+};
+
+static struct init_tree_node rdma_tx_root_fs = {
+ .type = FS_TYPE_NAMESPACE,
+ .ar_size = 1,
+ .children = (struct init_tree_node[]) {
+ ADD_PRIO(0, MLX5_BY_PASS_NUM_PRIOS, 0,
+ FS_CHAINING_CAPS_RDMA_TX,
+ ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_PRIOS,
+ BY_PASS_PRIO_NUM_LEVELS))),
}
};
@@ -158,7 +244,7 @@
};
static const struct rhashtable_params rhash_fte = {
- .key_len = FIELD_SIZEOF(struct fs_fte, val),
+ .key_len = sizeof_field(struct fs_fte, val),
.key_offset = offsetof(struct fs_fte, val),
.head_offset = offsetof(struct fs_fte, hash),
.automatic_shrinking = true,
@@ -166,7 +252,7 @@
};
static const struct rhashtable_params rhash_fg = {
- .key_len = FIELD_SIZEOF(struct mlx5_flow_group, mask),
+ .key_len = sizeof_field(struct mlx5_flow_group, mask),
.key_offset = offsetof(struct mlx5_flow_group, mask),
.head_offset = offsetof(struct mlx5_flow_group, hash),
.automatic_shrinking = true,
@@ -182,7 +268,7 @@
static void del_sw_fte(struct fs_node *node);
static void del_sw_prio(struct fs_node *node);
static void del_sw_ns(struct fs_node *node);
-/* Delete rule (destination) is special case that
+/* Delete rule (destination) is special case that
* requires to lock the FTE for all the deletion process.
*/
static void del_sw_hw_rule(struct fs_node *node);
@@ -242,10 +328,11 @@
}
}
-static void down_write_ref_node(struct fs_node *node)
+static void down_write_ref_node(struct fs_node *node, bool locked)
{
if (node) {
- down_write(&node->lock);
+ if (!locked)
+ down_write(&node->lock);
refcount_inc(&node->refcount);
}
}
@@ -256,13 +343,14 @@
up_read(&node->lock);
}
-static void up_write_ref_node(struct fs_node *node)
+static void up_write_ref_node(struct fs_node *node, bool locked)
{
refcount_dec(&node->refcount);
- up_write(&node->lock);
+ if (!locked)
+ up_write(&node->lock);
}
-static void tree_put_node(struct fs_node *node)
+static void tree_put_node(struct fs_node *node, bool locked)
{
struct fs_node *parent_node = node->parent;
@@ -270,30 +358,25 @@
if (node->del_hw_func)
node->del_hw_func(node);
if (parent_node) {
- /* Only root namespace doesn't have parent and we just
- * need to free its node.
- */
- down_write_ref_node(parent_node);
+ down_write_ref_node(parent_node, locked);
list_del_init(&node->list);
- if (node->del_sw_func)
- node->del_sw_func(node);
- up_write_ref_node(parent_node);
- } else {
- kfree(node);
}
+ node->del_sw_func(node);
+ if (parent_node)
+ up_write_ref_node(parent_node, locked);
node = NULL;
}
if (!node && parent_node)
- tree_put_node(parent_node);
+ tree_put_node(parent_node, locked);
}
-static int tree_remove_node(struct fs_node *node)
+static int tree_remove_node(struct fs_node *node, bool locked)
{
if (refcount_read(&node->refcount) > 1) {
refcount_dec(&node->refcount);
return -EEXIST;
}
- tree_put_node(node);
+ tree_put_node(node, locked);
return 0;
}
@@ -308,6 +391,12 @@
}
return NULL;
+}
+
+static bool is_fwd_next_action(u32 action)
+{
+ return action & (MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO |
+ MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_NS);
}
static bool check_valid_spec(const struct mlx5_flow_spec *spec)
@@ -377,9 +466,10 @@
fs_get_obj(ft, node);
dev = get_dev(&ft->node);
root = find_root(&ft->node);
+ trace_mlx5_fs_del_ft(ft);
if (node->active) {
- err = root->cmds->destroy_flow_table(dev, ft);
+ err = root->cmds->destroy_flow_table(root, ft);
if (err)
mlx5_core_warn(dev, "flow steering can't destroy ft\n");
}
@@ -393,29 +483,43 @@
fs_get_obj(ft, node);
rhltable_destroy(&ft->fgs_hash);
- fs_get_obj(prio, ft->node.parent);
- prio->num_ft--;
+ if (ft->node.parent) {
+ fs_get_obj(prio, ft->node.parent);
+ prio->num_ft--;
+ }
kfree(ft);
+}
+
+static void modify_fte(struct fs_fte *fte)
+{
+ struct mlx5_flow_root_namespace *root;
+ struct mlx5_flow_table *ft;
+ struct mlx5_flow_group *fg;
+ struct mlx5_core_dev *dev;
+ int err;
+
+ fs_get_obj(fg, fte->node.parent);
+ fs_get_obj(ft, fg->node.parent);
+ dev = get_dev(&fte->node);
+
+ root = find_root(&ft->node);
+ err = root->cmds->update_fte(root, ft, fg, fte->modify_mask, fte);
+ if (err)
+ mlx5_core_warn(dev,
+ "%s can't del rule fg id=%d fte_index=%d\n",
+ __func__, fg->id, fte->index);
+ fte->modify_mask = 0;
}
static void del_sw_hw_rule(struct fs_node *node)
{
- struct mlx5_flow_root_namespace *root;
struct mlx5_flow_rule *rule;
- struct mlx5_flow_table *ft;
- struct mlx5_flow_group *fg;
struct fs_fte *fte;
- int modify_mask;
- struct mlx5_core_dev *dev = get_dev(node);
- int err;
- bool update_fte = false;
fs_get_obj(rule, node);
fs_get_obj(fte, rule->node.parent);
- fs_get_obj(fg, fte->node.parent);
- fs_get_obj(ft, fg->node.parent);
trace_mlx5_fs_del_rule(rule);
- if (rule->sw_action == MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO) {
+ if (is_fwd_next_action(rule->sw_action)) {
mutex_lock(&rule->dest_attr.ft->lock);
list_del(&rule->next_ft);
mutex_unlock(&rule->dest_attr.ft->lock);
@@ -423,27 +527,26 @@
if (rule->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_COUNTER &&
--fte->dests_size) {
- modify_mask = BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION) |
- BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_FLOW_COUNTERS);
+ fte->modify_mask |=
+ BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION) |
+ BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_FLOW_COUNTERS);
fte->action.action &= ~MLX5_FLOW_CONTEXT_ACTION_COUNT;
- update_fte = true;
+ goto out;
+ }
+
+ if (rule->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_PORT &&
+ --fte->dests_size) {
+ fte->modify_mask |= BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION);
+ fte->action.action &= ~MLX5_FLOW_CONTEXT_ACTION_ALLOW;
goto out;
}
if ((fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) &&
--fte->dests_size) {
- modify_mask = BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_DESTINATION_LIST);
- update_fte = true;
+ fte->modify_mask |=
+ BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_DESTINATION_LIST);
}
out:
- root = find_root(&ft->node);
- if (update_fte && fte->dests_size) {
- err = root->cmds->update_fte(dev, ft, fg->id, modify_mask, fte);
- if (err)
- mlx5_core_warn(dev,
- "%s can't del rule fg id=%d fte_index=%d\n",
- __func__, fg->id, fte->index);
- }
kfree(rule);
}
@@ -464,7 +567,7 @@
dev = get_dev(&ft->node);
root = find_root(&ft->node);
if (node->active) {
- err = root->cmds->delete_fte(dev, ft, fte);
+ err = root->cmds->delete_fte(root, ft, fte);
if (err)
mlx5_core_warn(dev,
"flow steering can't delete fte in index %d of flow group id %d\n",
@@ -504,7 +607,7 @@
trace_mlx5_fs_del_fg(fg);
root = find_root(&ft->node);
- if (fg->node.active && root->cmds->destroy_flow_group(dev, ft, fg->id))
+ if (fg->node.active && root->cmds->destroy_flow_group(root, ft, fg))
mlx5_core_warn(dev, "flow steering can't destroy fg %d of ft %d\n",
fg->id, ft->id);
}
@@ -521,7 +624,9 @@
rhashtable_destroy(&fg->ftes_hash);
ida_destroy(&fg->fte_allocator);
- if (ft->autogroup.active && fg->max_ftes == ft->autogroup.group_size)
+ if (ft->autogroup.active &&
+ fg->max_ftes == ft->autogroup.group_size &&
+ fg->start_index < ft->autogroup.max_fte)
ft->autogroup.num_groups--;
err = rhltable_remove(&ft->fgs_hash,
&fg->hash,
@@ -556,7 +661,7 @@
}
static struct fs_fte *alloc_fte(struct mlx5_flow_table *ft,
- u32 *match_value,
+ const struct mlx5_flow_spec *spec,
struct mlx5_flow_act *flow_act)
{
struct mlx5_flow_steering *steering = get_steering(&ft->node);
@@ -566,9 +671,10 @@
if (!fte)
return ERR_PTR(-ENOMEM);
- memcpy(fte->val, match_value, sizeof(fte->val));
+ memcpy(fte->val, &spec->match_value, sizeof(fte->val));
fte->node.type = FS_TYPE_FLOW_ENTRY;
fte->action = *flow_act;
+ fte->flow_context = spec->flow_context;
tree_init_node(&fte->node, del_hw_fte, del_sw_fte);
@@ -584,7 +690,7 @@
static struct mlx5_flow_group *alloc_flow_group(struct mlx5_flow_steering *steering,
u8 match_criteria_enable,
- void *match_criteria,
+ const void *match_criteria,
int start_index,
int end_index)
{
@@ -599,7 +705,8 @@
if (ret) {
kmem_cache_free(steering->fgs_cache, fg);
return ERR_PTR(ret);
-}
+ }
+
ida_init(&fg->fte_allocator);
fg->mask.match_criteria_enable = match_criteria_enable;
memcpy(&fg->mask.match_criteria, match_criteria,
@@ -613,7 +720,7 @@
static struct mlx5_flow_group *alloc_insert_flow_group(struct mlx5_flow_table *ft,
u8 match_criteria_enable,
- void *match_criteria,
+ const void *match_criteria,
int start_index,
int end_index,
struct list_head *prev)
@@ -711,20 +818,42 @@
return ft;
}
-/* If reverse if false then return the first flow table in next priority of
- * prio in the tree, else return the last flow table in the previous priority
- * of prio in the tree.
