Commit 684a95c0 authored by David S. Miller's avatar David S. Miller

Merge branch 'mlxsw-ipv4-unicast-routing'

Jiri Pirko says:

====================
mlxsw: Implement IPV4 unicast routing

This patchset enables IPv4 unicast routing in the Mellanox Spectrum ASIC
switch driver. This builds upon the work that was done by a couple of
previous patchsets.

Patches 1,2,6 add a couple of dependencies outside the driver. Namely, the
ability to propagate ndo_neigh_construct()/destroy() through stacked devices and
a notification whenever DELAY_PROBE_TIME changes. When propagated down, the
ndos allow drivers to add and remove neighbour entries from their private
neighbour table. The DELAY_PROBE_TIME notification gives drivers the ability to
correctly configure their polling interval for neighbour activity, so that
active neighbour won't be marked as STALE.

Patches 3-5,7-8 add the neighbour offloading infrastructure, where patch 7 uses
the DELAY_PROBE_TIME notification in order to correctly configure the device's
polling interval. Patch 8 finally programs neighbours to the device's table
based on NEIGH_UPDATE notifications, so that directly connected routes can
be used.

Patches 9-16 build upon the previous patches and extend the router with
remote routes (nexthop) support.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parents 9046a745 0b2361d9
......@@ -4137,6 +4137,8 @@ static const struct net_device_ops bond_netdev_ops = {
.ndo_add_slave = bond_enslave,
.ndo_del_slave = bond_release,
.ndo_fix_features = bond_fix_features,
.ndo_neigh_construct = netdev_default_l2upper_neigh_construct,
.ndo_neigh_destroy = netdev_default_l2upper_neigh_destroy,
.ndo_bridge_setlink = switchdev_port_bridge_setlink,
.ndo_bridge_getlink = switchdev_port_bridge_getlink,
.ndo_bridge_dellink = switchdev_port_bridge_dellink,
......
......@@ -7,5 +7,6 @@ obj-$(CONFIG_MLXSW_SWITCHX2) += mlxsw_switchx2.o
mlxsw_switchx2-objs := switchx2.o
obj-$(CONFIG_MLXSW_SPECTRUM) += mlxsw_spectrum.o
mlxsw_spectrum-objs := spectrum.o spectrum_buffers.o \
spectrum_switchdev.o spectrum_router.o
spectrum_switchdev.o spectrum_router.o \
spectrum_kvdl.o
mlxsw_spectrum-$(CONFIG_MLXSW_SPECTRUM_DCB) += spectrum_dcb.o
......@@ -607,6 +607,24 @@ MLXSW_ITEM32(cmd_mbox, config_profile,
*/
MLXSW_ITEM32(cmd_mbox, config_profile, set_ar_sec, 0x0C, 15, 1);
/* cmd_mbox_config_set_kvd_linear_size
* Capability bit. Setting a bit to 1 configures the profile
* according to the mailbox contents.
*/
MLXSW_ITEM32(cmd_mbox, config_profile, set_kvd_linear_size, 0x0C, 24, 1);
/* cmd_mbox_config_set_kvd_hash_single_size
* Capability bit. Setting a bit to 1 configures the profile
* according to the mailbox contents.
*/
MLXSW_ITEM32(cmd_mbox, config_profile, set_kvd_hash_single_size, 0x0C, 25, 1);
/* cmd_mbox_config_set_kvd_hash_double_size
* Capability bit. Setting a bit to 1 configures the profile
* according to the mailbox contents.
*/
MLXSW_ITEM32(cmd_mbox, config_profile, set_kvd_hash_double_size, 0x0C, 26, 1);
/* cmd_mbox_config_profile_max_vepa_channels
* Maximum number of VEPA channels per port (0 through 16)
* 0 - multi-channel VEPA is disabled
......@@ -733,6 +751,31 @@ MLXSW_ITEM32(cmd_mbox, config_profile, adaptive_routing_group_cap, 0x4C, 0, 16);
*/
MLXSW_ITEM32(cmd_mbox, config_profile, arn, 0x50, 31, 1);
/* cmd_mbox_config_kvd_linear_size
* KVD Linear Size
* Valid for Spectrum only
* Allowed values are 128*N where N=0 or higher
*/
MLXSW_ITEM32(cmd_mbox, config_profile, kvd_linear_size, 0x54, 0, 24);
/* cmd_mbox_config_kvd_hash_single_size
* KVD Hash single-entries size
* Valid for Spectrum only
* Allowed values are 128*N where N=0 or higher
* Must be greater or equal to cap_min_kvd_hash_single_size
* Must be smaller or equal to cap_kvd_size - kvd_linear_size
*/
MLXSW_ITEM32(cmd_mbox, config_profile, kvd_hash_single_size, 0x58, 0, 24);
/* cmd_mbox_config_kvd_hash_double_size
* KVD Hash double-entries size (units of single-size entries)
* Valid for Spectrum only
* Allowed values are 128*N where N=0 or higher
* Must be either 0 or greater or equal to cap_min_kvd_hash_double_size
* Must be smaller or equal to cap_kvd_size - kvd_linear_size
*/
MLXSW_ITEM32(cmd_mbox, config_profile, kvd_hash_double_size, 0x5C, 0, 24);
/* cmd_mbox_config_profile_swid_config_mask
* Modify Switch Partition Configuration mask. When set, the configu-
* ration value for the Switch Partition are taken from the mailbox.
