svn commit: r359541 - in stable/11/sys: dev/mlx5 dev/mlx5/mlx5_core modules/mlx5
Konstantin Belousov
kib at FreeBSD.org
Wed Apr 1 22:44:37 UTC 2020
Author: kib
Date: Wed Apr 1 22:44:26 2020
New Revision: 359541
URL: https://svnweb.freebsd.org/changeset/base/359541
Log:
MFC r359100:
mlx5: Restore eswitch management code from attic.
Added:
stable/11/sys/dev/mlx5/mlx5_core/eswitch.h
- copied unchanged from r359100, head/sys/dev/mlx5/mlx5_core/eswitch.h
stable/11/sys/dev/mlx5/mlx5_core/flow_table.h
- copied unchanged from r359100, head/sys/dev/mlx5/mlx5_core/flow_table.h
stable/11/sys/dev/mlx5/mlx5_core/mlx5_eswitch.c
- copied unchanged from r359100, head/sys/dev/mlx5/mlx5_core/mlx5_eswitch.c
Modified:
stable/11/sys/dev/mlx5/driver.h
stable/11/sys/modules/mlx5/Makefile
Directory Properties:
stable/11/ (props changed)
Modified: stable/11/sys/dev/mlx5/driver.h
==============================================================================
--- stable/11/sys/dev/mlx5/driver.h Wed Apr 1 22:42:42 2020 (r359540)
+++ stable/11/sys/dev/mlx5/driver.h Wed Apr 1 22:44:26 2020 (r359541)
@@ -606,6 +606,8 @@ struct mlx5_priv {
spinlock_t ctx_lock;
unsigned long pci_dev_data;
struct mlx5_pme_stats pme_stats;
+
+ struct mlx5_eswitch *eswitch;
};
enum mlx5_device_state {
Copied: stable/11/sys/dev/mlx5/mlx5_core/eswitch.h (from r359100, head/sys/dev/mlx5/mlx5_core/eswitch.h)
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ stable/11/sys/dev/mlx5/mlx5_core/eswitch.h Wed Apr 1 22:44:26 2020 (r359541, copy of r359100, head/sys/dev/mlx5/mlx5_core/eswitch.h)
@@ -0,0 +1,169 @@
+/*-
+ * Copyright (c) 2013-2017, Mellanox Technologies, Ltd. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * $FreeBSD$
+ */
+
+#ifndef __MLX5_ESWITCH_H__
+#define __MLX5_ESWITCH_H__
+
+#include <linux/if_ether.h>
+#include <dev/mlx5/device.h>
+
+#define MLX5_MAX_UC_PER_VPORT(dev) \
+ (1 << MLX5_CAP_GEN(dev, log_max_current_uc_list))
+
+#define MLX5_MAX_MC_PER_VPORT(dev) \
+ (1 << MLX5_CAP_GEN(dev, log_max_current_mc_list))
+
+#define MLX5_L2_ADDR_HASH_SIZE (BIT(BITS_PER_BYTE))
+#define MLX5_L2_ADDR_HASH(addr) (addr[5])
+
+/* L2 -mac address based- hash helpers */
+struct l2addr_node {
+ struct hlist_node hlist;
+ u8 addr[ETH_ALEN];
+};
+
+#define for_each_l2hash_node(hn, tmp, hash, i) \
+ for (i = 0; i < MLX5_L2_ADDR_HASH_SIZE; i++) \
+ hlist_for_each_entry_safe(hn, tmp, &hash[i], hlist)
+
+#define l2addr_hash_find(hash, mac, type) ({ \
+ int ix = MLX5_L2_ADDR_HASH(mac); \
+ bool found = false; \
+ type *ptr = NULL; \
+ \
+ hlist_for_each_entry(ptr, &hash[ix], node.hlist) \
+ if (ether_addr_equal(ptr->node.addr, mac)) {\
+ found = true; \
+ break; \
+ } \
+ if (!found) \
+ ptr = NULL; \
+ ptr; \
+})
+
+#define l2addr_hash_add(hash, mac, type, gfp) ({ \
+ int ix = MLX5_L2_ADDR_HASH(mac); \
+ type *ptr = NULL; \
+ \
+ ptr = kzalloc(sizeof(type), gfp); \
+ if (ptr) { \
+ ether_addr_copy(ptr->node.addr, mac); \
+ hlist_add_head(&ptr->node.hlist, &hash[ix]);\
+ } \
+ ptr; \
+})
+
+#define l2addr_hash_del(ptr) ({ \
+ hlist_del(&ptr->node.hlist); \
+ kfree(ptr); \
+})
+
+struct vport_ingress {
+ struct mlx5_flow_table *acl;
+ struct mlx5_flow_group *drop_grp;
+ struct mlx5_flow_rule *drop_rule;
+};
+
+struct vport_egress {
+ struct mlx5_flow_table *acl;
+ struct mlx5_flow_group *allowed_vlans_grp;
+ struct mlx5_flow_group *drop_grp;
+ struct mlx5_flow_rule *allowed_vlan;
+ struct mlx5_flow_rule *drop_rule;
+};
+
+struct mlx5_vport {
+ struct mlx5_core_dev *dev;
+ int vport;
+ struct hlist_head uc_list[MLX5_L2_ADDR_HASH_SIZE];
+ struct hlist_head mc_list[MLX5_L2_ADDR_HASH_SIZE];
+ struct work_struct vport_change_handler;
+
+ struct vport_ingress ingress;
+ struct vport_egress egress;
+
+ u16 vlan;
+ u8 qos;
+ struct mutex state_lock; /* protect dynamic state changes */
+ /* This spinlock protects access to vport data, between
+ * "esw_vport_disable" and ongoing interrupt "mlx5_eswitch_vport_event"
+ * once vport marked as disabled new interrupts are discarded.