- */
-static struct mlx5_flow_table *find_closest_ft(struct fs_prio *prio, bool reverse)
+static struct fs_node *find_prio_chains_parent(struct fs_node *parent,
+ struct fs_node **child)
{
+ struct fs_node *node = NULL;
+
+ while (parent && parent->type != FS_TYPE_PRIO_CHAINS) {
+ node = parent;
+ parent = parent->parent;
+ }
+
+ if (child)
+ *child = node;
+
+ return parent;
+}
+
+/* If reverse is false then return the first flow table next to the passed node
+ * in the tree, else return the last flow table before the node in the tree.
+ * If skip is true, skip the flow tables in the same prio_chains prio.
+ */
+static struct mlx5_flow_table *find_closest_ft(struct fs_node *node, bool reverse,
+ bool skip)
+{
+ struct fs_node *prio_chains_parent = NULL;
struct mlx5_flow_table *ft = NULL;
struct fs_node *curr_node;
struct fs_node *parent;
- parent = prio->node.parent;
- curr_node = &prio->node;
+ if (skip)
+ prio_chains_parent = find_prio_chains_parent(node, NULL);
+ parent = node->parent;
+ curr_node = node;
while (!ft && parent) {
- ft = find_closest_ft_recursive(parent, &curr_node->list, reverse);
+ if (parent != prio_chains_parent)
+ ft = find_closest_ft_recursive(parent, &curr_node->list,
+ reverse);
curr_node = parent;
parent = curr_node->parent;
}
@@ -732,15 +861,27 @@
}
/* Assuming all the tree is locked by mutex chain lock */
-static struct mlx5_flow_table *find_next_chained_ft(struct fs_prio *prio)
+static struct mlx5_flow_table *find_next_chained_ft(struct fs_node *node)
{
- return find_closest_ft(prio, false);
+ return find_closest_ft(node, false, true);
}
/* Assuming all the tree is locked by mutex chain lock */
-static struct mlx5_flow_table *find_prev_chained_ft(struct fs_prio *prio)
+static struct mlx5_flow_table *find_prev_chained_ft(struct fs_node *node)
{
- return find_closest_ft(prio, true);
+ return find_closest_ft(node, true, true);
+}
+
+static struct mlx5_flow_table *find_next_fwd_ft(struct mlx5_flow_table *ft,
+ struct mlx5_flow_act *flow_act)
+{
+ struct fs_prio *prio;
+ bool next_ns;
+
+ next_ns = flow_act->action & MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_NS;
+ fs_get_obj(prio, next_ns ? ft->ns->node.parent : ft->node.parent);
+
+ return find_next_chained_ft(&prio->node);
}
static int connect_fts_in_prio(struct mlx5_core_dev *dev,
@@ -749,22 +890,34 @@
{
struct mlx5_flow_root_namespace *root = find_root(&prio->node);
struct mlx5_flow_table *iter;
- int i = 0;
int err;
fs_for_each_ft(iter, prio) {
- i++;
- err = root->cmds->modify_flow_table(dev, iter, ft);
+ err = root->cmds->modify_flow_table(root, iter, ft);
if (err) {
- mlx5_core_warn(dev, "Failed to modify flow table %d\n",
- iter->id);
+ mlx5_core_err(dev,
+ "Failed to modify flow table id %d, type %d, err %d\n",
+ iter->id, iter->type, err);
/* The driver is out of sync with the FW */
- if (i > 1)
- WARN_ON(true);
return err;
}
}
return 0;
+}
+
+static struct mlx5_flow_table *find_closet_ft_prio_chains(struct fs_node *node,
+ struct fs_node *parent,
+ struct fs_node **child,
+ bool reverse)
+{
+ struct mlx5_flow_table *ft;
+
+ ft = find_closest_ft(node, reverse, false);
+
+ if (ft && parent == find_prio_chains_parent(&ft->node, child))
+ return ft;
+
+ return NULL;
}
/* Connect flow tables from previous priority of prio to ft */
@@ -772,16 +925,35 @@
struct mlx5_flow_table *ft,
struct fs_prio *prio)
{
+ struct fs_node *prio_parent, *parent = NULL, *child, *node;
struct mlx5_flow_table *prev_ft;
+ int err = 0;
- prev_ft = find_prev_chained_ft(prio);
- if (prev_ft) {
+ prio_parent = find_prio_chains_parent(&prio->node, &child);
+
+ /* return directly if not under the first sub ns of prio_chains prio */
+ if (prio_parent && !list_is_first(&child->list, &prio_parent->children))
+ return 0;
+
+ prev_ft = find_prev_chained_ft(&prio->node);
+ while (prev_ft) {
struct fs_prio *prev_prio;
fs_get_obj(prev_prio, prev_ft->node.parent);
- return connect_fts_in_prio(dev, prev_prio, ft);
+ err = connect_fts_in_prio(dev, prev_prio, ft);
+ if (err)
+ break;
+
+ if (!parent) {
+ parent = find_prio_chains_parent(&prev_prio->node, &child);
+ if (!parent)
+ break;
+ }
+
+ node = child;
+ prev_ft = find_closet_ft_prio_chains(node, parent, &child, true);
}
- return 0;
+ return err;
}
static int update_root_ft_create(struct mlx5_flow_table *ft, struct fs_prio
@@ -790,7 +962,7 @@
struct mlx5_flow_root_namespace *root = find_root(&prio->node);
struct mlx5_ft_underlay_qp *uqp;
int min_level = INT_MAX;
- int err;
+ int err = 0;
u32 qpn;
if (root->root_ft)
@@ -802,11 +974,11 @@
if (list_empty(&root->underlay_qpns)) {
/* Don't set any QPN (zero) in case QPN list is empty */
qpn = 0;
- err = root->cmds->update_root_ft(root->dev, ft, qpn, false);
+ err = root->cmds->update_root_ft(root, ft, qpn, false);
} else {
list_for_each_entry(uqp, &root->underlay_qpns, list) {
qpn = uqp->qpn;
- err = root->cmds->update_root_ft(root->dev, ft,
+ err = root->cmds->update_root_ft(root, ft,
qpn, false);
if (err)
break;
@@ -836,15 +1008,15 @@
fs_get_obj(fte, rule->node.parent);
if (!(fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST))
return -EINVAL;
- down_write_ref_node(&fte->node);
+ down_write_ref_node(&fte->node, false);
fs_get_obj(fg, fte->node.parent);
fs_get_obj(ft, fg->node.parent);
memcpy(&rule->dest_attr, dest, sizeof(*dest));
root = find_root(&ft->node);
- err = root->cmds->update_fte(get_dev(&ft->node), ft, fg->id,
+ err = root->cmds->update_fte(root, ft, fg,
modify_mask, fte);
- up_write_ref_node(&fte->node);
+ up_write_ref_node(&fte->node, false);
return err;
}
@@ -893,6 +1065,10 @@
list_splice_init(&old_next_ft->fwd_rules, &new_next_ft->fwd_rules);
mutex_unlock(&old_next_ft->lock);
list_for_each_entry(iter, &new_next_ft->fwd_rules, next_ft) {
+ if ((iter->sw_action & MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_NS) &&
+ iter->ft->ns == new_next_ft->ns)
+ continue;
+
err = _mlx5_modify_rule_destination(iter, &dest);
if (err)
pr_err("mlx5_core: failed to modify rule to point on flow table %d\n",
@@ -916,7 +1092,7 @@
if (err)
return err;
- next_ft = first_ft ? first_ft : find_next_chained_ft(prio);
+ next_ft = first_ft ? first_ft : find_next_chained_ft(&prio->node);
err = connect_fwd_rules(dev, ft, next_ft);
if (err)
return err;
@@ -948,7 +1124,8 @@
u16 vport)
{
struct mlx5_flow_root_namespace *root = find_root(&ns->node);
- struct mlx5_flow_table *next_ft = NULL;
+ bool unmanaged = ft_attr->flags & MLX5_FLOW_TABLE_UNMANAGED;
+ struct mlx5_flow_table *next_ft;
struct fs_prio *fs_prio = NULL;
struct mlx5_flow_table *ft;
int log_table_sz;
@@ -965,14 +1142,21 @@
err = -EINVAL;
goto unlock_root;
}
- if (ft_attr->level >= fs_prio->num_levels) {
- err = -ENOSPC;
- goto unlock_root;
+ if (!unmanaged) {
+ /* The level is related to the
+ * priority level range.
+ */
+ if (ft_attr->level >= fs_prio->num_levels) {
+ err = -ENOSPC;
+ goto unlock_root;
+ }
+
+ ft_attr->level += fs_prio->start_level;
}
+
/* The level is related to the
* priority level range.