......
......@@ -190,7 +190,8 @@ struct mlxsw_config_profile {
used_max_ib_mc:1,
used_max_pkey:1,
used_ar_sec:1,
used_adaptive_routing_group_cap:1;
used_adaptive_routing_group_cap:1,
used_kvd_sizes:1;
u8 max_vepa_channels;
u16 max_lag;
u16 max_port_per_lag;
......@@ -211,6 +212,9 @@ struct mlxsw_config_profile {
u8 ar_sec;
u16 adaptive_routing_group_cap;
u8 arn;
u32 kvd_linear_size;
u32 kvd_hash_single_size;
u32 kvd_hash_double_size;
struct mlxsw_swid_config swid_config[MLXSW_CONFIG_PROFILE_SWID_COUNT];
};
......
......@@ -1255,6 +1255,20 @@ static int mlxsw_pci_config_profile(struct mlxsw_pci *mlxsw_pci, char *mbox,
mlxsw_cmd_mbox_config_profile_adaptive_routing_group_cap_set(
mbox, profile->adaptive_routing_group_cap);
}
if (profile->used_kvd_sizes) {
mlxsw_cmd_mbox_config_profile_set_kvd_linear_size_set(
mbox, 1);
mlxsw_cmd_mbox_config_profile_kvd_linear_size_set(
mbox, profile->kvd_linear_size);
mlxsw_cmd_mbox_config_profile_set_kvd_hash_single_size_set(
mbox, 1);
mlxsw_cmd_mbox_config_profile_kvd_hash_single_size_set(
mbox, profile->kvd_hash_single_size);
mlxsw_cmd_mbox_config_profile_set_kvd_hash_double_size_set(
mbox, 1);
mlxsw_cmd_mbox_config_profile_kvd_hash_double_size_set(
mbox, profile->kvd_hash_double_size);
}
for (i = 0; i < MLXSW_CONFIG_PROFILE_SWID_COUNT; i++)
mlxsw_pci_config_profile_swid_config(mlxsw_pci, mbox, i,
......
This diff is collapsed.
......@@ -803,6 +803,8 @@ static const struct net_device_ops mlxsw_sp_port_netdev_ops = {
.ndo_get_stats64 = mlxsw_sp_port_get_stats64,
.ndo_vlan_rx_add_vid = mlxsw_sp_port_add_vid,
.ndo_vlan_rx_kill_vid = mlxsw_sp_port_kill_vid,
.ndo_neigh_construct = mlxsw_sp_router_neigh_construct,
.ndo_neigh_destroy = mlxsw_sp_router_neigh_destroy,
.ndo_fdb_add = switchdev_port_fdb_add,
.ndo_fdb_del = switchdev_port_fdb_del,
.ndo_fdb_dump = switchdev_port_fdb_dump,
......@@ -2354,6 +2356,10 @@ static struct mlxsw_config_profile mlxsw_sp_config_profile = {
.max_ib_mc = 0,
.used_max_pkey = 1,
.max_pkey = 0,
.used_kvd_sizes = 1,
.kvd_linear_size = MLXSW_SP_KVD_LINEAR_SIZE,
.kvd_hash_single_size = MLXSW_SP_KVD_HASH_SINGLE_SIZE,
.kvd_hash_double_size = MLXSW_SP_KVD_HASH_DOUBLE_SIZE,
.swid_config = {
{
.used_type = 1,
......