+ */
+ spinlock_t lock; /* vport events sync */
+ bool enabled;
+ u16 enabled_events;
+};
+
+struct mlx5_l2_table {
+ struct hlist_head l2_hash[MLX5_L2_ADDR_HASH_SIZE];
+ u32 size;
+ unsigned long *bitmap;
+};
+
+struct mlx5_eswitch_fdb {
+ void *fdb;
+ struct mlx5_flow_group *addr_grp;
+};
+
+struct mlx5_eswitch {
+ struct mlx5_core_dev *dev;
+ struct mlx5_l2_table l2_table;
+ struct mlx5_eswitch_fdb fdb_table;
+ struct hlist_head mc_table[MLX5_L2_ADDR_HASH_SIZE];
+ struct workqueue_struct *work_queue;
+ struct mlx5_vport *vports;
+ int total_vports;
+ int enabled_vports;
+};
+
+struct mlx5_esw_vport_info {
+ __u32 vf;
+ __u8 mac[32];
+ __u32 vlan;
+ __u32 qos;
+ __u32 spoofchk;
+ __u32 linkstate;
+ __u32 min_tx_rate;
+ __u32 max_tx_rate;
+};
+
+/* E-Switch API */
+int mlx5_eswitch_init(struct mlx5_core_dev *dev);
+void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw);
+void mlx5_eswitch_vport_event(struct mlx5_eswitch *esw, struct mlx5_eqe *eqe);
+int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs);
+void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw);
+int mlx5_eswitch_set_vport_mac(struct mlx5_eswitch *esw,
+ int vport, u8 mac[ETH_ALEN]);
+int mlx5_eswitch_set_vport_state(struct mlx5_eswitch *esw,
+ int vport, int link_state);
+int mlx5_eswitch_set_vport_vlan(struct mlx5_eswitch *esw,
+ int vport, u16 vlan, u8 qos);
+int mlx5_eswitch_get_vport_config(struct mlx5_eswitch *esw,
+ int vport, struct mlx5_esw_vport_info *evi);
+
+#endif /* __MLX5_ESWITCH_H__ */
Copied: stable/11/sys/dev/mlx5/mlx5_core/flow_table.h (from r359100, head/sys/dev/mlx5/mlx5_core/flow_table.h)
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ stable/11/sys/dev/mlx5/mlx5_core/flow_table.h Wed Apr 1 22:44:26 2020 (r359541, copy of r359100, head/sys/dev/mlx5/mlx5_core/flow_table.h)
@@ -0,0 +1,65 @@
+/*-
+ * Copyright (c) 2013-2017, Mellanox Technologies, Ltd. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * $FreeBSD$
+ */
+
+#ifndef MLX5_FLOW_TABLE_H
+#define MLX5_FLOW_TABLE_H
+
+#include <dev/mlx5/driver.h>
+
+#define MLX5_SET_FLOW_TABLE_ROOT_OPMOD_SET 0x0
+#define MLX5_SET_FLOW_TABLE_ROOT_OPMOD_RESET 0x1
+
+struct mlx5_flow_table_group {
+ u8 log_sz;
+ u8 match_criteria_enable;
+ u32 match_criteria[MLX5_ST_SZ_DW(fte_match_param)];
+};
+
+void *mlx5_create_flow_table(struct mlx5_core_dev *dev, u8 level, u8 table_type,
+ u16 vport,
+ u16 num_groups,
+ struct mlx5_flow_table_group *group);
+void mlx5_destroy_flow_table(void *flow_table);
+int mlx5_add_flow_table_entry(void *flow_table, u8 match_criteria_enable,
+ void *match_criteria, void *flow_context,
+ u32 *flow_index);
+int mlx5_del_flow_table_entry(void *flow_table, u32 flow_index);
+u32 mlx5_get_flow_table_id(void *flow_table);
+int mlx5_set_flow_table_root(struct mlx5_core_dev *mdev, u16 op_mod,
+ u8 vport_num, u8 table_type, u32 table_id,
+ u32 underlay_qpn);
+void *mlx5_get_flow_table_properties(void *flow_table);
+u32 mlx5_set_flow_table_miss_id(void *flow_table, u32 miss_ft_id);
+
+int mlx5_create_flow_counter(struct mlx5_core_dev *dev, u16 *cnt_id);
+void mlx5_destroy_flow_counter(struct mlx5_core_dev *dev, u16 cnt_id);
+int mlx5_query_flow_counters(struct mlx5_core_dev *dev,
+ u32 num_counters, u16 *cnt_ids,
+ struct mlx5_traffic_counter *cnt_data);
+int mlx5_reset_flow_counter(struct mlx5_core_dev *dev, u16 cnt_id);
+
+#endif /* MLX5_FLOW_TABLE_H */
Copied: stable/11/sys/dev/mlx5/mlx5_core/mlx5_eswitch.c (from r359100, head/sys/dev/mlx5/mlx5_core/mlx5_eswitch.c)
