linux_dsm_epyc7002/net/dsa/dsa.c
Alexander Lobakin e131a56348 net: dsa: add GRO support via gro_cells
gro_cells lib is used by different encapsulating netdevices, such as
geneve, macsec, vxlan etc. to speed up decapsulated traffic processing.
CPU tag is a sort of "encapsulation", and we can use the same mechs to
greatly improve overall DSA performance.
skbs are passed to the GRO layer after removing CPU tags, so we don't
need any new packet offload types as it was firstly proposed by me in
the first GRO-over-DSA variant [1].

The size of struct gro_cells is sizeof(void *), so hot struct
dsa_slave_priv becomes only 4/8 bytes bigger, and all critical fields
remain in one 32-byte cacheline.
The other positive side effect is that drivers for network devices
that can be shipped as CPU ports of DSA-driven switches can now use
napi_gro_frags() to pass skbs to kernel. Packets built that way are
completely non-linear and are likely being dropped without GRO.

This was tested on to-be-mainlined-soon Ethernet driver that uses
napi_gro_frags(), and the overall performance was on par with the
variant from [1], sometimes even better due to minimal overhead.
net.core.gro_normal_batch tuning may help to push it to the limit
on particular setups and platforms.

iperf3 IPoE VLAN NAT TCP forwarding (port1.218 -> port0) setup
on 1.2 GHz MIPS board:

5.7-rc2 baseline:

[ID]  Interval         Transfer     Bitrate        Retr
[ 5]  0.00-120.01 sec  9.00 GBytes  644 Mbits/sec  413  sender
[ 5]  0.00-120.00 sec  8.99 GBytes  644 Mbits/sec       receiver

Iface      RX packets  TX packets
eth0       7097731     7097702
port0      426050      6671829
port1      6671681     425862
port1.218  6671677     425851

With this patch:

[ID]  Interval         Transfer     Bitrate        Retr
[ 5]  0.00-120.01 sec  12.2 GBytes  870 Mbits/sec  122  sender
[ 5]  0.00-120.00 sec  12.2 GBytes  870 Mbits/sec       receiver

Iface      RX packets  TX packets
eth0       9474792     9474777
port0      455200      353288
port1      9019592     455035
port1.218  353144      455024

v2:
 - Add some performance examples in the commit message;
 - No functional changes.

[1] https://lore.kernel.org/netdev/20191230143028.27313-1-alobakin@dlink.ru/

Signed-off-by: Alexander Lobakin <bloodyreaper@yandex.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
2020-04-23 12:32:13 -07:00