......@@ -39,6 +39,7 @@
#include <linux/types.h>
#include <linux/netdevice.h>
#include <linux/rhashtable.h>
#include <linux/bitops.h>
#include <linux/if_vlan.h>
#include <linux/list.h>
......@@ -75,6 +76,10 @@
#define MLXSW_SP_BYTES_TO_CELLS(b) DIV_ROUND_UP(b, MLXSW_SP_BYTES_PER_CELL)
#define MLXSW_SP_CELLS_TO_BYTES(c) (c * MLXSW_SP_BYTES_PER_CELL)
#define MLXSW_SP_KVD_LINEAR_SIZE 65536 /* entries */
#define MLXSW_SP_KVD_HASH_SINGLE_SIZE 163840 /* entries */
#define MLXSW_SP_KVD_HASH_DOUBLE_SIZE 32768 /* entries */
/* Maximum delay buffer needed in case of PAUSE frames, in cells.
* Assumes 100m cable and maximum MTU.
*/
......@@ -212,6 +217,15 @@ struct mlxsw_sp_vr {
struct mlxsw_sp_router {
struct mlxsw_sp_lpm_tree lpm_trees[MLXSW_SP_LPM_TREE_COUNT];
struct mlxsw_sp_vr vrs[MLXSW_SP_VIRTUAL_ROUTER_MAX];
struct rhashtable neigh_ht;
struct {
struct delayed_work dw;
unsigned long interval; /* ms */
} neighs_update;
struct delayed_work nexthop_probe_dw;
#define MLXSW_SP_UNRESOLVED_NH_PROBE_INTERVAL 5000 /* ms */
struct list_head nexthop_group_list;
struct list_head nexthop_neighs_list;
};
struct mlxsw_sp {
......@@ -243,6 +257,9 @@ struct mlxsw_sp {
u8 port_to_module[MLXSW_PORT_MAX_PORTS];
struct mlxsw_sp_sb sb;
struct mlxsw_sp_router router;
struct {
DECLARE_BITMAP(usage, MLXSW_SP_KVD_LINEAR_SIZE);
} kvdl;
};
static inline struct mlxsw_sp_upper *
......@@ -524,5 +541,12 @@ int mlxsw_sp_router_fib4_add(struct mlxsw_sp_port *mlxsw_sp_port,
struct switchdev_trans *trans);
int mlxsw_sp_router_fib4_del(struct mlxsw_sp_port *mlxsw_sp_port,
const struct switchdev_obj_ipv4_fib *fib4);
int mlxsw_sp_router_neigh_construct(struct net_device *dev,
struct neighbour *n);
void mlxsw_sp_router_neigh_destroy(struct net_device *dev,
struct neighbour *n);
int mlxsw_sp_kvdl_alloc(struct mlxsw_sp *mlxsw_sp, unsigned int entry_count);
void mlxsw_sp_kvdl_free(struct mlxsw_sp *mlxsw_sp, int entry_index);
#endif
/*
* drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c
* Copyright (c) 2016 Mellanox Technologies. All rights reserved.
* Copyright (c) 2016 Jiri Pirko <jiri@mellanox.com>
*
* 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.
* 3. Neither the names of the copyright holders nor the names of its
* contributors may be used to endorse or promote products derived from
* this software without specific prior written permission.
*
* Alternatively, this software may be distributed under the terms of the
* GNU General Public License ("GPL") version 2 as published by the Free
* Software Foundation.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS 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 THE COPYRIGHT OWNER 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.