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ stable/11/sys/dev/mlx5/mlx5_core/mlx5_eswitch.c Wed Apr 1 22:44:26 2020 (r359541, copy of r359100, head/sys/dev/mlx5/mlx5_core/mlx5_eswitch.c)
@@ -0,0 +1,1382 @@
+/*-
+ * Copyright (c) 2013-2017, Mellanox Technologies, Ltd. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * $FreeBSD$
+ */
+
+#include <linux/etherdevice.h>
+#include <dev/mlx5/driver.h>
+#include <dev/mlx5/mlx5_ifc.h>
+#include <dev/mlx5/vport.h>
+#include <dev/mlx5/fs.h>
+#include "mlx5_core.h"
+#include "eswitch.h"
+
+#define UPLINK_VPORT 0xFFFF
+
+#define MLX5_DEBUG_ESWITCH_MASK BIT(3)
+
+#define esw_info(dev, format, ...) \
+ printf("mlx5_core: INFO: ""(%s): E-Switch: " format, (dev)->priv.name, ##__VA_ARGS__)
+
+#define esw_warn(dev, format, ...) \
+ printf("mlx5_core: WARN: ""(%s): E-Switch: " format, (dev)->priv.name, ##__VA_ARGS__)
+
+#define esw_debug(dev, format, ...) \
+ mlx5_core_dbg_mask(dev, MLX5_DEBUG_ESWITCH_MASK, format, ##__VA_ARGS__)
+
+enum {
+ MLX5_ACTION_NONE = 0,
+ MLX5_ACTION_ADD = 1,
+ MLX5_ACTION_DEL = 2,
+};
+
+/* E-Switch UC L2 table hash node */
+struct esw_uc_addr {
+ struct l2addr_node node;
+ u32 table_index;
+ u32 vport;
+};
+
+/* E-Switch MC FDB table hash node */
+struct esw_mc_addr { /* SRIOV only */
+ struct l2addr_node node;
+ struct mlx5_flow_rule *uplink_rule; /* Forward to uplink rule */
+ u32 refcnt;
+};
+
+/* Vport UC/MC hash node */
+struct vport_addr {
+ struct l2addr_node node;
+ u8 action;
+ u32 vport;
+ struct mlx5_flow_rule *flow_rule; /* SRIOV only */
+};
+
+enum {
+ UC_ADDR_CHANGE = BIT(0),
+ MC_ADDR_CHANGE = BIT(1),
+};
+
+/* Vport context events */
+#define SRIOV_VPORT_EVENTS (UC_ADDR_CHANGE | \
+ MC_ADDR_CHANGE)
+
+static int arm_vport_context_events_cmd(struct mlx5_core_dev *dev, u16 vport,
+ u32 events_mask)
+{
+ int in[MLX5_ST_SZ_DW(modify_nic_vport_context_in)] = {0};
+ int out[MLX5_ST_SZ_DW(modify_nic_vport_context_out)] = {0};
+ void *nic_vport_ctx;
+
+ MLX5_SET(modify_nic_vport_context_in, in,
+ opcode, MLX5_CMD_OP_MODIFY_NIC_VPORT_CONTEXT);
+ MLX5_SET(modify_nic_vport_context_in, in, field_select.change_event, 1);
+ MLX5_SET(modify_nic_vport_context_in, in, vport_number, vport);
+ if (vport)
+ MLX5_SET(modify_nic_vport_context_in, in, other_vport, 1);
+ nic_vport_ctx = MLX5_ADDR_OF(modify_nic_vport_context_in,
+ in, nic_vport_context);
+
+ MLX5_SET(nic_vport_context, nic_vport_ctx, arm_change_event, 1);
+
+ if (events_mask & UC_ADDR_CHANGE)
+ MLX5_SET(nic_vport_context, nic_vport_ctx,
+ event_on_uc_address_change, 1);
+ if (events_mask & MC_ADDR_CHANGE)
+ MLX5_SET(nic_vport_context, nic_vport_ctx,
+ event_on_mc_address_change, 1);
+
+ return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
+}
+
+/* E-Switch vport context HW commands */
+static int query_esw_vport_context_cmd(struct mlx5_core_dev *mdev, u32 vport,
+ u32 *out, int outlen)
+{
+ u32 in[MLX5_ST_SZ_DW(query_esw_vport_context_in)] = {0};
+
+ MLX5_SET(query_nic_vport_context_in, in, opcode,
+ MLX5_CMD_OP_QUERY_ESW_VPORT_CONTEXT);
+
+ MLX5_SET(query_esw_vport_context_in, in, vport_number, vport);
+ if (vport)
+ MLX5_SET(query_esw_vport_context_in, in, other_vport, 1);
+
+ return mlx5_cmd_exec(mdev, in, sizeof(in), out, outlen);
+}
+
+static int query_esw_vport_cvlan(struct mlx5_core_dev *dev, u32 vport,
+ u16 *vlan, u8 *qos)
+{
+ u32 out[MLX5_ST_SZ_DW(query_esw_vport_context_out)] = {0};
+ int err;
+ bool cvlan_strip;
+ bool cvlan_insert;
+
+ *vlan = 0;
+ *qos = 0;
+
+ if (!MLX5_CAP_ESW(dev, vport_cvlan_strip) ||