456 lines
10 KiB
C

// SPDX-License-Identifier: GPL-2.0-or-later
/*
* net/dsa/dsa.c - Hardware switch handling
* Copyright (c) 2008-2009 Marvell Semiconductor
* Copyright (c) 2013 Florian Fainelli <florian@openwrt.org>
*/
#include <linux/device.h>
#include <linux/list.h>
#include <linux/platform_device.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/notifier.h>
#include <linux/of.h>
#include <linux/of_mdio.h>
#include <linux/of_platform.h>
#include <linux/of_net.h>
#include <linux/netdevice.h>
#include <linux/sysfs.h>
#include <linux/phy_fixed.h>
#include <linux/ptp_classify.h>
#include <linux/etherdevice.h>
#include "dsa_priv.h"
static LIST_HEAD(dsa_tag_drivers_list);
static DEFINE_MUTEX(dsa_tag_drivers_lock);
static struct sk_buff *dsa_slave_notag_xmit(struct sk_buff *skb,
struct net_device *dev)
{
/* Just return the original SKB */
return skb;
}
static const struct dsa_device_ops none_ops = {
.name = "none",
.proto = DSA_TAG_PROTO_NONE,
.xmit = dsa_slave_notag_xmit,
.rcv = NULL,
};
DSA_TAG_DRIVER(none_ops);
static void dsa_tag_driver_register(struct dsa_tag_driver *dsa_tag_driver,
struct module *owner)
{
dsa_tag_driver->owner = owner;
mutex_lock(&dsa_tag_drivers_lock);
list_add_tail(&dsa_tag_driver->list, &dsa_tag_drivers_list);
mutex_unlock(&dsa_tag_drivers_lock);
}
void dsa_tag_drivers_register(struct dsa_tag_driver *dsa_tag_driver_array[],
unsigned int count, struct module *owner)
{
unsigned int i;
for (i = 0; i < count; i++)
dsa_tag_driver_register(dsa_tag_driver_array[i], owner);
}
static void dsa_tag_driver_unregister(struct dsa_tag_driver *dsa_tag_driver)
{
mutex_lock(&dsa_tag_drivers_lock);
list_del(&dsa_tag_driver->list);
mutex_unlock(&dsa_tag_drivers_lock);
}
EXPORT_SYMBOL_GPL(dsa_tag_drivers_register);
void dsa_tag_drivers_unregister(struct dsa_tag_driver *dsa_tag_driver_array[],
unsigned int count)
{
unsigned int i;
for (i = 0; i < count; i++)
dsa_tag_driver_unregister(dsa_tag_driver_array[i]);
}
EXPORT_SYMBOL_GPL(dsa_tag_drivers_unregister);
const char *dsa_tag_protocol_to_str(const struct dsa_device_ops *ops)
{
return ops->name;
};
const struct dsa_device_ops *dsa_tag_driver_get(int tag_protocol)
{
struct dsa_tag_driver *dsa_tag_driver;
const struct dsa_device_ops *ops;
bool found = false;
request_module("%s%d", DSA_TAG_DRIVER_ALIAS, tag_protocol);
mutex_lock(&dsa_tag_drivers_lock);
list_for_each_entry(dsa_tag_driver, &dsa_tag_drivers_list, list) {
ops = dsa_tag_driver->ops;
if (ops->proto == tag_protocol) {
found = true;
break;
}
}
if (found) {
if (!try_module_get(dsa_tag_driver->owner))
ops = ERR_PTR(-ENOPROTOOPT);
} else {
ops = ERR_PTR(-ENOPROTOOPT);
}
mutex_unlock(&dsa_tag_drivers_lock);
return ops;
}
void dsa_tag_driver_put(const struct dsa_device_ops *ops)
{
struct dsa_tag_driver *dsa_tag_driver;
mutex_lock(&dsa_tag_drivers_lock);
list_for_each_entry(dsa_tag_driver, &dsa_tag_drivers_list, list) {
if (dsa_tag_driver->ops == ops) {
module_put(dsa_tag_driver->owner);
break;
}
}
mutex_unlock(&dsa_tag_drivers_lock);
}
static int dev_is_class(struct device *dev, void *class)
{
if (dev->class != NULL && !strcmp(dev->class->name, class))
return 1;
return 0;
}
static struct device *dev_find_class(struct device *parent, char *class)
{
if (dev_is_class(parent, class)) {
get_device(parent);
return parent;
}
return device_find_child(parent, class, dev_is_class);
}