*/
- ft_attr->level += fs_prio->start_level;
ft = alloc_flow_table(ft_attr->level,
vport,
ft_attr->max_fte ? roundup_pow_of_two(ft_attr->max_fte) : 0,
@@ -985,26 +1169,35 @@
tree_init_node(&ft->node, del_hw_flow_table, del_sw_flow_table);
log_table_sz = ft->max_fte ? ilog2(ft->max_fte) : 0;
- next_ft = find_next_chained_ft(fs_prio);
- err = root->cmds->create_flow_table(root->dev, ft->vport, ft->op_mod,
- ft->type, ft->level, log_table_sz,
- next_ft, &ft->id, ft->flags);
+ next_ft = unmanaged ? ft_attr->next_ft :
+ find_next_chained_ft(&fs_prio->node);
+ ft->def_miss_action = ns->def_miss_action;
+ ft->ns = ns;
+ err = root->cmds->create_flow_table(root, ft, log_table_sz, next_ft);
if (err)
goto free_ft;
- err = connect_flow_table(root->dev, ft, fs_prio);
- if (err)
- goto destroy_ft;
+ if (!unmanaged) {
+ err = connect_flow_table(root->dev, ft, fs_prio);
+ if (err)
+ goto destroy_ft;
+ }
+
ft->node.active = true;
- down_write_ref_node(&fs_prio->node);
- tree_add_node(&ft->node, &fs_prio->node);
- list_add_flow_table(ft, fs_prio);
+ down_write_ref_node(&fs_prio->node, false);
+ if (!unmanaged) {
+ tree_add_node(&ft->node, &fs_prio->node);
+ list_add_flow_table(ft, fs_prio);
+ } else {
+ ft->node.root = fs_prio->node.root;
+ }
fs_prio->num_ft++;
- up_write_ref_node(&fs_prio->node);
+ up_write_ref_node(&fs_prio->node, false);
mutex_unlock(&root->chain_lock);
+ trace_mlx5_fs_add_ft(ft);
return ft;
destroy_ft:
- root->cmds->destroy_flow_table(root->dev, ft);
+ root->cmds->destroy_flow_table(root, ft);
free_ft:
rhltable_destroy(&ft->fgs_hash);
kfree(ft);
@@ -1046,31 +1239,27 @@
struct mlx5_flow_table*
mlx5_create_auto_grouped_flow_table(struct mlx5_flow_namespace *ns,
- int prio,
- int num_flow_table_entries,
- int max_num_groups,
- u32 level,
- u32 flags)
+ struct mlx5_flow_table_attr *ft_attr)
{
- struct mlx5_flow_table_attr ft_attr = {};
+ int num_reserved_entries = ft_attr->autogroup.num_reserved_entries;
+ int autogroups_max_fte = ft_attr->max_fte - num_reserved_entries;
+ int max_num_groups = ft_attr->autogroup.max_num_groups;
struct mlx5_flow_table *ft;
- if (max_num_groups > num_flow_table_entries)
+ if (max_num_groups > autogroups_max_fte)
+ return ERR_PTR(-EINVAL);
+ if (num_reserved_entries > ft_attr->max_fte)
return ERR_PTR(-EINVAL);
- ft_attr.max_fte = num_flow_table_entries;
- ft_attr.prio = prio;
- ft_attr.level = level;
- ft_attr.flags = flags;
-
- ft = mlx5_create_flow_table(ns, &ft_attr);
+ ft = mlx5_create_flow_table(ns, ft_attr);
if (IS_ERR(ft))
return ft;
ft->autogroup.active = true;
ft->autogroup.required_groups = max_num_groups;
+ ft->autogroup.max_fte = autogroups_max_fte;
/* We save place for flow groups in addition to max types */
- ft->autogroup.group_size = ft->max_fte / (max_num_groups + 1);
+ ft->autogroup.group_size = autogroups_max_fte / (max_num_groups + 1);
return ft;
}
@@ -1089,24 +1278,23 @@
start_flow_index);
int end_index = MLX5_GET(create_flow_group_in, fg_in,
end_flow_index);
- struct mlx5_core_dev *dev = get_dev(&ft->node);
struct mlx5_flow_group *fg;
int err;
- if (ft->autogroup.active)
+ if (ft->autogroup.active && start_index < ft->autogroup.max_fte)
return ERR_PTR(-EPERM);
- down_write_ref_node(&ft->node);
+ down_write_ref_node(&ft->node, false);
fg = alloc_insert_flow_group(ft, match_criteria_enable, match_criteria,
start_index, end_index,
ft->node.children.prev);
- up_write_ref_node(&ft->node);
+ up_write_ref_node(&ft->node, false);
if (IS_ERR(fg))
return fg;
- err = root->cmds->create_flow_group(dev, ft, fg_in, &fg->id);
+ err = root->cmds->create_flow_group(root, ft, fg_in, fg);
if (err) {
- tree_put_node(&fg->node);
+ tree_put_node(&fg->node, false);
return ERR_PTR(err);
}
trace_mlx5_fs_add_fg(fg);
@@ -1244,17 +1432,15 @@
fs_get_obj(ft, fg->node.parent);
root = find_root(&fg->node);
if (!(fte->status & FS_FTE_STATUS_EXISTING))
- err = root->cmds->create_fte(get_dev(&ft->node),
- ft, fg, fte);
+ err = root->cmds->create_fte(root, ft, fg, fte);
else
- err = root->cmds->update_fte(get_dev(&ft->node), ft, fg->id,
- modify_mask, fte);
+ err = root->cmds->update_fte(root, ft, fg, modify_mask, fte);
if (err)
goto free_handle;
fte->node.active = true;
fte->status |= FS_FTE_STATUS_EXISTING;
- atomic_inc(&fte->node.version);
+ atomic_inc(&fg->node.version);
out:
return handle;
@@ -1265,12 +1451,13 @@
}
static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft,
- struct mlx5_flow_spec *spec)
+ const struct mlx5_flow_spec *spec)
{
struct list_head *prev = &ft->node.children;
- struct mlx5_flow_group *fg;
+ u32 max_fte = ft->autogroup.max_fte;
unsigned int candidate_index = 0;
unsigned int group_size = 0;
+ struct mlx5_flow_group *fg;
if (!ft->autogroup.active)
return ERR_PTR(-ENOENT);
@@ -1278,7 +1465,7 @@
if (ft->autogroup.num_groups < ft->autogroup.required_groups)
group_size = ft->autogroup.group_size;
- /* ft->max_fte == ft->autogroup.max_types */
+ /* max_fte == ft->autogroup.max_types */
if (group_size == 0)
group_size = 1;
@@ -1291,7 +1478,7 @@
prev = &fg->node.list;
}
- if (candidate_index + group_size > ft->max_fte)
+ if (candidate_index + group_size > max_fte)
return ERR_PTR(-ENOSPC);
fg = alloc_insert_flow_group(ft,
@@ -1314,7 +1501,6 @@
struct mlx5_flow_group *fg)
{
struct mlx5_flow_root_namespace *root = find_root(&ft->node);
- 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;
@@ -1344,7 +1530,7 @@
memcpy(match_criteria_addr, fg->mask.match_criteria,
sizeof(fg->mask.match_criteria));
- err = root->cmds->create_flow_group(dev, ft, in, &fg->id);
+ err = root->cmds->create_flow_group(root, ft, in, fg);
if (!err) {
fg->node.active = true;
trace_mlx5_fs_add_fg(fg);
@@ -1359,7 +1545,13 @@
{
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->vport.flags == d2->vport.flags &&
+ ((d1->vport.flags & MLX5_FLOW_DEST_VPORT_VHCA_ID) ?
+ (d1->vport.vhca_id == d2->vport.vhca_id) : true) &&
+ ((d1->vport.flags & MLX5_FLOW_DEST_VPORT_REFORMAT_ID) ?
+ (d1->vport.pkt_reformat->id ==
+ d2->vport.pkt_reformat->id) : true)) ||
(d1->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE &&
d1->ft == d2->ft) ||
(d1->type == MLX5_FLOW_DESTINATION_TYPE_TIR &&
@@ -1384,9 +1576,22 @@
return NULL;
}
-static bool check_conflicting_actions(u32 action1, u32 action2)
+static bool check_conflicting_actions_vlan(const struct mlx5_fs_vlan *vlan0,
+ const struct mlx5_fs_vlan *vlan1)
{
- u32 xored_actions = action1 ^ action2;
+ return vlan0->ethtype != vlan1->ethtype ||
+ vlan0->vid != vlan1->vid ||
+ vlan0->prio != vlan1->prio;
+}
+
+static bool check_conflicting_actions(const struct mlx5_flow_act *act1,
+ const struct mlx5_flow_act *act2)
+{
+ u32 action1 = act1->action;
+ u32 action2 = act2->action;
+ u32 xored_actions;
+
+ xored_actions = action1 ^ action2;
/* if one rule only wants to count, it's ok */
if (action1 == MLX5_FLOW_CONTEXT_ACTION_COUNT ||
@@ -1394,7 +1599,7 @@
return false;
if (xored_actions & (MLX5_FLOW_CONTEXT_ACTION_DROP |
- MLX5_FLOW_CONTEXT_ACTION_ENCAP |
+ MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT |
MLX5_FLOW_CONTEXT_ACTION_DECAP |
MLX5_FLOW_CONTEXT_ACTION_MOD_HDR |
MLX5_FLOW_CONTEXT_ACTION_VLAN_POP |
@@ -1403,23 +1608,41 @@
MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH_2))
return true;
+ if (action1 & MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT &&
+ act1->pkt_reformat != act2->pkt_reformat)
+ return true;
+
+ if (action1 & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR &&
+ act1->modify_hdr != act2->modify_hdr)
+ return true;
+
+ if (action1 & MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH &&
+ check_conflicting_actions_vlan(&act1->vlan[0], &act2->vlan[0]))
+ return true;
+
+ if (action1 & MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH_2 &&
+ check_conflicting_actions_vlan(&act1->vlan[1], &act2->vlan[1]))
+ return true;
+
return false;
}
-static int check_conflicting_ftes(struct fs_fte *fte, const struct mlx5_flow_act *flow_act)
+static int check_conflicting_ftes(struct fs_fte *fte,
+ const struct mlx5_flow_context *flow_context,
+ const struct mlx5_flow_act *flow_act)
{
- if (check_conflicting_actions(flow_act->action, fte->action.action)) {
+ if (check_conflicting_actions(flow_act, &fte->action)) {
mlx5_core_warn(get_dev(&fte->node),
"Found two FTEs with conflicting actions\n");
return -EEXIST;
}
- if (flow_act->has_flow_tag &&
- fte->action.flow_tag != flow_act->flow_tag) {
+ if ((flow_context->flags & FLOW_CONTEXT_HAS_TAG) &&
+ fte->flow_context.flow_tag != flow_context->flow_tag) {
mlx5_core_warn(get_dev(&fte->node),
"FTE flow tag %u already exists with different flow tag %u\n",
- fte->action.flow_tag,
- flow_act->flow_tag);
+ fte->flow_context.flow_tag,
+ flow_context->flow_tag);
return -EEXIST;
}
@@ -1427,7 +1650,7 @@
}
static struct mlx5_flow_handle *add_rule_fg(struct mlx5_flow_group *fg,
- u32 *match_value,
+ const struct mlx5_flow_spec *spec,
struct mlx5_flow_act *flow_act,
struct mlx5_flow_destination *dest,
int dest_num,
@@ -1438,7 +1661,7 @@
int i;
int ret;
- ret = check_conflicting_ftes(fte, flow_act);
+ ret = check_conflicting_ftes(fte, &spec->flow_context, flow_act);
if (ret)
return ERR_PTR(ret);
@@ -1461,46 +1684,39 @@
return handle;
}
-struct mlx5_fc *mlx5_flow_rule_counter(struct mlx5_flow_handle *handle)
+static bool counter_is_valid(u32 action)
{
- struct mlx5_flow_rule *dst;
- struct fs_fte *fte;
-
- fs_get_obj(fte, handle->rule[0]->node.parent);
-
- fs_for_each_dst(dst, fte) {
- if (dst->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_COUNTER)
- return dst->dest_attr.counter;
- }
-
- return NULL;
-}
-
-static bool counter_is_valid(struct mlx5_fc *counter, u32 action)
-{
- if (!(action & MLX5_FLOW_CONTEXT_ACTION_COUNT))
- return !counter;
-
- if (!counter)
- return false;
-
return (action & (MLX5_FLOW_CONTEXT_ACTION_DROP |
+ MLX5_FLOW_CONTEXT_ACTION_ALLOW |
MLX5_FLOW_CONTEXT_ACTION_FWD_DEST));
}
static bool dest_is_valid(struct mlx5_flow_destination *dest,
- u32 action,
+ struct mlx5_flow_act *flow_act,
struct mlx5_flow_table *ft)
{
+ bool ignore_level = flow_act->flags & FLOW_ACT_IGNORE_FLOW_LEVEL;
+ u32 action = flow_act->action;
+
if (dest && (dest->type == MLX5_FLOW_DESTINATION_TYPE_COUNTER))
- return counter_is_valid(dest->counter, action);
+ return counter_is_valid(action);
if (!(action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST))
return true;
+ if (ignore_level) {
+ if (ft->type != FS_FT_FDB &&
+ ft->type != FS_FT_NIC_RX)
+ return false;
+
+ if (dest->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE &&
+ ft->type != dest->ft->type)
+ return false;
+ }
+
if (!dest || ((dest->type ==
MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE) &&
- (dest->ft->level <= ft->level)))
+ (dest->ft->level <= ft->level && !ignore_level)))
return false;
return true;
}
@@ -1510,30 +1726,22 @@
struct mlx5_flow_group *g;
};
-struct match_list_head {
- struct list_head list;
- struct match_list first;
-};
-
-static void free_match_list(struct match_list_head *head)
+static void free_match_list(struct match_list *head, bool ft_locked)
{
- if (!list_empty(&head->list)) {
- struct match_list *iter, *match_tmp;
+ struct match_list *iter, *match_tmp;
- list_del(&head->first.list);
- tree_put_node(&head->first.g->node);
- list_for_each_entry_safe(iter, match_tmp, &head->list,
- list) {
- tree_put_node(&iter->g->node);
- list_del(&iter->list);
- kfree(iter);
- }
+ list_for_each_entry_safe(iter, match_tmp, &head->list,
+ list) {
+ tree_put_node(&iter->g->node, ft_locked);
+ list_del(&iter->list);
+ kfree(iter);
}
}
-static int build_match_list(struct match_list_head *match_head,
+static int build_match_list(struct match_list *match_head,
struct mlx5_flow_table *ft,
- struct mlx5_flow_spec *spec)
+ const struct mlx5_flow_spec *spec,
+ bool ft_locked)
{
struct rhlist_head *tmp, *list;
struct mlx5_flow_group *g;
@@ -1547,24 +1755,14 @@
rhl_for_each_entry_rcu(g, tmp, list, hash) {
struct match_list *curr_match;
- if (likely(list_empty(&match_head->list))) {
- if (!tree_get_node(&g->node))
- continue;
- match_head->first.g = g;
- list_add_tail(&match_head->first.list,
- &match_head->list);
+ if (unlikely(!tree_get_node(&g->node)))
continue;
- }
curr_match = kmalloc(sizeof(*curr_match), GFP_ATOMIC);
if (!curr_match) {
rcu_read_unlock();
- free_match_list(match_head);
+ free_match_list(match_head, ft_locked);
return -ENOMEM;
- }
- if (!tree_get_node(&g->node)) {
- kfree(curr_match);
- continue;
}
curr_match->g = g;
list_add_tail(&curr_match->list, &match_head->list);
@@ -1585,7 +1783,7 @@
static struct fs_fte *
lookup_fte_locked(struct mlx5_flow_group *g,
- u32 *match_value,
+ const u32 *match_value,
bool take_write)
{
struct fs_fte *fte_tmp;
@@ -1601,7 +1799,7 @@
goto out;
}
if (!fte_tmp->node.active) {
- tree_put_node(&fte_tmp->node);
+ tree_put_node(&fte_tmp->node, false);
fte_tmp = NULL;
goto out;
}
@@ -1609,7 +1807,7 @@
nested_down_write_ref_node(&fte_tmp->node, FS_LOCK_CHILD);
out:
if (take_write)
- up_write_ref_node(&g->node);
+ up_write_ref_node(&g->node, false);
else
up_read_ref_node(&g->node);
return fte_tmp;
@@ -1618,7 +1816,7 @@
static struct mlx5_flow_handle *
try_add_to_existing_fg(struct mlx5_flow_table *ft,
struct list_head *match_head,
- struct mlx5_flow_spec *spec,
+ const struct mlx5_flow_spec *spec,
struct mlx5_flow_act *flow_act,
struct mlx5_flow_destination *dest,
int dest_num,
@@ -1630,16 +1828,20 @@
struct match_list *iter;
bool take_write = false;
struct fs_fte *fte;
- u64 version;
+ u64 version = 0;
int err;
- fte = alloc_fte(ft, spec->match_value, flow_act);
+ fte = alloc_fte(ft, spec, flow_act);
if (IS_ERR(fte))
return ERR_PTR(-ENOMEM);
search_again_locked:
+ if (flow_act->flags & FLOW_ACT_NO_APPEND)
+ goto skip_search;
version = matched_fgs_get_version(match_head);
- /* Try to find a fg that already contains a matching fte */
+ /* Try to find an fte with identical match value and attempt update its
+ * action.
+ */
list_for_each_entry(iter, match_head, list) {
struct fs_fte *fte_tmp;
@@ -1647,13 +1849,18 @@
fte_tmp = lookup_fte_locked(g, spec->match_value, take_write);
if (!fte_tmp)
continue;
- rule = add_rule_fg(g, spec->match_value,
- flow_act, dest, dest_num, fte_tmp);
- up_write_ref_node(&fte_tmp->node);
- tree_put_node(&fte_tmp->node);
+ rule = add_rule_fg(g, spec, flow_act, dest, dest_num, fte_tmp);
+ /* No error check needed here, because insert_fte() is not called */
+ up_write_ref_node(&fte_tmp->node, false);
+ tree_put_node(&fte_tmp->node, false);
kmem_cache_free(steering->ftes_cache, fte);
return rule;
}
+
+skip_search:
+ /* No group with matching fte found, or we skipped the search.
+ * Try to add a new fte to any matching fg.
+ */
/* Check the ft version, for case that new flow group
* was added while the fgs weren't locked
@@ -1663,10 +1870,12 @@
goto out;
}
- /* Check the fgs version, for case the new FTE with the
- * same values was added while the fgs weren't locked
+ /* Check the fgs version. If version have changed it could be that an
+ * FTE with the same match value was added while the fgs weren't
+ * locked.
*/
- if (version != matched_fgs_get_version(match_head)) {
+ if (!(flow_act->flags & FLOW_ACT_NO_APPEND) &&
+ version != matched_fgs_get_version(match_head)) {
take_write = true;
goto search_again_locked;
}
@@ -1674,14 +1883,16 @@
list_for_each_entry(iter, match_head, list) {
g = iter->g;
- if (!g->node.active)
- continue;
-
nested_down_write_ref_node(&g->node, FS_LOCK_PARENT);
+
+ if (!g->node.active) {
+ up_write_ref_node(&g->node, false);
+ continue;
+ }
err = insert_fte(g, fte);
if (err) {
- up_write_ref_node(&g->node);
+ up_write_ref_node(&g->node, false);
if (err == -ENOSPC)
continue;
kmem_cache_free(steering->ftes_cache, fte);
@@ -1689,11 +1900,11 @@
}
nested_down_write_ref_node(&fte->node, FS_LOCK_CHILD);
- up_write_ref_node(&g->node);
- rule = add_rule_fg(g, spec->match_value,
- flow_act, dest, dest_num, fte);
- up_write_ref_node(&fte->node);
- tree_put_node(&fte->node);
+ up_write_ref_node(&g->node, false);
+ rule = add_rule_fg(g, spec, flow_act, dest, dest_num, fte);
+ up_write_ref_node(&fte->node, false);
+ if (IS_ERR(rule))
+ tree_put_node(&fte->node, false);
return rule;
}
rule = ERR_PTR(-ENOENT);
@@ -1704,16 +1915,16 @@
static struct mlx5_flow_handle *
_mlx5_add_flow_rules(struct mlx5_flow_table *ft,
- struct mlx5_flow_spec *spec,
+ const struct mlx5_flow_spec *spec,
struct mlx5_flow_act *flow_act,
struct mlx5_flow_destination *dest,
int dest_num)
{
struct mlx5_flow_steering *steering = get_steering(&ft->node);
- struct mlx5_flow_group *g;
struct mlx5_flow_handle *rule;
- struct match_list_head match_head;
+ struct match_list match_head;
+ struct mlx5_flow_group *g;
bool take_write = false;
struct fs_fte *fte;
int version;
@@ -1724,7 +1935,7 @@
return ERR_PTR(-EINVAL);
for (i = 0; i < dest_num; i++) {
- if (!dest_is_valid(&dest[i], flow_act->action, ft))
+ if (!dest_is_valid(&dest[i], flow_act, ft))
return ERR_PTR(-EINVAL);
}
nested_down_read_ref_node(&ft->node, FS_LOCK_GRANDPARENT);
@@ -1732,10 +1943,10 @@
version = atomic_read(&ft->node.version);
/* Collect all fgs which has a matching match_criteria */
- err = build_match_list(&match_head, ft, spec);
+ err = build_match_list(&match_head, ft, spec, take_write);
if (err) {
if (take_write)
- up_write_ref_node(&ft->node);
+ up_write_ref_node(&ft->node, false);
else
up_read_ref_node(&ft->node);
return ERR_PTR(err);
@@ -1746,11 +1957,11 @@
rule = try_add_to_existing_fg(ft, &match_head.list, spec, flow_act, dest,
dest_num, version);
- free_match_list(&match_head);
+ free_match_list(&match_head, take_write);
if (!IS_ERR(rule) ||
(PTR_ERR(rule) != -ENOENT && PTR_ERR(rule) != -EAGAIN)) {
if (take_write)
- up_write_ref_node(&ft->node);
+ up_write_ref_node(&ft->node, false);
return rule;
}
@@ -1766,41 +1977,42 @@
g = alloc_auto_flow_group(ft, spec);
if (IS_ERR(g)) {
rule = ERR_CAST(g);
- up_write_ref_node(&ft->node);
+ up_write_ref_node(&ft->node, false);
return rule;
}
+ fte = alloc_fte(ft, spec, flow_act);
+ if (IS_ERR(fte)) {
+ up_write_ref_node(&ft->node, false);
+ err = PTR_ERR(fte);
+ goto err_alloc_fte;
+ }
+
nested_down_write_ref_node(&g->node, FS_LOCK_PARENT);
- up_write_ref_node(&ft->node);
+ up_write_ref_node(&ft->node, false);
err = create_auto_flow_group(ft, g);
if (err)
goto err_release_fg;
- fte = alloc_fte(ft, spec->match_value, flow_act);
- if (IS_ERR(fte)) {
- err = PTR_ERR(fte);
- goto err_release_fg;
- }
-
err = insert_fte(g, fte);
- if (err) {
- kmem_cache_free(steering->ftes_cache, fte);
+ if (err)
goto err_release_fg;
- }
nested_down_write_ref_node(&fte->node, FS_LOCK_CHILD);
- up_write_ref_node(&g->node);
- rule = add_rule_fg(g, spec->match_value, flow_act, dest,
- dest_num, fte);
- up_write_ref_node(&fte->node);
- tree_put_node(&fte->node);
- tree_put_node(&g->node);
+ up_write_ref_node(&g->node, false);
+ rule = add_rule_fg(g, spec, flow_act, dest, dest_num, fte);
+ up_write_ref_node(&fte->node, false);
+ if (IS_ERR(rule))
+ tree_put_node(&fte->node, false);
+ tree_put_node(&g->node, false);
return rule;
err_release_fg:
- up_write_ref_node(&g->node);
- tree_put_node(&g->node);
+ up_write_ref_node(&g->node, false);
+ kmem_cache_free(steering->ftes_cache, fte);
+err_alloc_fte:
+ tree_put_node(&g->node, false);
return ERR_PTR(err);
}
@@ -1812,61 +2024,104 @@
struct mlx5_flow_handle *
mlx5_add_flow_rules(struct mlx5_flow_table *ft,
- struct mlx5_flow_spec *spec,
+ const struct mlx5_flow_spec *spec,
struct mlx5_flow_act *flow_act,
struct mlx5_flow_destination *dest,
int num_dest)
{
struct mlx5_flow_root_namespace *root = find_root(&ft->node);
- struct mlx5_flow_destination gen_dest = {};
+ static const struct mlx5_flow_spec zero_spec = {};
+ struct mlx5_flow_destination *gen_dest = NULL;
struct mlx5_flow_table *next_ft = NULL;
struct mlx5_flow_handle *handle = NULL;
u32 sw_action = flow_act->action;
- struct fs_prio *prio;
+ int i;
- fs_get_obj(prio, ft->node.parent);
- if (flow_act->action == MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO) {
- if (!fwd_next_prio_supported(ft))
- return ERR_PTR(-EOPNOTSUPP);
- if (num_dest)
- return ERR_PTR(-EINVAL);
- mutex_lock(&root->chain_lock);
- next_ft = find_next_chained_ft(prio);
- if (next_ft) {
- gen_dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
- gen_dest.ft = next_ft;
- dest = &gen_dest;
- num_dest = 1;
- flow_act->action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
- } else {
- mutex_unlock(&root->chain_lock);
- return ERR_PTR(-EOPNOTSUPP);
- }
+ if (!spec)
+ spec = &zero_spec;
+
+ if (!is_fwd_next_action(sw_action))
+ return _mlx5_add_flow_rules(ft, spec, flow_act, dest, num_dest);
+
+ if (!fwd_next_prio_supported(ft))
+ return ERR_PTR(-EOPNOTSUPP);
+
+ mutex_lock(&root->chain_lock);
+ next_ft = find_next_fwd_ft(ft, flow_act);
+ if (!next_ft) {
+ handle = ERR_PTR(-EOPNOTSUPP);
+ goto unlock;
}
+ gen_dest = kcalloc(num_dest + 1, sizeof(*dest),
+ GFP_KERNEL);
+ if (!gen_dest) {
+ handle = ERR_PTR(-ENOMEM);
+ goto unlock;
+ }
+ for (i = 0; i < num_dest; i++)
+ gen_dest[i] = dest[i];
+ gen_dest[i].type =
+ MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
+ gen_dest[i].ft = next_ft;
+ dest = gen_dest;
+ num_dest++;
+ flow_act->action &= ~(MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO |
+ MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_NS);
+ flow_act->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
handle = _mlx5_add_flow_rules(ft, spec, flow_act, dest, num_dest);
+ if (IS_ERR(handle))
+ goto unlock;
- if (sw_action == MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO) {
- if (!IS_ERR_OR_NULL(handle) &&
- (list_empty(&handle->rule[0]->next_ft))) {
- mutex_lock(&next_ft->lock);
- list_add(&handle->rule[0]->next_ft,
- &next_ft->fwd_rules);
- mutex_unlock(&next_ft->lock);
- handle->rule[0]->sw_action = MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO;
- }
- mutex_unlock(&root->chain_lock);
+ if (list_empty(&handle->rule[num_dest - 1]->next_ft)) {
+ mutex_lock(&next_ft->lock);
+ list_add(&handle->rule[num_dest - 1]->next_ft,
+ &next_ft->fwd_rules);
+ mutex_unlock(&next_ft->lock);
+ handle->rule[num_dest - 1]->sw_action = sw_action;
+ handle->rule[num_dest - 1]->ft = ft;
}
+unlock:
+ mutex_unlock(&root->chain_lock);
+ kfree(gen_dest);
return handle;
}
EXPORT_SYMBOL(mlx5_add_flow_rules);
void mlx5_del_flow_rules(struct mlx5_flow_handle *handle)
{
+ struct fs_fte *fte;
int i;
+ /* In order to consolidate the HW changes we lock the FTE for other
+ * changes, and increase its refcount, in order not to perform the
+ * "del" functions of the FTE. Will handle them here.
+ * The removal of the rules is done under locked FTE.
+ * After removing all the handle's rules, if there are remaining
+ * rules, it means we just need to modify the FTE in FW, and
+ * unlock/decrease the refcount we increased before.
+ * Otherwise, it means the FTE should be deleted. First delete the
+ * FTE in FW. Then, unlock the FTE, and proceed the tree_put_node of
+ * the FTE, which will handle the last decrease of the refcount, as
+ * well as required handling of its parent.