+ !MLX5_CAP_ESW(dev, vport_cvlan_insert_if_not_exist))
+ return -ENOTSUPP;
+
+ err = query_esw_vport_context_cmd(dev, vport, out, sizeof(out));
+ if (err)
+ goto out;
+
+ cvlan_strip = MLX5_GET(query_esw_vport_context_out, out,
+ esw_vport_context.vport_cvlan_strip);
+
+ cvlan_insert = MLX5_GET(query_esw_vport_context_out, out,
+ esw_vport_context.vport_cvlan_insert);
+
+ if (cvlan_strip || cvlan_insert) {
+ *vlan = MLX5_GET(query_esw_vport_context_out, out,
+ esw_vport_context.cvlan_id);
+ *qos = MLX5_GET(query_esw_vport_context_out, out,
+ esw_vport_context.cvlan_pcp);
+ }
+
+ esw_debug(dev, "Query Vport[%d] cvlan: VLAN %d qos=%d\n",
+ vport, *vlan, *qos);
+out:
+ return err;
+}
+
+static int modify_esw_vport_context_cmd(struct mlx5_core_dev *dev, u16 vport,
+ void *in, int inlen)
+{
+ u32 out[MLX5_ST_SZ_DW(modify_esw_vport_context_out)] = {0};
+
+ MLX5_SET(modify_esw_vport_context_in, in, vport_number, vport);
+ if (vport)
+ MLX5_SET(modify_esw_vport_context_in, in, other_vport, 1);
+
+ MLX5_SET(modify_esw_vport_context_in, in, opcode,
+ MLX5_CMD_OP_MODIFY_ESW_VPORT_CONTEXT);
+
+ return mlx5_cmd_exec(dev, in, inlen, out, sizeof(out));
+}
+
+static int modify_esw_vport_cvlan(struct mlx5_core_dev *dev, u32 vport,
+ u16 vlan, u8 qos, bool set)
+{
+ u32 in[MLX5_ST_SZ_DW(modify_esw_vport_context_in)] = {0};
+
+ if (!MLX5_CAP_ESW(dev, vport_cvlan_strip) ||
+ !MLX5_CAP_ESW(dev, vport_cvlan_insert_if_not_exist))
+ return -ENOTSUPP;
+
+ esw_debug(dev, "Set Vport[%d] VLAN %d qos %d set=%d\n",
+ vport, vlan, qos, set);
+
+ if (set) {
+ MLX5_SET(modify_esw_vport_context_in, in,
+ esw_vport_context.vport_cvlan_strip, 1);
+ /* insert only if no vlan in packet */
+ MLX5_SET(modify_esw_vport_context_in, in,
+ esw_vport_context.vport_cvlan_insert, 1);
+ MLX5_SET(modify_esw_vport_context_in, in,
+ esw_vport_context.cvlan_pcp, qos);
+ MLX5_SET(modify_esw_vport_context_in, in,
+ esw_vport_context.cvlan_id, vlan);
+ }
+
+ MLX5_SET(modify_esw_vport_context_in, in,
+ field_select.vport_cvlan_strip, 1);
+ MLX5_SET(modify_esw_vport_context_in, in,
+ field_select.vport_cvlan_insert, 1);
+
+ return modify_esw_vport_context_cmd(dev, vport, in, sizeof(in));
+}
+
+/* HW L2 Table (MPFS) management */
+static int set_l2_table_entry_cmd(struct mlx5_core_dev *dev, u32 index,
+ u8 *mac, u8 vlan_valid, u16 vlan)
+{
+ u32 in[MLX5_ST_SZ_DW(set_l2_table_entry_in)] = {0};
+ u32 out[MLX5_ST_SZ_DW(set_l2_table_entry_out)] = {0};
+ u8 *in_mac_addr;
+
+ MLX5_SET(set_l2_table_entry_in, in, opcode,
+ MLX5_CMD_OP_SET_L2_TABLE_ENTRY);
+ MLX5_SET(set_l2_table_entry_in, in, table_index, index);
+ MLX5_SET(set_l2_table_entry_in, in, vlan_valid, vlan_valid);
+ MLX5_SET(set_l2_table_entry_in, in, vlan, vlan);
+
+ in_mac_addr = MLX5_ADDR_OF(set_l2_table_entry_in, in, mac_address);
+ ether_addr_copy(&in_mac_addr[2], mac);
+
+ return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
+}
+
+static int del_l2_table_entry_cmd(struct mlx5_core_dev *dev, u32 index)
+{
+ u32 in[MLX5_ST_SZ_DW(delete_l2_table_entry_in)] = {0};
+ u32 out[MLX5_ST_SZ_DW(delete_l2_table_entry_out)] = {0};
+
+ MLX5_SET(delete_l2_table_entry_in, in, opcode,
+ MLX5_CMD_OP_DELETE_L2_TABLE_ENTRY);
+ MLX5_SET(delete_l2_table_entry_in, in, table_index, index);
+ return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
+}
+
+static int alloc_l2_table_index(struct mlx5_l2_table *l2_table, u32 *ix)
+{
+ int err = 0;
+
+ *ix = find_first_zero_bit(l2_table->bitmap, l2_table->size);
+ if (*ix >= l2_table->size)
+ err = -ENOSPC;
+ else
+ __set_bit(*ix, l2_table->bitmap);