struct net_device *dsa_dev_to_net_device(struct device *dev)
{
struct device *d;
d = dev_find_class(dev, "net");
if (d != NULL) {
struct net_device *nd;
nd = to_net_dev(d);
dev_hold(nd);
put_device(d);
return nd;
}
return NULL;
}
EXPORT_SYMBOL_GPL(dsa_dev_to_net_device);
/* Determine if we should defer delivery of skb until we have a rx timestamp.
*
* Called from dsa_switch_rcv. For now, this will only work if tagging is
* enabled on the switch. Normally the MAC driver would retrieve the hardware
* timestamp when it reads the packet out of the hardware. However in a DSA
* switch, the DSA driver owning the interface to which the packet is
* delivered is never notified unless we do so here.
*/
static bool dsa_skb_defer_rx_timestamp(struct dsa_slave_priv *p,
struct sk_buff *skb)
{
struct dsa_switch *ds = p->dp->ds;
unsigned int type;
if (skb_headroom(skb) < ETH_HLEN)
return false;
__skb_push(skb, ETH_HLEN);
type = ptp_classify_raw(skb);
__skb_pull(skb, ETH_HLEN);
if (type == PTP_CLASS_NONE)
return false;
if (likely(ds->ops->port_rxtstamp))
return ds->ops->port_rxtstamp(ds, p->dp->index, skb, type);
return false;
}
static int dsa_switch_rcv(struct sk_buff *skb, struct net_device *dev,
struct packet_type *pt, struct net_device *unused)
{
struct dsa_port *cpu_dp = dev->dsa_ptr;
struct sk_buff *nskb = NULL;
struct pcpu_sw_netstats *s;
struct dsa_slave_priv *p;
if (unlikely(!cpu_dp)) {
kfree_skb(skb);
return 0;
}
skb = skb_unshare(skb, GFP_ATOMIC);
if (!skb)
return 0;
nskb = cpu_dp->rcv(skb, dev, pt);
if (!nskb) {
kfree_skb(skb);
return 0;
}
skb = nskb;
p = netdev_priv(skb->dev);
skb_push(skb, ETH_HLEN);
skb->pkt_type = PACKET_HOST;
skb->protocol = eth_type_trans(skb, skb->dev);
s = this_cpu_ptr(p->stats64);
u64_stats_update_begin(&s->syncp);
s->rx_packets++;
s->rx_bytes += skb->len;
u64_stats_update_end(&s->syncp);
if (dsa_skb_defer_rx_timestamp(p, skb))
return 0;
gro_cells_receive(&p->gcells, skb);
return 0;
}
#ifdef CONFIG_PM_SLEEP
static bool dsa_is_port_initialized(struct dsa_switch *ds, int p)
{
const struct dsa_port *dp = dsa_to_port(ds, p);
return dp->type == DSA_PORT_TYPE_USER && dp->slave;
}
int dsa_switch_suspend(struct dsa_switch *ds)
{
int i, ret = 0;
/* Suspend slave network devices */
for (i = 0; i < ds->num_ports; i++) {
if (!dsa_is_port_initialized(ds, i))
continue;
ret = dsa_slave_suspend(dsa_to_port(ds, i)->slave);
if (ret)
return ret;
}
if (ds->ops->suspend)
ret = ds->ops->suspend(ds);
return ret;
}
EXPORT_SYMBOL_GPL(dsa_switch_suspend);
int dsa_switch_resume(struct dsa_switch *ds)
{
int i, ret = 0;
if (ds->ops->resume)
ret = ds->ops->resume(ds);
if (ret)
return ret;
/* Resume slave network devices */
for (i = 0; i < ds->num_ports; i++) {
if (!dsa_is_port_initialized(ds, i))
continue;
ret = dsa_slave_resume(dsa_to_port(ds, i)->slave);
if (ret)
return ret;
}
return 0;
}
EXPORT_SYMBOL_GPL(dsa_switch_resume);
#endif
static struct packet_type dsa_pack_type __read_mostly = {
.type = cpu_to_be16(ETH_P_XDSA),
.func = dsa_switch_rcv,
};
static struct workqueue_struct *dsa_owq;
bool dsa_schedule_work(struct work_struct *work)
{
return queue_work(dsa_owq, work);
}
static ATOMIC_NOTIFIER_HEAD(dsa_notif_chain);
int register_dsa_notifier(struct notifier_block *nb)
{
return atomic_notifier_chain_register(&dsa_notif_chain, nb);
}
EXPORT_SYMBOL_GPL(register_dsa_notifier);
int unregister_dsa_notifier(struct notifier_block *nb)
{