+ */
+ fs_get_obj(fte, handle->rule[0]->node.parent);
+ down_write_ref_node(&fte->node, false);
for (i = handle->num_rules - 1; i >= 0; i--)
- tree_remove_node(&handle->rule[i]->node);
+ tree_remove_node(&handle->rule[i]->node, true);
+ if (list_empty(&fte->node.children)) {
+ del_hw_fte(&fte->node);
+ /* Avoid double call to del_hw_fte */
+ fte->node.del_hw_func = NULL;
+ up_write_ref_node(&fte->node, false);
+ tree_put_node(&fte->node, false);
+ } else if (fte->dests_size) {
+ if (fte->modify_mask)
+ modify_fte(fte);
+ up_write_ref_node(&fte->node, false);
+ } else {
+ up_write_ref_node(&fte->node, false);
+ }
kfree(handle);
}
EXPORT_SYMBOL(mlx5_del_flow_rules);
@@ -1874,13 +2129,20 @@
/* Assuming prio->node.children(flow tables) is sorted by level */
static struct mlx5_flow_table *find_next_ft(struct mlx5_flow_table *ft)
{
+ struct fs_node *prio_parent, *child;
struct fs_prio *prio;
fs_get_obj(prio, ft->node.parent);
if (!list_is_last(&ft->node.list, &prio->node.children))
return list_next_entry(ft, node.list);
- return find_next_chained_ft(prio);
+
+ prio_parent = find_prio_chains_parent(&prio->node, &child);
+
+ if (prio_parent && list_is_first(&child->list, &prio_parent->children))
+ return find_closest_ft(&prio->node, false, false);
+
+ return find_next_chained_ft(&prio->node);
}
static int update_root_ft_destroy(struct mlx5_flow_table *ft)
@@ -1903,12 +2165,12 @@
if (list_empty(&root->underlay_qpns)) {
/* Don't set any QPN (zero) in case QPN list is empty */
qpn = 0;
- err = root->cmds->update_root_ft(root->dev, new_root_ft,
+ err = root->cmds->update_root_ft(root, new_root_ft,
qpn, false);
} else {
list_for_each_entry(uqp, &root->underlay_qpns, list) {
qpn = uqp->qpn;
- err = root->cmds->update_root_ft(root->dev,
+ err = root->cmds->update_root_ft(root,
new_root_ft, qpn,
false);
if (err)
@@ -1964,12 +2226,13 @@
int err = 0;
mutex_lock(&root->chain_lock);
- err = disconnect_flow_table(ft);
+ if (!(ft->flags & MLX5_FLOW_TABLE_UNMANAGED))
+ err = disconnect_flow_table(ft);
if (err) {
mutex_unlock(&root->chain_lock);
return err;
}
- if (tree_remove_node(&ft->node))
+ if (tree_remove_node(&ft->node, false))
mlx5_core_warn(get_dev(&ft->node), "Flow table %d wasn't destroyed, refcount > 1\n",
ft->id);
mutex_unlock(&root->chain_lock);
@@ -1980,17 +2243,29 @@
void mlx5_destroy_flow_group(struct mlx5_flow_group *fg)
{
- if (tree_remove_node(&fg->node))
+ if (tree_remove_node(&fg->node, false))
mlx5_core_warn(get_dev(&fg->node), "Flow group %d wasn't destroyed, refcount > 1\n",
fg->id);
}
+
+struct mlx5_flow_namespace *mlx5_get_fdb_sub_ns(struct mlx5_core_dev *dev,
+ int n)
+{
+ struct mlx5_flow_steering *steering = dev->priv.steering;
+
+ if (!steering || !steering->fdb_sub_ns)
+ return NULL;
+
+ return steering->fdb_sub_ns[n];
+}
+EXPORT_SYMBOL(mlx5_get_fdb_sub_ns);
struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev,
enum mlx5_flow_namespace_type type)
{
struct mlx5_flow_steering *steering = dev->priv.steering;
struct mlx5_flow_root_namespace *root_ns;
- int prio;
+ int prio = 0;
struct fs_prio *fs_prio;
struct mlx5_flow_namespace *ns;
@@ -1998,40 +2273,39 @@
return NULL;
switch (type) {
- case MLX5_FLOW_NAMESPACE_BYPASS:
- case MLX5_FLOW_NAMESPACE_LAG:
- case MLX5_FLOW_NAMESPACE_OFFLOADS:
- case MLX5_FLOW_NAMESPACE_ETHTOOL:
- case MLX5_FLOW_NAMESPACE_KERNEL:
- case MLX5_FLOW_NAMESPACE_LEFTOVERS:
- case MLX5_FLOW_NAMESPACE_ANCHOR:
- prio = type;
- break;
case MLX5_FLOW_NAMESPACE_FDB:
if (steering->fdb_root_ns)
return &steering->fdb_root_ns->ns;
- else
- return NULL;
+ return NULL;
case MLX5_FLOW_NAMESPACE_SNIFFER_RX:
if (steering->sniffer_rx_root_ns)
return &steering->sniffer_rx_root_ns->ns;
- else
- return NULL;
+ return NULL;
case MLX5_FLOW_NAMESPACE_SNIFFER_TX:
if (steering->sniffer_tx_root_ns)
return &steering->sniffer_tx_root_ns->ns;
- else
- return NULL;
- case MLX5_FLOW_NAMESPACE_EGRESS:
- if (steering->egress_root_ns)
- return &steering->egress_root_ns->ns;
- else
- return NULL;
- default:
return NULL;
+ default:
+ break;
}
- root_ns = steering->root_ns;
+ if (type == MLX5_FLOW_NAMESPACE_EGRESS ||
+ type == MLX5_FLOW_NAMESPACE_EGRESS_KERNEL) {
+ root_ns = steering->egress_root_ns;
+ prio = type - MLX5_FLOW_NAMESPACE_EGRESS;
+ } else if (type == MLX5_FLOW_NAMESPACE_RDMA_RX) {
+ root_ns = steering->rdma_rx_root_ns;
+ prio = RDMA_RX_BYPASS_PRIO;
+ } else if (type == MLX5_FLOW_NAMESPACE_RDMA_RX_KERNEL) {
+ root_ns = steering->rdma_rx_root_ns;
+ prio = RDMA_RX_KERNEL_PRIO;
+ } else if (type == MLX5_FLOW_NAMESPACE_RDMA_TX) {
+ root_ns = steering->rdma_tx_root_ns;
+ } else { /* Must be NIC RX */
+ root_ns = steering->root_ns;
+ prio = type;
+ }
+
if (!root_ns)
return NULL;
@@ -2053,7 +2327,7 @@
{
struct mlx5_flow_steering *steering = dev->priv.steering;
- if (!steering || vport >= MLX5_TOTAL_VPORTS(dev))
+ if (!steering || vport >= mlx5_eswitch_get_total_vports(dev))
return NULL;
switch (type) {
@@ -2074,8 +2348,10 @@
}
}
-static struct fs_prio *fs_create_prio(struct mlx5_flow_namespace *ns,
- unsigned int prio, int num_levels)
+static struct fs_prio *_fs_create_prio(struct mlx5_flow_namespace *ns,
+ unsigned int prio,
+ int num_levels,
+ enum fs_node_type type)
{
struct fs_prio *fs_prio;
@@ -2083,7 +2359,7 @@
if (!fs_prio)
return ERR_PTR(-ENOMEM);
- fs_prio->node.type = FS_TYPE_PRIO;
+ fs_prio->node.type = type;
tree_init_node(&fs_prio->node, NULL, del_sw_prio);
tree_add_node(&fs_prio->node, &ns->node);
fs_prio->num_levels = num_levels;
@@ -2091,6 +2367,19 @@
list_add_tail(&fs_prio->node.list, &ns->node.children);
return fs_prio;
+}
+
+static struct fs_prio *fs_create_prio_chained(struct mlx5_flow_namespace *ns,
+ unsigned int prio,
+ int num_levels)
+{
+ return _fs_create_prio(ns, prio, num_levels, FS_TYPE_PRIO_CHAINS);
+}
+
+static struct fs_prio *fs_create_prio(struct mlx5_flow_namespace *ns,
+ unsigned int prio, int num_levels)
+{
+ return _fs_create_prio(ns, prio, num_levels, FS_TYPE_PRIO);
}
static struct mlx5_flow_namespace *fs_init_namespace(struct mlx5_flow_namespace
@@ -2101,7 +2390,8 @@
return ns;
}
-static struct mlx5_flow_namespace *fs_create_namespace(struct fs_prio *prio)
+static struct mlx5_flow_namespace *fs_create_namespace(struct fs_prio *prio,
+ int def_miss_act)
{
struct mlx5_flow_namespace *ns;
@@ -2110,6 +2400,7 @@
return ERR_PTR(-ENOMEM);
fs_init_namespace(ns);
+ ns->def_miss_action = def_miss_act;
tree_init_node(&ns->node, NULL, del_sw_ns);
tree_add_node(&ns->node, &prio->node);
list_add_tail(&ns->node.list, &prio->node.children);
@@ -2176,7 +2467,7 @@
base = &fs_prio->node;
} else if (init_node->type == FS_TYPE_NAMESPACE) {
fs_get_obj(fs_prio, fs_parent_node);
- fs_ns = fs_create_namespace(fs_prio);
+ fs_ns = fs_create_namespace(fs_prio, init_node->def_miss_action);
if (IS_ERR(fs_ns))
return PTR_ERR(fs_ns);
base = &fs_ns->node;
@@ -2217,6 +2508,17 @@
return 0;
}
+static void del_sw_root_ns(struct fs_node *node)
+{
+ struct mlx5_flow_root_namespace *root_ns;
+ struct mlx5_flow_namespace *ns;
+
+ fs_get_obj(ns, node);
+ root_ns = container_of(ns, struct mlx5_flow_root_namespace, ns);
+ mutex_destroy(&root_ns->chain_lock);
+ kfree(node);
+}
+
static struct mlx5_flow_root_namespace
*create_root_ns(struct mlx5_flow_steering *steering,
enum fs_flow_table_type table_type)
@@ -2225,7 +2527,7 @@
struct mlx5_flow_root_namespace *root_ns;
struct mlx5_flow_namespace *ns;
- if (mlx5_accel_ipsec_device_caps(steering->dev) & MLX5_ACCEL_IPSEC_CAP_DEVICE &&
+ if (mlx5_fpga_ipsec_device_caps(steering->dev) & MLX5_ACCEL_IPSEC_CAP_DEVICE &&
(table_type == FS_FT_NIC_RX || table_type == FS_FT_NIC_TX))
cmds = mlx5_fs_cmd_get_default_ipsec_fpga_cmds(table_type);
@@ -2243,7 +2545,7 @@
ns = &root_ns->ns;
fs_init_namespace(ns);
mutex_init(&root_ns->chain_lock);
- tree_init_node(&ns->node, NULL, NULL);
+ tree_init_node(&ns->node, NULL, del_sw_root_ns);
tree_add_node(&ns->node, NULL);
return root_ns;
@@ -2269,9 +2571,17 @@
int acc_level_ns = acc_level;
prio->start_level = acc_level;
- fs_for_each_ns(ns, prio)
+ fs_for_each_ns(ns, prio) {
/* This updates start_level and num_levels of ns's priority descendants */
acc_level_ns = set_prio_attrs_in_ns(ns, acc_level);
+
+ /* If this a prio with chains, and we can jump from one chain
+ * (namepsace) to another, so we accumulate the levels
+ */
+ if (prio->node.type == FS_TYPE_PRIO_CHAINS)
+ acc_level = acc_level_ns;
+ }
+
if (!prio->num_levels)
prio->num_levels = acc_level_ns - prio->start_level;
WARN_ON(prio->num_levels < acc_level_ns - prio->start_level);
@@ -2348,8 +2658,8 @@
tree_get_node(node);
list_for_each_entry_safe(iter, temp, &node->children, list)
clean_tree(iter);
- tree_put_node(node);
- tree_remove_node(node);
+ tree_put_node(node, false);
+ tree_remove_node(node, false);
}
}
@@ -2369,7 +2679,7 @@
if (!steering->esw_egress_root_ns)
return;
- for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++)
+ for (i = 0; i < mlx5_eswitch_get_total_vports(dev); i++)
cleanup_root_ns(steering->esw_egress_root_ns[i]);
kfree(steering->esw_egress_root_ns);
@@ -2384,7 +2694,7 @@
if (!steering->esw_ingress_root_ns)
return;
- for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++)
+ for (i = 0; i < mlx5_eswitch_get_total_vports(dev); i++)
cleanup_root_ns(steering->esw_ingress_root_ns[i]);
kfree(steering->esw_ingress_root_ns);
@@ -2399,8 +2709,13 @@
cleanup_egress_acls_root_ns(dev);
cleanup_ingress_acls_root_ns(dev);
cleanup_root_ns(steering->fdb_root_ns);
+ steering->fdb_root_ns = NULL;
+ kfree(steering->fdb_sub_ns);
+ steering->fdb_sub_ns = NULL;
cleanup_root_ns(steering->sniffer_rx_root_ns);
cleanup_root_ns(steering->sniffer_tx_root_ns);
+ cleanup_root_ns(steering->rdma_rx_root_ns);
+ cleanup_root_ns(steering->rdma_tx_root_ns);
cleanup_root_ns(steering->egress_root_ns);
mlx5_cleanup_fc_stats(dev);
kmem_cache_destroy(steering->ftes_cache);
@@ -2418,11 +2733,7 @@
/* Create single prio */
prio = fs_create_prio(&steering->sniffer_tx_root_ns->ns, 0, 1);
- if (IS_ERR(prio)) {
- cleanup_root_ns(steering->sniffer_tx_root_ns);
- return PTR_ERR(prio);
- }
- return 0;
+ return PTR_ERR_OR_ZERO(prio);
}
static int init_sniffer_rx_root_ns(struct mlx5_flow_steering *steering)
@@ -2435,36 +2746,184 @@
/* Create single prio */
prio = fs_create_prio(&steering->sniffer_rx_root_ns->ns, 0, 1);
- if (IS_ERR(prio)) {
- cleanup_root_ns(steering->sniffer_rx_root_ns);
- return PTR_ERR(prio);
+ return PTR_ERR_OR_ZERO(prio);
+}
+
+static int init_rdma_rx_root_ns(struct mlx5_flow_steering *steering)
+{
+ int err;
+
+ steering->rdma_rx_root_ns = create_root_ns(steering, FS_FT_RDMA_RX);
+ if (!steering->rdma_rx_root_ns)
+ return -ENOMEM;
+
+ err = init_root_tree(steering, &rdma_rx_root_fs,
+ &steering->rdma_rx_root_ns->ns.node);
+ if (err)
+ goto out_err;
+
+ set_prio_attrs(steering->rdma_rx_root_ns);
+
+ return 0;
+
+out_err:
+ cleanup_root_ns(steering->rdma_rx_root_ns);
+ steering->rdma_rx_root_ns = NULL;
+ return err;
+}
+
+static int init_rdma_tx_root_ns(struct mlx5_flow_steering *steering)
+{
+ int err;
+
+ steering->rdma_tx_root_ns = create_root_ns(steering, FS_FT_RDMA_TX);
+ if (!steering->rdma_tx_root_ns)
+ return -ENOMEM;
+
+ err = init_root_tree(steering, &rdma_tx_root_fs,
+ &steering->rdma_tx_root_ns->ns.node);
+ if (err)
+ goto out_err;
+
+ set_prio_attrs(steering->rdma_tx_root_ns);
+
+ return 0;
+
+out_err:
+ cleanup_root_ns(steering->rdma_tx_root_ns);
+ steering->rdma_tx_root_ns = NULL;
+ return err;
+}
+
+/* FT and tc chains are stored in the same array so we can re-use the
+ * mlx5_get_fdb_sub_ns() and tc api for FT chains.
+ * When creating a new ns for each chain store it in the first available slot.
+ * Assume tc chains are created and stored first and only then the FT chain.
+ */
+static void store_fdb_sub_ns_prio_chain(struct mlx5_flow_steering *steering,
+ struct mlx5_flow_namespace *ns)
+{
+ int chain = 0;
+
+ while (steering->fdb_sub_ns[chain])
+ ++chain;
+
+ steering->fdb_sub_ns[chain] = ns;
+}
+
+static int create_fdb_sub_ns_prio_chain(struct mlx5_flow_steering *steering,
+ struct fs_prio *maj_prio)
+{
+ struct mlx5_flow_namespace *ns;
+ struct fs_prio *min_prio;
+ int prio;
+
+ ns = fs_create_namespace(maj_prio, MLX5_FLOW_TABLE_MISS_ACTION_DEF);
+ if (IS_ERR(ns))
+ return PTR_ERR(ns);
+
+ for (prio = 0; prio < FDB_TC_MAX_PRIO; prio++) {
+ min_prio = fs_create_prio(ns, prio, FDB_TC_LEVELS_PER_PRIO);
+ if (IS_ERR(min_prio))
+ return PTR_ERR(min_prio);
}
+
+ store_fdb_sub_ns_prio_chain(steering, ns);
+
+ return 0;
+}
+
+static int create_fdb_chains(struct mlx5_flow_steering *steering,
+ int fs_prio,
+ int chains)
+{
+ struct fs_prio *maj_prio;
+ int levels;
+ int chain;
+ int err;
+
+ levels = FDB_TC_LEVELS_PER_PRIO * FDB_TC_MAX_PRIO * chains;
+ maj_prio = fs_create_prio_chained(&steering->fdb_root_ns->ns,
+ fs_prio,
+ levels);
+ if (IS_ERR(maj_prio))
+ return PTR_ERR(maj_prio);
+
+ for (chain = 0; chain < chains; chain++) {
+ err = create_fdb_sub_ns_prio_chain(steering, maj_prio);
+ if (err)
+ return err;
+ }
+
+ return 0;
+}
+
+static int create_fdb_fast_path(struct mlx5_flow_steering *steering)
+{
+ int err;
+
+ steering->fdb_sub_ns = kcalloc(FDB_NUM_CHAINS,
+ sizeof(*steering->fdb_sub_ns),
+ GFP_KERNEL);
+ if (!steering->fdb_sub_ns)
+ return -ENOMEM;
+
+ err = create_fdb_chains(steering, FDB_TC_OFFLOAD, FDB_TC_MAX_CHAIN + 1);
+ if (err)
+ return err;
+
+ err = create_fdb_chains(steering, FDB_FT_OFFLOAD, 1);
+ if (err)
+ return err;
+
return 0;
}
static int init_fdb_root_ns(struct mlx5_flow_steering *steering)
{
- struct fs_prio *prio;
+ struct fs_prio *maj_prio;
+ int err;
steering->fdb_root_ns = create_root_ns(steering, FS_FT_FDB);
if (!steering->fdb_root_ns)
return -ENOMEM;
- prio = fs_create_prio(&steering->fdb_root_ns->ns, 0, 2);
- if (IS_ERR(prio))
+ maj_prio = fs_create_prio(&steering->fdb_root_ns->ns, FDB_BYPASS_PATH,
+ 1);
+ if (IS_ERR(maj_prio)) {
+ err = PTR_ERR(maj_prio);
+ goto out_err;
+ }
+ err = create_fdb_fast_path(steering);
+ if (err)
goto out_err;
- prio = fs_create_prio(&steering->fdb_root_ns->ns, 1, 1);
- if (IS_ERR(prio))
+ maj_prio = fs_create_prio(&steering->fdb_root_ns->ns, FDB_SLOW_PATH, 1);
+ if (IS_ERR(maj_prio)) {
+ err = PTR_ERR(maj_prio);
goto out_err;
+ }
+
+ /* We put this priority last, knowing that nothing will get here
+ * unless explicitly forwarded to. This is possible because the
+ * slow path tables have catch all rules and nothing gets passed
+ * those tables.