+
+ return err;
+}
+
+static void free_l2_table_index(struct mlx5_l2_table *l2_table, u32 ix)
+{
+ __clear_bit(ix, l2_table->bitmap);
+}
+
+static int set_l2_table_entry(struct mlx5_core_dev *dev, u8 *mac,
+ u8 vlan_valid, u16 vlan,
+ u32 *index)
+{
+ struct mlx5_l2_table *l2_table = &dev->priv.eswitch->l2_table;
+ int err;
+
+ err = alloc_l2_table_index(l2_table, index);
+ if (err)
+ return err;
+
+ err = set_l2_table_entry_cmd(dev, *index, mac, vlan_valid, vlan);
+ if (err)
+ free_l2_table_index(l2_table, *index);
+
+ return err;
+}
+
+static void del_l2_table_entry(struct mlx5_core_dev *dev, u32 index)
+{
+ struct mlx5_l2_table *l2_table = &dev->priv.eswitch->l2_table;
+
+ del_l2_table_entry_cmd(dev, index);
+ free_l2_table_index(l2_table, index);
+}
+
+/* E-Switch FDB */
+static struct mlx5_flow_rule *
+esw_fdb_set_vport_rule(struct mlx5_eswitch *esw, u8 mac[ETH_ALEN], u32 vport)
+{
+ int match_header = MLX5_MATCH_OUTER_HEADERS;
+ struct mlx5_flow_destination dest;
+ struct mlx5_flow_rule *flow_rule = NULL;
+ u32 *match_v;
+ u32 *match_c;
+ u8 *dmac_v;
+ u8 *dmac_c;
+
+ match_v = kzalloc(MLX5_ST_SZ_BYTES(fte_match_param), GFP_KERNEL);
+ match_c = kzalloc(MLX5_ST_SZ_BYTES(fte_match_param), GFP_KERNEL);
+ if (!match_v || !match_c) {
+ printf("mlx5_core: WARN: ""FDB: Failed to alloc match parameters\n");
+ goto out;
+ }
+ dmac_v = MLX5_ADDR_OF(fte_match_param, match_v,
+ outer_headers.dmac_47_16);
+ dmac_c = MLX5_ADDR_OF(fte_match_param, match_c,
+ outer_headers.dmac_47_16);
+
+ ether_addr_copy(dmac_v, mac);
+ /* Match criteria mask */
+ memset(dmac_c, 0xff, 6);
+
+ dest.type = MLX5_FLOW_CONTEXT_DEST_TYPE_VPORT;
+ dest.vport_num = vport;
+
+ esw_debug(esw->dev,
+ "\tFDB add rule dmac_v(%pM) dmac_c(%pM) -> vport(%d)\n",
+ dmac_v, dmac_c, vport);
+ flow_rule =
+ mlx5_add_flow_rule(esw->fdb_table.fdb,
+ match_header,
+ match_c,
+ match_v,
+ MLX5_FLOW_CONTEXT_ACTION_FWD_DEST,
+ 0, &dest);
+ if (IS_ERR_OR_NULL(flow_rule)) {
+ printf("mlx5_core: WARN: ""FDB: Failed to add flow rule: dmac_v(%pM) dmac_c(%pM) -> vport(%d), err(%ld)\n", dmac_v, dmac_c, vport, PTR_ERR(flow_rule));
+ flow_rule = NULL;
+ }
+out:
+ kfree(match_v);
+ kfree(match_c);
+ return flow_rule;
+}
+
+static int esw_create_fdb_table(struct mlx5_eswitch *esw)
+{
+ int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
+ struct mlx5_core_dev *dev = esw->dev;
+ struct mlx5_flow_namespace *root_ns;
+ struct mlx5_flow_table *fdb;
+ struct mlx5_flow_group *g;
+ void *match_criteria;
+ int table_size;
+ u32 *flow_group_in;
+ u8 *dmac;
+ int err = 0;
+
+ esw_debug(dev, "Create FDB log_max_size(%d)\n",
+ MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size));
+
+ root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB);
+ if (!root_ns) {
+ esw_warn(dev, "Failed to get FDB flow namespace\n");
+ return -ENOMEM;
+ }
+
+ flow_group_in = mlx5_vzalloc(inlen);
+ if (!flow_group_in)
+ return -ENOMEM;
+ memset(flow_group_in, 0, inlen);
+
+ /* (-2) Since MaorG said so .. */
+ table_size = BIT(MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size)) - 2;
+
+ fdb = mlx5_create_flow_table(root_ns, 0, "FDB", table_size);
+ if (IS_ERR_OR_NULL(fdb)) {
+ err = PTR_ERR(fdb);
+ esw_warn(dev, "Failed to create FDB Table err %d\n", err);
+ goto out;
+ }
+
+ MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable,
+ MLX5_MATCH_OUTER_HEADERS);
+ match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria);
+ dmac = MLX5_ADDR_OF(fte_match_param, match_criteria, outer_headers.dmac_47_16);
+ MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0);
+ MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, table_size - 1);
+ eth_broadcast_addr(dmac);
+
+ g = mlx5_create_flow_group(fdb, flow_group_in);
+ if (IS_ERR_OR_NULL(g)) {
+ err = PTR_ERR(g);
+ esw_warn(dev, "Failed to create flow group err(%d)\n", err);
+ goto out;
+ }
+
+ esw->fdb_table.addr_grp = g;
+ esw->fdb_table.fdb = fdb;
+out:
+ kfree(flow_group_in);
+ if (err && !IS_ERR_OR_NULL(fdb))
+ mlx5_destroy_flow_table(fdb);
+ return err;
+}
+
+static void esw_destroy_fdb_table(struct mlx5_eswitch *esw)
+{
+ if (!esw->fdb_table.fdb)
+ return;
+
+ esw_debug(esw->dev, "Destroy FDB Table\n");
+ mlx5_destroy_flow_group(esw->fdb_table.addr_grp);
+ mlx5_destroy_flow_table(esw->fdb_table.fdb);
+ esw->fdb_table.fdb = NULL;
+ esw->fdb_table.addr_grp = NULL;
+}
+
+/* E-Switch vport UC/MC lists management */
+typedef int (*vport_addr_action)(struct mlx5_eswitch *esw,
+ struct vport_addr *vaddr);
+
+static int esw_add_uc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr)
+{
+ struct hlist_head *hash = esw->l2_table.l2_hash;
+ struct esw_uc_addr *esw_uc;
+ u8 *mac = vaddr->node.addr;
+ u32 vport = vaddr->vport;
+ int err;
+
+ esw_uc = l2addr_hash_find(hash, mac, struct esw_uc_addr);
+ if (esw_uc) {
+ esw_warn(esw->dev,
+ "Failed to set L2 mac(%pM) for vport(%d), mac is already in use by vport(%d)\n",
+ mac, vport, esw_uc->vport);
+ return -EEXIST;
+ }
+
+ esw_uc = l2addr_hash_add(hash, mac, struct esw_uc_addr, GFP_KERNEL);
+ if (!esw_uc)
+ return -ENOMEM;
+ esw_uc->vport = vport;
+
+ err = set_l2_table_entry(esw->dev, mac, 0, 0, &esw_uc->table_index);
+ if (err)
+ goto abort;
+
+ if (esw->fdb_table.fdb) /* SRIOV is enabled: Forward UC MAC to vport */
+ vaddr->flow_rule = esw_fdb_set_vport_rule(esw, mac, vport);
+
+ esw_debug(esw->dev, "\tADDED UC MAC: vport[%d] %pM index:%d fr(%p)\n",
+ vport, mac, esw_uc->table_index, vaddr->flow_rule);
+ return err;
+abort:
+ l2addr_hash_del(esw_uc);
+ return err;
+}
+
+static int esw_del_uc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr)
+{
+ struct hlist_head *hash = esw->l2_table.l2_hash;
+ struct esw_uc_addr *esw_uc;
+ u8 *mac = vaddr->node.addr;
+ u32 vport = vaddr->vport;
+
+ esw_uc = l2addr_hash_find(hash, mac, struct esw_uc_addr);
+ if (!esw_uc || esw_uc->vport != vport) {
+ esw_debug(esw->dev,
+ "MAC(%pM) doesn't belong to vport (%d)\n",
+ mac, vport);
+ return -EINVAL;
+ }
+ esw_debug(esw->dev, "\tDELETE UC MAC: vport[%d] %pM index:%d fr(%p)\n",
+ vport, mac, esw_uc->table_index, vaddr->flow_rule);
+
+ del_l2_table_entry(esw->dev, esw_uc->table_index);
+
+ if (vaddr->flow_rule)
+ mlx5_del_flow_rule(vaddr->flow_rule);
+ vaddr->flow_rule = NULL;
+
+ l2addr_hash_del(esw_uc);
+ return 0;
+}
+
+static int esw_add_mc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr)
+{
+ struct hlist_head *hash = esw->mc_table;
+ struct esw_mc_addr *esw_mc;
+ u8 *mac = vaddr->node.addr;
+ u32 vport = vaddr->vport;
+
+ if (!esw->fdb_table.fdb)
+ return 0;
+
+ esw_mc = l2addr_hash_find(hash, mac, struct esw_mc_addr);
+ if (esw_mc)
+ goto add;
+
+ esw_mc = l2addr_hash_add(hash, mac, struct esw_mc_addr, GFP_KERNEL);
+ if (!esw_mc)
+ return -ENOMEM;
+
+ esw_mc->uplink_rule = /* Forward MC MAC to Uplink */
+ esw_fdb_set_vport_rule(esw, mac, UPLINK_VPORT);
+add:
+ esw_mc->refcnt++;
+ /* Forward MC MAC to vport */
+ vaddr->flow_rule = esw_fdb_set_vport_rule(esw, mac, vport);
+ esw_debug(esw->dev,
+ "\tADDED MC MAC: vport[%d] %pM fr(%p) refcnt(%d) uplinkfr(%p)\n",
+ vport, mac, vaddr->flow_rule,
+ esw_mc->refcnt, esw_mc->uplink_rule);
+ return 0;
+}
+
+static int esw_del_mc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr)
+{
+ struct hlist_head *hash = esw->mc_table;
+ struct esw_mc_addr *esw_mc;
+ u8 *mac = vaddr->node.addr;
+ u32 vport = vaddr->vport;
+
+ if (!esw->fdb_table.fdb)
+ return 0;
+
+ esw_mc = l2addr_hash_find(hash, mac, struct esw_mc_addr);
+ if (!esw_mc) {
+ esw_warn(esw->dev,
+ "Failed to find eswitch MC addr for MAC(%pM) vport(%d)",
+ mac, vport);
+ return -EINVAL;
+ }
+ esw_debug(esw->dev,
+ "\tDELETE MC MAC: vport[%d] %pM fr(%p) refcnt(%d) uplinkfr(%p)\n",
+ vport, mac, vaddr->flow_rule, esw_mc->refcnt,
+ esw_mc->uplink_rule);
+
+ if (vaddr->flow_rule)
+ mlx5_del_flow_rule(vaddr->flow_rule);
+ vaddr->flow_rule = NULL;
+
+ if (--esw_mc->refcnt)
+ return 0;
+
+ if (esw_mc->uplink_rule)
+ mlx5_del_flow_rule(esw_mc->uplink_rule);
+
+ l2addr_hash_del(esw_mc);
+ return 0;
+}
+
+/* Apply vport UC/MC list to HW l2 table and FDB table */
+static void esw_apply_vport_addr_list(struct mlx5_eswitch *esw,
+ u32 vport_num, int list_type)
+{
+ struct mlx5_vport *vport = &esw->vports[vport_num];
+ bool is_uc = list_type == MLX5_NIC_VPORT_LIST_TYPE_UC;
+ vport_addr_action vport_addr_add;
+ vport_addr_action vport_addr_del;
+ struct vport_addr *addr;
+ struct l2addr_node *node;
+ struct hlist_head *hash;
+ struct hlist_node *tmp;
+ int hi;
+
+ vport_addr_add = is_uc ? esw_add_uc_addr :
+ esw_add_mc_addr;
+ vport_addr_del = is_uc ? esw_del_uc_addr :
+ esw_del_mc_addr;
+
+ hash = is_uc ? vport->uc_list : vport->mc_list;
+ for_each_l2hash_node(node, tmp, hash, hi) {
+ addr = container_of(node, struct vport_addr, node);
+ switch (addr->action) {
+ case MLX5_ACTION_ADD:
+ vport_addr_add(esw, addr);
+ addr->action = MLX5_ACTION_NONE;
+ break;
+ case MLX5_ACTION_DEL:
+ vport_addr_del(esw, addr);
+ l2addr_hash_del(addr);
+ break;
+ }
+ }
+}
+
+/* Sync vport UC/MC list from vport context */
+static void esw_update_vport_addr_list(struct mlx5_eswitch *esw,
+ u32 vport_num, int list_type)
+{
+ struct mlx5_vport *vport = &esw->vports[vport_num];
+ bool is_uc = list_type == MLX5_NIC_VPORT_LIST_TYPE_UC;
+ u8 (*mac_list)[ETH_ALEN];
+ struct l2addr_node *node;
+ struct vport_addr *addr;
+ struct hlist_head *hash;
+ struct hlist_node *tmp;
+ int size;
+ int err;
+ int hi;
+ int i;
+
+ size = is_uc ? MLX5_MAX_UC_PER_VPORT(esw->dev) :
+ MLX5_MAX_MC_PER_VPORT(esw->dev);
+
+ mac_list = kcalloc(size, ETH_ALEN, GFP_KERNEL);
+ if (!mac_list)
+ return;
+
+ hash = is_uc ? vport->uc_list : vport->mc_list;
+
+ for_each_l2hash_node(node, tmp, hash, hi) {
+ addr = container_of(node, struct vport_addr, node);
+ addr->action = MLX5_ACTION_DEL;
+ }
+
+ err = mlx5_query_nic_vport_mac_list(esw->dev, vport_num, list_type,
+ mac_list, &size);
+ if (err)
+ return;
+ esw_debug(esw->dev, "vport[%d] context update %s list size (%d)\n",
+ vport_num, is_uc ? "UC" : "MC", size);
+
+ for (i = 0; i < size; i++) {
+ if (is_uc && !is_valid_ether_addr(mac_list[i]))
+ continue;
+
+ if (!is_uc && !is_multicast_ether_addr(mac_list[i]))
+ continue;
+
+ addr = l2addr_hash_find(hash, mac_list[i], struct vport_addr);
+ if (addr) {
+ addr->action = MLX5_ACTION_NONE;
+ continue;
+ }
+
+ addr = l2addr_hash_add(hash, mac_list[i], struct vport_addr,
+ GFP_KERNEL);
+ if (!addr) {
+ esw_warn(esw->dev,
+ "Failed to add MAC(%pM) to vport[%d] DB\n",
+ mac_list[i], vport_num);
+ continue;
+ }
+ addr->vport = vport_num;
+ addr->action = MLX5_ACTION_ADD;
+ }
+ kfree(mac_list);
+}
+
+static void esw_vport_change_handler(struct work_struct *work)
+{
+ struct mlx5_vport *vport =
+ container_of(work, struct mlx5_vport, vport_change_handler);
+ struct mlx5_core_dev *dev = vport->dev;
+ struct mlx5_eswitch *esw = dev->priv.eswitch;
+ u8 mac[ETH_ALEN];
+
+ mlx5_query_nic_vport_mac_address(dev, vport->vport, mac);
+ esw_debug(dev, "vport[%d] Context Changed: perm mac: %pM\n",
+ vport->vport, mac);
+
+ if (vport->enabled_events & UC_ADDR_CHANGE) {
+ esw_update_vport_addr_list(esw, vport->vport,
+ MLX5_NIC_VPORT_LIST_TYPE_UC);
+ esw_apply_vport_addr_list(esw, vport->vport,
+ MLX5_NIC_VPORT_LIST_TYPE_UC);
+ }
+
+ if (vport->enabled_events & MC_ADDR_CHANGE) {
+ esw_update_vport_addr_list(esw, vport->vport,
+ MLX5_NIC_VPORT_LIST_TYPE_MC);
+ esw_apply_vport_addr_list(esw, vport->vport,
+ MLX5_NIC_VPORT_LIST_TYPE_MC);
+ }
+
+ esw_debug(esw->dev, "vport[%d] Context Changed: Done\n", vport->vport);
+ if (vport->enabled)
+ arm_vport_context_events_cmd(dev, vport->vport,
+ vport->enabled_events);
+}
+
+static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw,
+ struct mlx5_vport *vport)
+{
+ int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
+ struct mlx5_flow_group *vlan_grp = NULL;
+ struct mlx5_flow_group *drop_grp = NULL;
+ struct mlx5_core_dev *dev = esw->dev;
+ struct mlx5_flow_namespace *root_ns;
+ struct mlx5_flow_table *acl;
+ void *match_criteria;
+ char table_name[32];
+ u32 *flow_group_in;
+ int table_size = 2;
+ int err = 0;
+
+ if (!MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support))
+ return;
+
+ esw_debug(dev, "Create vport[%d] egress ACL log_max_size(%d)\n",
+ vport->vport, MLX5_CAP_ESW_EGRESS_ACL(dev, log_max_ft_size));
+
+ root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_EGRESS);
+ if (!root_ns) {
+ esw_warn(dev, "Failed to get E-Switch egress flow namespace\n");
+ return;
+ }
+
+ flow_group_in = mlx5_vzalloc(inlen);
+ if (!flow_group_in)
+ return;
+
+ snprintf(table_name, 32, "egress_%d", vport->vport);
+ acl = mlx5_create_vport_flow_table(root_ns, vport->vport, 0, table_name, table_size);
+ if (IS_ERR_OR_NULL(acl)) {
+ err = PTR_ERR(acl);
+ esw_warn(dev, "Failed to create E-Switch vport[%d] egress flow Table, err(%d)\n",
+ vport->vport, err);
+ goto out;
+ }
+
+ MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS);
+ match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria);
+ MLX5_SET_TO_ONES(fte_match_param, match_criteria, outer_headers.cvlan_tag);
+ MLX5_SET_TO_ONES(fte_match_param, match_criteria, outer_headers.first_vid);
+ MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0);
+ MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 0);
+
+ vlan_grp = mlx5_create_flow_group(acl, flow_group_in);
+ if (IS_ERR_OR_NULL(vlan_grp)) {
+ err = PTR_ERR(vlan_grp);
+ esw_warn(dev, "Failed to create E-Switch vport[%d] egress allowed vlans flow group, err(%d)\n",
+ vport->vport, err);
+ goto out;
+ }
+
+ memset(flow_group_in, 0, inlen);
+ MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 1);
+ MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 1);
+ drop_grp = mlx5_create_flow_group(acl, flow_group_in);
+ if (IS_ERR_OR_NULL(drop_grp)) {
+ err = PTR_ERR(drop_grp);
+ esw_warn(dev, "Failed to create E-Switch vport[%d] egress drop flow group, err(%d)\n",
+ vport->vport, err);
+ goto out;
+ }
+
+ vport->egress.acl = acl;
+ vport->egress.drop_grp = drop_grp;
+ vport->egress.allowed_vlans_grp = vlan_grp;
*** DIFF OUTPUT TRUNCATED AT 1000 LINES ***
More information about the svn-src-all
mailing list