return atomic_notifier_chain_unregister(&dsa_notif_chain, nb);
}
EXPORT_SYMBOL_GPL(unregister_dsa_notifier);
int call_dsa_notifiers(unsigned long val, struct net_device *dev,
struct dsa_notifier_info *info)
{
info->dev = dev;
return atomic_notifier_call_chain(&dsa_notif_chain, val, info);
}
EXPORT_SYMBOL_GPL(call_dsa_notifiers);
int dsa_devlink_param_get(struct devlink *dl, u32 id,
struct devlink_param_gset_ctx *ctx)
{
struct dsa_devlink_priv *dl_priv;
struct dsa_switch *ds;
dl_priv = devlink_priv(dl);
ds = dl_priv->ds;
if (!ds->ops->devlink_param_get)
return -EOPNOTSUPP;
return ds->ops->devlink_param_get(ds, id, ctx);
}
EXPORT_SYMBOL_GPL(dsa_devlink_param_get);
int dsa_devlink_param_set(struct devlink *dl, u32 id,
struct devlink_param_gset_ctx *ctx)
{
struct dsa_devlink_priv *dl_priv;
struct dsa_switch *ds;
dl_priv = devlink_priv(dl);
ds = dl_priv->ds;
if (!ds->ops->devlink_param_set)
return -EOPNOTSUPP;
return ds->ops->devlink_param_set(ds, id, ctx);
}
EXPORT_SYMBOL_GPL(dsa_devlink_param_set);
int dsa_devlink_params_register(struct dsa_switch *ds,
const struct devlink_param *params,
size_t params_count)
{
return devlink_params_register(ds->devlink, params, params_count);
}
EXPORT_SYMBOL_GPL(dsa_devlink_params_register);
void dsa_devlink_params_unregister(struct dsa_switch *ds,
const struct devlink_param *params,
size_t params_count)
{
devlink_params_unregister(ds->devlink, params, params_count);
}
EXPORT_SYMBOL_GPL(dsa_devlink_params_unregister);
int dsa_devlink_resource_register(struct dsa_switch *ds,
const char *resource_name,
u64 resource_size,
u64 resource_id,
u64 parent_resource_id,
const struct devlink_resource_size_params *size_params)
{
return devlink_resource_register(ds->devlink, resource_name,
resource_size, resource_id,
parent_resource_id,
size_params);
}
EXPORT_SYMBOL_GPL(dsa_devlink_resource_register);
void dsa_devlink_resources_unregister(struct dsa_switch *ds)
{
devlink_resources_unregister(ds->devlink, NULL);
}
EXPORT_SYMBOL_GPL(dsa_devlink_resources_unregister);
void dsa_devlink_resource_occ_get_register(struct dsa_switch *ds,
u64 resource_id,
devlink_resource_occ_get_t *occ_get,
void *occ_get_priv)
{
return devlink_resource_occ_get_register(ds->devlink, resource_id,
occ_get, occ_get_priv);
}
EXPORT_SYMBOL_GPL(dsa_devlink_resource_occ_get_register);
void dsa_devlink_resource_occ_get_unregister(struct dsa_switch *ds,
u64 resource_id)
{
devlink_resource_occ_get_unregister(ds->devlink, resource_id);
}
EXPORT_SYMBOL_GPL(dsa_devlink_resource_occ_get_unregister);
static int __init dsa_init_module(void)
{
int rc;
dsa_owq = alloc_ordered_workqueue("dsa_ordered",
WQ_MEM_RECLAIM);
if (!dsa_owq)
return -ENOMEM;
rc = dsa_slave_register_notifier();
if (rc)
goto register_notifier_fail;
dev_add_pack(&dsa_pack_type);
dsa_tag_driver_register(&DSA_TAG_DRIVER_NAME(none_ops),
THIS_MODULE);
return 0;
register_notifier_fail:
destroy_workqueue(dsa_owq);
return rc;
}
module_init(dsa_init_module);
static void __exit dsa_cleanup_module(void)
{
dsa_tag_driver_unregister(&DSA_TAG_DRIVER_NAME(none_ops));
dsa_slave_unregister_notifier();
dev_remove_pack(&dsa_pack_type);
destroy_workqueue(dsa_owq);
}
module_exit(dsa_cleanup_module);
MODULE_AUTHOR("Lennert Buytenhek <buytenh@wantstofly.org>");
MODULE_DESCRIPTION("Driver for Distributed Switch Architecture switch chips");
MODULE_LICENSE("GPL");
MODULE_ALIAS("platform:dsa");