+ */
+ maj_prio = fs_create_prio(&steering->fdb_root_ns->ns, FDB_PER_VPORT, 1);
+ if (IS_ERR(maj_prio)) {
+ err = PTR_ERR(maj_prio);
+ goto out_err;
+ }
set_prio_attrs(steering->fdb_root_ns);
return 0;
out_err:
cleanup_root_ns(steering->fdb_root_ns);
+ kfree(steering->fdb_sub_ns);
+ steering->fdb_sub_ns = NULL;
steering->fdb_root_ns = NULL;
- return PTR_ERR(prio);
+ return err;
}
static int init_egress_acl_root_ns(struct mlx5_flow_steering *steering, int vport)
@@ -2496,16 +2955,18 @@
static int init_egress_acls_root_ns(struct mlx5_core_dev *dev)
{
struct mlx5_flow_steering *steering = dev->priv.steering;
+ int total_vports = mlx5_eswitch_get_total_vports(dev);
int err;
int i;
- steering->esw_egress_root_ns = kcalloc(MLX5_TOTAL_VPORTS(dev),
- sizeof(*steering->esw_egress_root_ns),
- GFP_KERNEL);
+ steering->esw_egress_root_ns =
+ kcalloc(total_vports,
+ sizeof(*steering->esw_egress_root_ns),
+ GFP_KERNEL);
if (!steering->esw_egress_root_ns)
return -ENOMEM;
- for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++) {
+ for (i = 0; i < total_vports; i++) {
err = init_egress_acl_root_ns(steering, i);
if (err)
goto cleanup_root_ns;
@@ -2524,16 +2985,18 @@
static int init_ingress_acls_root_ns(struct mlx5_core_dev *dev)
{
struct mlx5_flow_steering *steering = dev->priv.steering;
+ int total_vports = mlx5_eswitch_get_total_vports(dev);
int err;
int i;
- steering->esw_ingress_root_ns = kcalloc(MLX5_TOTAL_VPORTS(dev),
- sizeof(*steering->esw_ingress_root_ns),
- GFP_KERNEL);
+ steering->esw_ingress_root_ns =
+ kcalloc(total_vports,
+ sizeof(*steering->esw_ingress_root_ns),
+ GFP_KERNEL);
if (!steering->esw_ingress_root_ns)
return -ENOMEM;
- for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++) {
+ for (i = 0; i < total_vports; i++) {
err = init_ingress_acl_root_ns(steering, i);
if (err)
goto cleanup_root_ns;
@@ -2551,16 +3014,23 @@
static int init_egress_root_ns(struct mlx5_flow_steering *steering)
{
- struct fs_prio *prio;
+ int err;
steering->egress_root_ns = create_root_ns(steering,
FS_FT_NIC_TX);
if (!steering->egress_root_ns)
return -ENOMEM;
- /* create 1 prio*/
- prio = fs_create_prio(&steering->egress_root_ns->ns, 0, 1);
- return PTR_ERR_OR_ZERO(prio);
+ err = init_root_tree(steering, &egress_root_fs,
+ &steering->egress_root_ns->ns.node);
+ if (err)
+ goto cleanup;
+ set_prio_attrs(steering->egress_root_ns);
+ return 0;
+cleanup:
+ cleanup_root_ns(steering->egress_root_ns);
+ steering->egress_root_ns = NULL;
+ return err;
}
int mlx5_init_fs(struct mlx5_core_dev *dev)
@@ -2628,7 +3098,21 @@
goto err;
}
- if (MLX5_IPSEC_DEV(dev)) {
+ if (MLX5_CAP_FLOWTABLE_RDMA_RX(dev, ft_support) &&
+ MLX5_CAP_FLOWTABLE_RDMA_RX(dev, table_miss_action_domain)) {
+ err = init_rdma_rx_root_ns(steering);
+ if (err)
+ goto err;
+ }
+
+ if (MLX5_CAP_FLOWTABLE_RDMA_TX(dev, ft_support)) {
+ err = init_rdma_tx_root_ns(steering);
+ if (err)
+ goto err;
+ }
+
+ if (mlx5_fpga_ipsec_device_caps(steering->dev) & MLX5_ACCEL_IPSEC_CAP_DEVICE ||
+ MLX5_CAP_FLOWTABLE_NIC_TX(dev, ft_support)) {
err = init_egress_root_ns(steering);
if (err)
goto err;
@@ -2657,7 +3141,7 @@
goto update_ft_fail;
}
- err = root->cmds->update_root_ft(dev, root->root_ft, underlay_qpn,
+ err = root->cmds->update_root_ft(root, root->root_ft, underlay_qpn,
false);
if (err) {
mlx5_core_warn(dev, "Failed adding underlay QPN (%u) to root FT err(%d)\n",
@@ -2701,7 +3185,7 @@
goto out;
}
- err = root->cmds->update_root_ft(dev, root->root_ft, underlay_qpn,
+ err = root->cmds->update_root_ft(root, root->root_ft, underlay_qpn,
true);
if (err)
mlx5_core_warn(dev, "Failed removing underlay QPN (%u) from root FT err(%d)\n",
@@ -2718,3 +3202,160 @@
return err;
}
EXPORT_SYMBOL(mlx5_fs_remove_rx_underlay_qpn);
+
+static struct mlx5_flow_root_namespace
+*get_root_namespace(struct mlx5_core_dev *dev, enum mlx5_flow_namespace_type ns_type)
+{
+ struct mlx5_flow_namespace *ns;
+
+ if (ns_type == MLX5_FLOW_NAMESPACE_ESW_EGRESS ||
+ ns_type == MLX5_FLOW_NAMESPACE_ESW_INGRESS)
+ ns = mlx5_get_flow_vport_acl_namespace(dev, ns_type, 0);
+ else
+ ns = mlx5_get_flow_namespace(dev, ns_type);
+ if (!ns)
+ return NULL;
+
+ return find_root(&ns->node);
+}
+
+struct mlx5_modify_hdr *mlx5_modify_header_alloc(struct mlx5_core_dev *dev,
+ u8 ns_type, u8 num_actions,
+ void *modify_actions)
+{
+ struct mlx5_flow_root_namespace *root;
+ struct mlx5_modify_hdr *modify_hdr;
+ int err;
+
+ root = get_root_namespace(dev, ns_type);
+ if (!root)
+ return ERR_PTR(-EOPNOTSUPP);
+
+ modify_hdr = kzalloc(sizeof(*modify_hdr), GFP_KERNEL);
+ if (!modify_hdr)
+ return ERR_PTR(-ENOMEM);
+
+ modify_hdr->ns_type = ns_type;
+ err = root->cmds->modify_header_alloc(root, ns_type, num_actions,
+ modify_actions, modify_hdr);
+ if (err) {
+ kfree(modify_hdr);
+ return ERR_PTR(err);
+ }
+
+ return modify_hdr;
+}
+EXPORT_SYMBOL(mlx5_modify_header_alloc);
+
+void mlx5_modify_header_dealloc(struct mlx5_core_dev *dev,
+ struct mlx5_modify_hdr *modify_hdr)
+{
+ struct mlx5_flow_root_namespace *root;
+
+ root = get_root_namespace(dev, modify_hdr->ns_type);
+ if (WARN_ON(!root))
+ return;
+ root->cmds->modify_header_dealloc(root, modify_hdr);
+ kfree(modify_hdr);
+}
+EXPORT_SYMBOL(mlx5_modify_header_dealloc);
+
+struct mlx5_pkt_reformat *mlx5_packet_reformat_alloc(struct mlx5_core_dev *dev,
+ int reformat_type,
+ size_t size,
+ void *reformat_data,
+ enum mlx5_flow_namespace_type ns_type)
+{
+ struct mlx5_pkt_reformat *pkt_reformat;
+ struct mlx5_flow_root_namespace *root;
+ int err;
+
+ root = get_root_namespace(dev, ns_type);
+ if (!root)
+ return ERR_PTR(-EOPNOTSUPP);
+
+ pkt_reformat = kzalloc(sizeof(*pkt_reformat), GFP_KERNEL);
+ if (!pkt_reformat)
+ return ERR_PTR(-ENOMEM);
+
+ pkt_reformat->ns_type = ns_type;
+ pkt_reformat->reformat_type = reformat_type;
+ err = root->cmds->packet_reformat_alloc(root, reformat_type, size,
+ reformat_data, ns_type,
+ pkt_reformat);
+ if (err) {
+ kfree(pkt_reformat);
+ return ERR_PTR(err);
+ }
+
+ return pkt_reformat;
+}
+EXPORT_SYMBOL(mlx5_packet_reformat_alloc);
+
+void mlx5_packet_reformat_dealloc(struct mlx5_core_dev *dev,
+ struct mlx5_pkt_reformat *pkt_reformat)
+{
+ struct mlx5_flow_root_namespace *root;
+
+ root = get_root_namespace(dev, pkt_reformat->ns_type);
+ if (WARN_ON(!root))
+ return;
+ root->cmds->packet_reformat_dealloc(root, pkt_reformat);
+ kfree(pkt_reformat);
+}
+EXPORT_SYMBOL(mlx5_packet_reformat_dealloc);
+
+int mlx5_flow_namespace_set_peer(struct mlx5_flow_root_namespace *ns,
+ struct mlx5_flow_root_namespace *peer_ns)
+{
+ if (peer_ns && ns->mode != peer_ns->mode) {
+ mlx5_core_err(ns->dev,
+ "Can't peer namespace of different steering mode\n");
+ return -EINVAL;
+ }
+
+ return ns->cmds->set_peer(ns, peer_ns);
+}
+
+/* This function should be called only at init stage of the namespace.
+ * It is not safe to call this function while steering operations
+ * are executed in the namespace.
+ */
+int mlx5_flow_namespace_set_mode(struct mlx5_flow_namespace *ns,
+ enum mlx5_flow_steering_mode mode)
+{
+ struct mlx5_flow_root_namespace *root;
+ const struct mlx5_flow_cmds *cmds;
+ int err;
+
+ root = find_root(&ns->node);
+ if (&root->ns != ns)
+ /* Can't set cmds to non root namespace */
+ return -EINVAL;
+
+ if (root->table_type != FS_FT_FDB)
+ return -EOPNOTSUPP;
+
+ if (root->mode == mode)
+ return 0;
+
+ if (mode == MLX5_FLOW_STEERING_MODE_SMFS)
+ cmds = mlx5_fs_cmd_get_dr_cmds();
+ else
+ cmds = mlx5_fs_cmd_get_fw_cmds();
+ if (!cmds)
+ return -EOPNOTSUPP;
+
+ err = cmds->create_ns(root);
+ if (err) {
+ mlx5_core_err(root->dev, "Failed to create flow namespace (%d)\n",
+ err);
+ return err;
+ }
+
+ root->cmds->destroy_ns(root);
+ root->cmds = cmds;
+ root->mode = mode;
+
+ return 0;
+}
--
Gitblit v1.6.2