*/
#include <linux/kernel.h>
#include <linux/bitops.h>
#include "spectrum.h"
#define MLXSW_SP_KVDL_SINGLE_BASE 0
#define MLXSW_SP_KVDL_SINGLE_SIZE 16384
#define MLXSW_SP_KVDL_CHUNKS_BASE \
(MLXSW_SP_KVDL_SINGLE_BASE + MLXSW_SP_KVDL_SINGLE_SIZE)
#define MLXSW_SP_KVDL_CHUNKS_SIZE \
(MLXSW_SP_KVD_LINEAR_SIZE - MLXSW_SP_KVDL_CHUNKS_BASE)
#define MLXSW_SP_CHUNK_MAX 32
int mlxsw_sp_kvdl_alloc(struct mlxsw_sp *mlxsw_sp, unsigned int entry_count)
{
int entry_index;
int size;
int type_base;
int type_size;
int type_entries;
if (entry_count == 0 || entry_count > MLXSW_SP_CHUNK_MAX) {
return -EINVAL;
} else if (entry_count == 1) {
type_base = MLXSW_SP_KVDL_SINGLE_BASE;
type_size = MLXSW_SP_KVDL_SINGLE_SIZE;
type_entries = 1;
} else {
type_base = MLXSW_SP_KVDL_CHUNKS_BASE;
type_size = MLXSW_SP_KVDL_CHUNKS_SIZE;
type_entries = MLXSW_SP_CHUNK_MAX;
}
entry_index = type_base;
size = type_base + type_size;
for_each_clear_bit_from(entry_index, mlxsw_sp->kvdl.usage, size) {
int i;
for (i = 0; i < type_entries; i++)
set_bit(entry_index + i, mlxsw_sp->kvdl.usage);
return entry_index;
}
return -ENOBUFS;
}
void mlxsw_sp_kvdl_free(struct mlxsw_sp *mlxsw_sp, int entry_index)
{
int type_entries;
int i;
if (entry_index < MLXSW_SP_KVDL_CHUNKS_BASE)
type_entries = 1;
else
type_entries = MLXSW_SP_CHUNK_MAX;
for (i = 0; i < type_entries; i++)
clear_bit(entry_index + i, mlxsw_sp->kvdl.usage);
}
......@@ -1996,7 +1996,8 @@ static int rocker_port_change_proto_down(struct net_device *dev,
return 0;
}
static void rocker_port_neigh_destroy(struct neighbour *n)
static void rocker_port_neigh_destroy(struct net_device *dev,
struct neighbour *n)
{
struct rocker_port *rocker_port = netdev_priv(n->dev);
int err;
......
......@@ -2002,6 +2002,8 @@ static const struct net_device_ops team_netdev_ops = {
.ndo_add_slave = team_add_slave,
.ndo_del_slave = team_del_slave,
.ndo_fix_features = team_fix_features,
.ndo_neigh_construct = netdev_default_l2upper_neigh_construct,
.ndo_neigh_destroy = netdev_default_l2upper_neigh_destroy,
.ndo_change_carrier = team_change_carrier,
.ndo_bridge_setlink = switchdev_port_bridge_setlink,
.ndo_bridge_getlink = switchdev_port_bridge_getlink,
......
......@@ -1209,8 +1209,10 @@ struct net_device_ops {
netdev_features_t features);
int (*ndo_set_features)(struct net_device *dev,
netdev_features_t features);
int (*ndo_neigh_construct)(struct neighbour *n);
void (*ndo_neigh_destroy)(struct neighbour *n);
int (*ndo_neigh_construct)(struct net_device *dev,
struct neighbour *n);
void (*ndo_neigh_destroy)(struct net_device *dev,
struct neighbour *n);
int (*ndo_fdb_add)(struct ndmsg *ndm,
struct nlattr *tb[],
......@@ -3843,6 +3845,10 @@ void *netdev_lower_dev_get_private(struct net_device *dev,
struct net_device *lower_dev);
void netdev_lower_state_changed(struct net_device *lower_dev,
void *lower_state_info);
int netdev_default_l2upper_neigh_construct(struct net_device *dev,
struct neighbour *n);
void netdev_default_l2upper_neigh_destroy(struct net_device *dev,
struct neighbour *n);
/* RSS keys are 40 or 52 bytes long */
#define NETDEV_RSS_KEY_LEN 52
......
......@@ -24,6 +24,7 @@ struct netevent_redirect {
enum netevent_notif_type {
NETEVENT_NEIGH_UPDATE = 1, /* arg is struct neighbour ptr */
NETEVENT_REDIRECT, /* arg is struct netevent_redirect ptr */
NETEVENT_DELAY_PROBE_TIME_UPDATE, /* arg is struct neigh_parms ptr */
};
int register_netevent_notifier(struct notifier_block *nb);
......
......@@ -790,6 +790,8 @@ static const struct net_device_ops vlan_netdev_ops = {
.ndo_netpoll_cleanup = vlan_dev_netpoll_cleanup,
#endif
.ndo_fix_features = vlan_dev_fix_features,
.ndo_neigh_construct = netdev_default_l2upper_neigh_construct,
.ndo_neigh_destroy = netdev_default_l2upper_neigh_destroy,
.ndo_fdb_add = switchdev_port_fdb_add,
.ndo_fdb_del = switchdev_port_fdb_del,
.ndo_fdb_dump = switchdev_port_fdb_dump,
......
......@@ -286,7 +286,7 @@ static const struct neigh_ops clip_neigh_ops = {
.connected_output = neigh_direct_output,
};
static int clip_constructor(struct neighbour *neigh)
static int clip_constructor(struct net_device *dev, struct neighbour *neigh)
{
struct atmarp_entry *entry = neighbour_priv(neigh);
......
......@@ -349,6 +349,8 @@ static const struct net_device_ops br_netdev_ops = {
.ndo_add_slave = br_add_slave,
.ndo_del_slave = br_del_slave,
.ndo_fix_features = br_fix_features,
.ndo_neigh_construct = netdev_default_l2upper_neigh_construct,
.ndo_neigh_destroy = netdev_default_l2upper_neigh_destroy,
.ndo_fdb_add = br_fdb_add,
.ndo_fdb_del = br_fdb_delete,
.ndo_fdb_dump = br_fdb_dump,
......
......@@ -6087,6 +6087,50 @@ void netdev_lower_state_changed(struct net_device *lower_dev,
}
EXPORT_SYMBOL(netdev_lower_state_changed);
int netdev_default_l2upper_neigh_construct(struct net_device *dev,
struct neighbour *n)
{
struct net_device *lower_dev, *stop_dev;
struct list_head *iter;
int err;
netdev_for_each_lower_dev(dev, lower_dev, iter) {
if (!lower_dev->netdev_ops->ndo_neigh_construct)
continue;
err = lower_dev->netdev_ops->ndo_neigh_construct(lower_dev, n);
if (err) {
stop_dev = lower_dev;
goto rollback;
}
}
return 0;
rollback:
netdev_for_each_lower_dev(dev, lower_dev, iter) {
if (lower_dev == stop_dev)
break;
if (!lower_dev->netdev_ops->ndo_neigh_destroy)
continue;
lower_dev->netdev_ops->ndo_neigh_destroy(lower_dev, n);
}
return err;
}
EXPORT_SYMBOL_GPL(netdev_default_l2upper_neigh_construct);
void netdev_default_l2upper_neigh_destroy(struct net_device *dev,
struct neighbour *n)
{
struct net_device *lower_dev;
struct list_head *iter;
netdev_for_each_lower_dev(dev, lower_dev, iter) {
if (!lower_dev->netdev_ops->ndo_neigh_destroy)
continue;
lower_dev->netdev_ops->ndo_neigh_destroy(lower_dev, n);
}
}
EXPORT_SYMBOL_GPL(netdev_default_l2upper_neigh_destroy);
static void dev_change_rx_flags(struct net_device *dev, int flags)
{
const struct net_device_ops *ops = dev->netdev_ops;
......
......@@ -473,7 +473,7 @@ struct neighbour *__neigh_create(struct neigh_table *tbl, const void *pkey,
}
if (dev->netdev_ops->ndo_neigh_construct) {
error = dev->netdev_ops->ndo_neigh_construct(n);
error = dev->netdev_ops->ndo_neigh_construct(dev, n);
if (error < 0) {
rc = ERR_PTR(error);
goto out_neigh_release;
......@@ -701,7 +701,7 @@ void neigh_destroy(struct neighbour *neigh)
neigh->arp_queue_len_bytes = 0;
if (dev->netdev_ops->ndo_neigh_destroy)
dev->netdev_ops->ndo_neigh_destroy(neigh);
dev->netdev_ops->ndo_neigh_destroy(dev, neigh);
dev_put(dev);
neigh_parms_put(neigh->parms);
......@@ -2047,6 +2047,7 @@ static int neightbl_set(struct sk_buff *skb, struct nlmsghdr *nlh)
case NDTPA_DELAY_PROBE_TIME:
NEIGH_VAR_SET(p, DELAY_PROBE_TIME,
nla_get_msecs(tbp[i]));
call_netevent_notifiers(NETEVENT_DELAY_PROBE_TIME_UPDATE, p);
break;
case NDTPA_RETRANS_TIME:
NEIGH_VAR_SET(p, RETRANS_TIME,
......@@ -2930,6 +2931,7 @@ static void neigh_proc_update(struct ctl_table *ctl, int write)
return;
set_bit(index, p->data_state);
call_netevent_notifiers(NETEVENT_DELAY_PROBE_TIME_UPDATE, p);
if (!dev) /* NULL dev means this is default value */
neigh_copy_dflt_parms(net, p, index);
}
......
......@@ -81,7 +81,7 @@ static int lowpan_stop(struct net_device *dev)
return 0;
}
static int lowpan_neigh_construct(struct neighbour *n)
static int lowpan_neigh_construct(struct net_device *dev, struct neighbour *n)
{
struct lowpan_802154_neigh *neigh = lowpan_802154_neigh(neighbour_priv(n));
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment