2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* INET An implementation of the TCP/IP protocol suite for the LINUX
|
|
|
|
* operating system. INET is implemented using the BSD Socket
|
|
|
|
* interface as the means of communication with the user level.
|
|
|
|
*
|
|
|
|
* Definitions for the Interfaces handler.
|
|
|
|
*
|
|
|
|
* Version: @(#)dev.h 1.0.10 08/12/93
|
|
|
|
*
|
2005-05-06 07:16:16 +08:00
|
|
|
* Authors: Ross Biro
|
2005-04-17 06:20:36 +08:00
|
|
|
* Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
|
|
|
|
* Corey Minyard <wf-rch!minyard@relay.EU.net>
|
|
|
|
* Donald J. Becker, <becker@cesdis.gsfc.nasa.gov>
|
2008-10-14 10:01:08 +08:00
|
|
|
* Alan Cox, <alan@lxorguk.ukuu.org.uk>
|
2005-04-17 06:20:36 +08:00
|
|
|
* Bjorn Ekwall. <bj0rn@blox.se>
|
|
|
|
* Pekka Riikonen <priikone@poseidon.pspt.fi>
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version
|
|
|
|
* 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* Moved to /usr/include/linux for NET3
|
|
|
|
*/
|
|
|
|
#ifndef _LINUX_NETDEVICE_H
|
|
|
|
#define _LINUX_NETDEVICE_H
|
|
|
|
|
2006-12-04 12:15:30 +08:00
|
|
|
#include <linux/timer.h>
|
2011-11-24 09:12:59 +08:00
|
|
|
#include <linux/bug.h>
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
#include <linux/delay.h>
|
2011-07-27 07:09:06 +08:00
|
|
|
#include <linux/atomic.h>
|
2014-10-08 23:19:27 +08:00
|
|
|
#include <linux/prefetch.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <asm/cache.h>
|
|
|
|
#include <asm/byteorder.h>
|
|
|
|
|
|
|
|
#include <linux/percpu.h>
|
2009-05-07 07:52:51 +08:00
|
|
|
#include <linux/rculist.h>
|
2006-06-18 12:24:58 +08:00
|
|
|
#include <linux/dmaengine.h>
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
#include <linux/workqueue.h>
|
2011-11-29 00:33:09 +08:00
|
|
|
#include <linux/dynamic_queue_limits.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-04-20 12:49:28 +08:00
|
|
|
#include <linux/ethtool.h>
|
2007-09-12 20:57:09 +08:00
|
|
|
#include <net/net_namespace.h>
|
2008-10-07 21:45:02 +08:00
|
|
|
#include <net/dsa.h>
|
2008-11-25 17:02:08 +08:00
|
|
|
#ifdef CONFIG_DCB
|
2008-11-21 12:52:10 +08:00
|
|
|
#include <net/dcbnl.h>
|
|
|
|
#endif
|
2011-11-22 13:10:51 +08:00
|
|
|
#include <net/netprio_cgroup.h>
|
2007-09-12 20:57:09 +08:00
|
|
|
|
2011-11-15 23:29:55 +08:00
|
|
|
#include <linux/netdev_features.h>
|
2012-04-15 14:43:56 +08:00
|
|
|
#include <linux/neighbour.h>
|
2012-10-13 17:46:48 +08:00
|
|
|
#include <uapi/linux/netdevice.h>
|
2015-02-03 22:48:29 +08:00
|
|
|
#include <uapi/linux/if_bonding.h>
|
2011-11-15 23:29:55 +08:00
|
|
|
|
2005-06-23 13:05:31 +08:00
|
|
|
struct netpoll_info;
|
2012-01-31 00:46:54 +08:00
|
|
|
struct device;
|
2010-07-17 16:49:36 +08:00
|
|
|
struct phy_device;
|
2007-04-24 03:20:05 +08:00
|
|
|
/* 802.11 specific */
|
|
|
|
struct wireless_dev;
|
2014-11-02 13:44:54 +08:00
|
|
|
/* 802.15.4 specific */
|
|
|
|
struct wpan_dev;
|
2015-04-22 18:14:37 +08:00
|
|
|
struct mpls_dev;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void netdev_set_default_ethtool_ops(struct net_device *dev,
|
|
|
|
const struct ethtool_ops *ops);
|
2013-01-11 07:19:10 +08:00
|
|
|
|
2009-11-15 15:20:12 +08:00
|
|
|
/* Backlog congestion levels */
|
|
|
|
#define NET_RX_SUCCESS 0 /* keep 'em coming, baby */
|
|
|
|
#define NET_RX_DROP 1 /* packet dropped */
|
|
|
|
|
2009-11-10 14:14:14 +08:00
|
|
|
/*
|
|
|
|
* Transmit return codes: transmit return codes originate from three different
|
|
|
|
* namespaces:
|
|
|
|
*
|
|
|
|
* - qdisc return codes
|
|
|
|
* - driver transmit return codes
|
|
|
|
* - errno values
|
|
|
|
*
|
|
|
|
* Drivers are allowed to return any one of those in their hard_start_xmit()
|
|
|
|
* function. Real network devices commonly used with qdiscs should only return
|
|
|
|
* the driver transmit return codes though - when qdiscs are used, the actual
|
|
|
|
* transmission happens asynchronously, so the value is not propagated to
|
|
|
|
* higher layers. Virtual network devices transmit synchronously, in this case
|
|
|
|
* the driver transmit return codes are consumed by dev_queue_xmit(), all
|
|
|
|
* others are propagated to higher layers.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* qdisc ->enqueue() return codes. */
|
|
|
|
#define NET_XMIT_SUCCESS 0x00
|
2009-11-15 15:20:12 +08:00
|
|
|
#define NET_XMIT_DROP 0x01 /* skb dropped */
|
|
|
|
#define NET_XMIT_CN 0x02 /* congestion notification */
|
|
|
|
#define NET_XMIT_POLICED 0x03 /* skb is shot by police */
|
|
|
|
#define NET_XMIT_MASK 0x0f /* qdisc flags in net/sch_generic.h */
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2006-11-14 21:21:36 +08:00
|
|
|
/* NET_XMIT_CN is special. It does not guarantee that this packet is lost. It
|
|
|
|
* indicates that the device will soon be dropping packets, or already drops
|
|
|
|
* some packets of the same priority; prompting us to send less aggressively. */
|
2009-11-10 14:14:14 +08:00
|
|
|
#define net_xmit_eval(e) ((e) == NET_XMIT_CN ? 0 : (e))
|
2005-04-17 06:20:36 +08:00
|
|
|
#define net_xmit_errno(e) ((e) != NET_XMIT_CN ? -ENOBUFS : 0)
|
|
|
|
|
2009-09-01 03:50:40 +08:00
|
|
|
/* Driver transmit return codes */
|
2009-11-15 15:20:12 +08:00
|
|
|
#define NETDEV_TX_MASK 0xf0
|
2009-11-10 14:14:14 +08:00
|
|
|
|
2009-09-01 03:50:40 +08:00
|
|
|
enum netdev_tx {
|
2009-11-10 14:14:14 +08:00
|
|
|
__NETDEV_TX_MIN = INT_MIN, /* make sure enum is signed */
|
2009-11-15 15:20:12 +08:00
|
|
|
NETDEV_TX_OK = 0x00, /* driver took care of packet */
|
|
|
|
NETDEV_TX_BUSY = 0x10, /* driver tx path was busy*/
|
|
|
|
NETDEV_TX_LOCKED = 0x20, /* driver tx lock was already taken */
|
2009-09-01 03:50:40 +08:00
|
|
|
};
|
|
|
|
typedef enum netdev_tx netdev_tx_t;
|
|
|
|
|
2009-11-15 15:20:12 +08:00
|
|
|
/*
|
|
|
|
* Current order: NETDEV_TX_MASK > NET_XMIT_MASK >= 0 is significant;
|
|
|
|
* hard_start_xmit() return < NET_XMIT_MASK means skb was consumed.
|
|
|
|
*/
|
|
|
|
static inline bool dev_xmit_complete(int rc)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Positive cases with an skb consumed by a driver:
|
|
|
|
* - successful transmission (rc == NETDEV_TX_OK)
|
|
|
|
* - error while transmitting (rc < 0)
|
|
|
|
* - error while queueing to a different device (rc & NET_XMIT_MASK)
|
|
|
|
*/
|
|
|
|
if (likely(rc < NET_XMIT_MASK))
|
|
|
|
return true;
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* Compute the worst case header length according to the protocols
|
|
|
|
* used.
|
|
|
|
*/
|
2009-02-06 13:26:19 +08:00
|
|
|
|
2011-11-25 22:40:26 +08:00
|
|
|
#if defined(CONFIG_WLAN) || IS_ENABLED(CONFIG_AX25)
|
2008-05-13 11:17:33 +08:00
|
|
|
# if defined(CONFIG_MAC80211_MESH)
|
|
|
|
# define LL_MAX_HEADER 128
|
|
|
|
# else
|
|
|
|
# define LL_MAX_HEADER 96
|
|
|
|
# endif
|
2005-04-17 06:20:36 +08:00
|
|
|
#else
|
2008-05-13 11:17:33 +08:00
|
|
|
# define LL_MAX_HEADER 32
|
2005-04-17 06:20:36 +08:00
|
|
|
#endif
|
|
|
|
|
2011-11-25 22:40:26 +08:00
|
|
|
#if !IS_ENABLED(CONFIG_NET_IPIP) && !IS_ENABLED(CONFIG_NET_IPGRE) && \
|
|
|
|
!IS_ENABLED(CONFIG_IPV6_SIT) && !IS_ENABLED(CONFIG_IPV6_TUNNEL)
|
2005-04-17 06:20:36 +08:00
|
|
|
#define MAX_HEADER LL_MAX_HEADER
|
|
|
|
#else
|
|
|
|
#define MAX_HEADER (LL_MAX_HEADER + 48)
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/*
|
2010-06-08 15:19:54 +08:00
|
|
|
* Old network device statistics. Fields are native words
|
|
|
|
* (unsigned long) so they can be read and written atomically.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2009-02-06 13:26:19 +08:00
|
|
|
|
2009-11-05 01:50:58 +08:00
|
|
|
struct net_device_stats {
|
2010-07-09 17:11:52 +08:00
|
|
|
unsigned long rx_packets;
|
|
|
|
unsigned long tx_packets;
|
|
|
|
unsigned long rx_bytes;
|
|
|
|
unsigned long tx_bytes;
|
|
|
|
unsigned long rx_errors;
|
|
|
|
unsigned long tx_errors;
|
|
|
|
unsigned long rx_dropped;
|
|
|
|
unsigned long tx_dropped;
|
|
|
|
unsigned long multicast;
|
2005-04-17 06:20:36 +08:00
|
|
|
unsigned long collisions;
|
|
|
|
unsigned long rx_length_errors;
|
2010-07-09 17:11:52 +08:00
|
|
|
unsigned long rx_over_errors;
|
|
|
|
unsigned long rx_crc_errors;
|
|
|
|
unsigned long rx_frame_errors;
|
|
|
|
unsigned long rx_fifo_errors;
|
|
|
|
unsigned long rx_missed_errors;
|
2005-04-17 06:20:36 +08:00
|
|
|
unsigned long tx_aborted_errors;
|
|
|
|
unsigned long tx_carrier_errors;
|
|
|
|
unsigned long tx_fifo_errors;
|
|
|
|
unsigned long tx_heartbeat_errors;
|
|
|
|
unsigned long tx_window_errors;
|
|
|
|
unsigned long rx_compressed;
|
|
|
|
unsigned long tx_compressed;
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
#include <linux/cache.h>
|
|
|
|
#include <linux/skbuff.h>
|
|
|
|
|
2011-11-17 11:13:26 +08:00
|
|
|
#ifdef CONFIG_RPS
|
2012-02-24 15:31:31 +08:00
|
|
|
#include <linux/static_key.h>
|
|
|
|
extern struct static_key rps_needed;
|
2011-11-17 11:13:26 +08:00
|
|
|
#endif
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
struct neighbour;
|
|
|
|
struct neigh_parms;
|
|
|
|
struct sk_buff;
|
|
|
|
|
2009-05-05 10:48:28 +08:00
|
|
|
struct netdev_hw_addr {
|
|
|
|
struct list_head list;
|
|
|
|
unsigned char addr[MAX_ADDR_LEN];
|
|
|
|
unsigned char type;
|
2009-05-23 07:22:17 +08:00
|
|
|
#define NETDEV_HW_ADDR_T_LAN 1
|
|
|
|
#define NETDEV_HW_ADDR_T_SAN 2
|
|
|
|
#define NETDEV_HW_ADDR_T_SLAVE 3
|
|
|
|
#define NETDEV_HW_ADDR_T_UNICAST 4
|
2010-04-02 05:22:57 +08:00
|
|
|
#define NETDEV_HW_ADDR_T_MULTICAST 5
|
|
|
|
bool global_use;
|
2013-04-15 17:54:25 +08:00
|
|
|
int sync_cnt;
|
2010-09-20 02:24:02 +08:00
|
|
|
int refcount;
|
2013-04-03 05:10:07 +08:00
|
|
|
int synced;
|
2009-05-05 10:48:28 +08:00
|
|
|
struct rcu_head rcu_head;
|
|
|
|
};
|
|
|
|
|
2009-06-17 09:12:19 +08:00
|
|
|
struct netdev_hw_addr_list {
|
|
|
|
struct list_head list;
|
|
|
|
int count;
|
|
|
|
};
|
|
|
|
|
2010-04-02 05:22:57 +08:00
|
|
|
#define netdev_hw_addr_list_count(l) ((l)->count)
|
|
|
|
#define netdev_hw_addr_list_empty(l) (netdev_hw_addr_list_count(l) == 0)
|
|
|
|
#define netdev_hw_addr_list_for_each(ha, l) \
|
|
|
|
list_for_each_entry(ha, &(l)->list, list)
|
2010-01-26 05:36:10 +08:00
|
|
|
|
2010-04-02 05:22:57 +08:00
|
|
|
#define netdev_uc_count(dev) netdev_hw_addr_list_count(&(dev)->uc)
|
|
|
|
#define netdev_uc_empty(dev) netdev_hw_addr_list_empty(&(dev)->uc)
|
|
|
|
#define netdev_for_each_uc_addr(ha, dev) \
|
|
|
|
netdev_hw_addr_list_for_each(ha, &(dev)->uc)
|
2010-02-05 02:22:25 +08:00
|
|
|
|
2010-04-02 05:22:57 +08:00
|
|
|
#define netdev_mc_count(dev) netdev_hw_addr_list_count(&(dev)->mc)
|
|
|
|
#define netdev_mc_empty(dev) netdev_hw_addr_list_empty(&(dev)->mc)
|
2010-04-08 07:40:09 +08:00
|
|
|
#define netdev_for_each_mc_addr(ha, dev) \
|
2010-04-02 05:22:57 +08:00
|
|
|
netdev_hw_addr_list_for_each(ha, &(dev)->mc)
|
2010-02-05 02:22:25 +08:00
|
|
|
|
2009-11-05 01:50:58 +08:00
|
|
|
struct hh_cache {
|
2011-07-14 22:53:20 +08:00
|
|
|
u16 hh_len;
|
2011-07-13 15:51:10 +08:00
|
|
|
u16 __pad;
|
2006-12-08 07:08:17 +08:00
|
|
|
seqlock_t hh_lock;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* cached hardware header; allow for machine alignment needs. */
|
|
|
|
#define HH_DATA_MOD 16
|
|
|
|
#define HH_DATA_OFF(__len) \
|
2005-06-03 07:48:05 +08:00
|
|
|
(HH_DATA_MOD - (((__len - 1) & (HH_DATA_MOD - 1)) + 1))
|
2005-04-17 06:20:36 +08:00
|
|
|
#define HH_DATA_ALIGN(__len) \
|
|
|
|
(((__len)+(HH_DATA_MOD-1))&~(HH_DATA_MOD - 1))
|
|
|
|
unsigned long hh_data[HH_DATA_ALIGN(LL_MAX_HEADER) / sizeof(long)];
|
|
|
|
};
|
|
|
|
|
|
|
|
/* Reserve HH_DATA_MOD byte aligned hard_header_len, but at least that much.
|
|
|
|
* Alternative is:
|
|
|
|
* dev->hard_header_len ? (dev->hard_header_len +
|
|
|
|
* (HH_DATA_MOD - 1)) & ~(HH_DATA_MOD - 1) : 0
|
|
|
|
*
|
|
|
|
* We could use other alignment values, but we must maintain the
|
|
|
|
* relationship HH alignment <= LL alignment.
|
|
|
|
*/
|
|
|
|
#define LL_RESERVED_SPACE(dev) \
|
2008-05-13 11:48:31 +08:00
|
|
|
((((dev)->hard_header_len+(dev)->needed_headroom)&~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
|
2005-04-17 06:20:36 +08:00
|
|
|
#define LL_RESERVED_SPACE_EXTRA(dev,extra) \
|
2008-05-13 11:48:31 +08:00
|
|
|
((((dev)->hard_header_len+(dev)->needed_headroom+(extra))&~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-10-09 16:40:57 +08:00
|
|
|
struct header_ops {
|
|
|
|
int (*create) (struct sk_buff *skb, struct net_device *dev,
|
|
|
|
unsigned short type, const void *daddr,
|
2012-04-15 13:58:06 +08:00
|
|
|
const void *saddr, unsigned int len);
|
2007-10-09 16:40:57 +08:00
|
|
|
int (*parse)(const struct sk_buff *skb, unsigned char *haddr);
|
2011-07-13 14:28:12 +08:00
|
|
|
int (*cache)(const struct neighbour *neigh, struct hh_cache *hh, __be16 type);
|
2007-10-09 16:40:57 +08:00
|
|
|
void (*cache_update)(struct hh_cache *hh,
|
|
|
|
const struct net_device *dev,
|
|
|
|
const unsigned char *haddr);
|
|
|
|
};
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* These flag bits are private to the generic network queueing
|
|
|
|
* layer, they may not be explicitly referenced by any other
|
|
|
|
* code.
|
|
|
|
*/
|
|
|
|
|
2009-11-05 01:50:58 +08:00
|
|
|
enum netdev_state_t {
|
2005-04-17 06:20:36 +08:00
|
|
|
__LINK_STATE_START,
|
|
|
|
__LINK_STATE_PRESENT,
|
|
|
|
__LINK_STATE_NOCARRIER,
|
2006-03-21 09:09:11 +08:00
|
|
|
__LINK_STATE_LINKWATCH_PENDING,
|
|
|
|
__LINK_STATE_DORMANT,
|
2005-04-17 06:20:36 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This structure holds at boot time configured netdevice settings. They
|
2009-02-06 13:26:19 +08:00
|
|
|
* are then used in the device probing.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
|
|
|
struct netdev_boot_setup {
|
|
|
|
char name[IFNAMSIZ];
|
|
|
|
struct ifmap map;
|
|
|
|
};
|
|
|
|
#define NETDEV_BOOT_SETUP_MAX 8
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int __init netdev_boot_setup(char *str);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/*
|
|
|
|
* Structure for NAPI scheduling similar to tasklet but with weighting
|
|
|
|
*/
|
|
|
|
struct napi_struct {
|
|
|
|
/* The poll_list must only be managed by the entity which
|
|
|
|
* changes the state of the NAPI_STATE_SCHED bit. This means
|
|
|
|
* whoever atomically sets that bit can add this napi_struct
|
|
|
|
* to the per-cpu poll_list, and whoever clears that bit
|
|
|
|
* can remove from the list right before clearing the bit.
|
|
|
|
*/
|
|
|
|
struct list_head poll_list;
|
|
|
|
|
|
|
|
unsigned long state;
|
|
|
|
int weight;
|
2012-09-26 15:07:47 +08:00
|
|
|
unsigned int gro_count;
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
int (*poll)(struct napi_struct *, int);
|
|
|
|
#ifdef CONFIG_NETPOLL
|
|
|
|
spinlock_t poll_lock;
|
|
|
|
int poll_owner;
|
|
|
|
#endif
|
2009-01-05 08:13:40 +08:00
|
|
|
struct net_device *dev;
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
struct sk_buff *gro_list;
|
2009-01-05 08:13:40 +08:00
|
|
|
struct sk_buff *skb;
|
net: gro: add a per device gro flush timer
Tuning coalescing parameters on NIC can be really hard.
Servers can handle both bulk and RPC like traffic, with conflicting
goals : bulk flows want as big GRO packets as possible, RPC want minimal
latencies.
To reach big GRO packets on 10Gbe NIC, one can use :
ethtool -C eth0 rx-usecs 4 rx-frames 44
But this penalizes rpc sessions, with an increase of latencies, up to
50% in some cases, as NICs generally do not force an interrupt when
a packet with TCP Push flag is received.
Some NICs do not have an absolute timer, only a timer rearmed for every
incoming packet.
This patch uses a different strategy : Let GRO stack decides what do do,
based on traffic pattern.
Packets with Push flag wont be delayed.
Packets without Push flag might be held in GRO engine, if we keep
receiving data.
This new mechanism is off by default, and shall be enabled by setting
/sys/class/net/ethX/gro_flush_timeout to a value in nanosecond.
To fully enable this mechanism, drivers should use napi_complete_done()
instead of napi_complete().
Tested:
Ran 200 netperf TCP_STREAM from A to B (10Gbe mlx4 link, 8 RX queues)
Without this feature, we send back about 305,000 ACK per second.
GRO aggregation ratio is low (811/305 = 2.65 segments per GRO packet)
Setting a timer of 2000 nsec is enough to increase GRO packet sizes
and reduce number of ACK packets. (811/19.2 = 42)
Receiver performs less calls to upper stacks, less wakes up.
This also reduces cpu usage on the sender, as it receives less ACK
packets.
Note that reducing number of wakes up increases cpu efficiency, but can
decrease QPS, as applications wont have the chance to warmup cpu caches
doing a partial read of RPC requests/answers if they fit in one skb.
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811269.80 305732.30 1199462.57 19705.72 0.00
0.00 0.50
B:~# echo 2000 >/sys/class/net/eth0/gro_flush_timeout
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811577.30 19230.80 1199916.51 1239.80 0.00
0.00 0.50
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-07 13:09:44 +08:00
|
|
|
struct hrtimer timer;
|
2012-09-26 15:07:47 +08:00
|
|
|
struct list_head dev_list;
|
2013-06-10 16:39:41 +08:00
|
|
|
struct hlist_node napi_hash_node;
|
|
|
|
unsigned int napi_id;
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
};
|
|
|
|
|
2009-11-05 01:50:58 +08:00
|
|
|
enum {
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
NAPI_STATE_SCHED, /* Poll is scheduled */
|
2008-01-08 12:35:07 +08:00
|
|
|
NAPI_STATE_DISABLE, /* Disable pending */
|
netpoll: fix race on poll_list resulting in garbage entry
A few months back a race was discused between the netpoll napi service
path, and the fast path through net_rx_action:
http://kerneltrap.org/mailarchive/linux-netdev/2007/10/16/345470
A patch was submitted for that bug, but I think we missed a case.
Consider the following scenario:
INITIAL STATE
CPU0 has one napi_struct A on its poll_list
CPU1 is calling netpoll_send_skb and needs to call poll_napi on the same
napi_struct A that CPU0 has on its list
CPU0 CPU1
net_rx_action poll_napi
!list_empty (returns true) locks poll_lock for A
poll_one_napi
napi->poll
netif_rx_complete
__napi_complete
(removes A from poll_list)
list_entry(list->next)
In the above scenario, net_rx_action assumes that the per-cpu poll_list is
exclusive to that cpu. netpoll of course violates that, and because the netpoll
path can dequeue from the poll list, its possible for CPU0 to detect a non-empty
list at the top of the while loop in net_rx_action, but have it become empty by
the time it calls list_entry. Since the poll_list isn't surrounded by any other
structure, the returned data from that list_entry call in this situation is
garbage, and any number of crashes can result based on what exactly that garbage
is.
Given that its not fasible for performance reasons to place exclusive locks
arround each cpus poll list to provide that mutal exclusion, I think the best
solution is modify the netpoll path in such a way that we continue to guarantee
that the poll_list for a cpu is in fact exclusive to that cpu. To do this I've
implemented the patch below. It adds an additional bit to the state field in
the napi_struct. When executing napi->poll from the netpoll_path, this bit will
be set. When a driver calls netif_rx_complete, if that bit is set, it will not
remove the napi_struct from the poll_list. That work will be saved for the next
iteration of net_rx_action.
I've tested this and it seems to work well. About the biggest drawback I can
see to it is the fact that it might result in an extra loop through
net_rx_action in the event that the device is actually contended for (i.e. the
netpoll path actually preforms all the needed work no the device, and the call
to net_rx_action winds up doing nothing, except removing the napi_struct from
the poll_list. However I think this is probably a small price to pay, given
that the alternative is a crash.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-10 15:22:26 +08:00
|
|
|
NAPI_STATE_NPSVC, /* Netpoll - don't dequeue from poll_list */
|
2013-06-10 16:39:41 +08:00
|
|
|
NAPI_STATE_HASHED, /* In NAPI hash */
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
};
|
|
|
|
|
2009-10-29 15:17:09 +08:00
|
|
|
enum gro_result {
|
2009-03-17 01:50:02 +08:00
|
|
|
GRO_MERGED,
|
|
|
|
GRO_MERGED_FREE,
|
|
|
|
GRO_HELD,
|
|
|
|
GRO_NORMAL,
|
|
|
|
GRO_DROP,
|
|
|
|
};
|
2009-10-29 15:17:09 +08:00
|
|
|
typedef enum gro_result gro_result_t;
|
2009-03-17 01:50:02 +08:00
|
|
|
|
2011-03-12 11:14:39 +08:00
|
|
|
/*
|
|
|
|
* enum rx_handler_result - Possible return values for rx_handlers.
|
|
|
|
* @RX_HANDLER_CONSUMED: skb was consumed by rx_handler, do not process it
|
|
|
|
* further.
|
|
|
|
* @RX_HANDLER_ANOTHER: Do another round in receive path. This is indicated in
|
|
|
|
* case skb->dev was changed by rx_handler.
|
|
|
|
* @RX_HANDLER_EXACT: Force exact delivery, no wildcard.
|
|
|
|
* @RX_HANDLER_PASS: Do nothing, passe the skb as if no rx_handler was called.
|
|
|
|
*
|
|
|
|
* rx_handlers are functions called from inside __netif_receive_skb(), to do
|
|
|
|
* special processing of the skb, prior to delivery to protocol handlers.
|
|
|
|
*
|
|
|
|
* Currently, a net_device can only have a single rx_handler registered. Trying
|
|
|
|
* to register a second rx_handler will return -EBUSY.
|
|
|
|
*
|
|
|
|
* To register a rx_handler on a net_device, use netdev_rx_handler_register().
|
|
|
|
* To unregister a rx_handler on a net_device, use
|
|
|
|
* netdev_rx_handler_unregister().
|
|
|
|
*
|
|
|
|
* Upon return, rx_handler is expected to tell __netif_receive_skb() what to
|
|
|
|
* do with the skb.
|
|
|
|
*
|
|
|
|
* If the rx_handler consumed to skb in some way, it should return
|
|
|
|
* RX_HANDLER_CONSUMED. This is appropriate when the rx_handler arranged for
|
|
|
|
* the skb to be delivered in some other ways.
|
|
|
|
*
|
|
|
|
* If the rx_handler changed skb->dev, to divert the skb to another
|
|
|
|
* net_device, it should return RX_HANDLER_ANOTHER. The rx_handler for the
|
|
|
|
* new device will be called if it exists.
|
|
|
|
*
|
|
|
|
* If the rx_handler consider the skb should be ignored, it should return
|
|
|
|
* RX_HANDLER_EXACT. The skb will only be delivered to protocol handlers that
|
2012-09-20 09:47:58 +08:00
|
|
|
* are registered on exact device (ptype->dev == skb->dev).
|
2011-03-12 11:14:39 +08:00
|
|
|
*
|
|
|
|
* If the rx_handler didn't changed skb->dev, but want the skb to be normally
|
|
|
|
* delivered, it should return RX_HANDLER_PASS.
|
|
|
|
*
|
|
|
|
* A device without a registered rx_handler will behave as if rx_handler
|
|
|
|
* returned RX_HANDLER_PASS.
|
|
|
|
*/
|
|
|
|
|
|
|
|
enum rx_handler_result {
|
|
|
|
RX_HANDLER_CONSUMED,
|
|
|
|
RX_HANDLER_ANOTHER,
|
|
|
|
RX_HANDLER_EXACT,
|
|
|
|
RX_HANDLER_PASS,
|
|
|
|
};
|
|
|
|
typedef enum rx_handler_result rx_handler_result_t;
|
|
|
|
typedef rx_handler_result_t rx_handler_func_t(struct sk_buff **pskb);
|
2010-06-02 05:52:08 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void __napi_schedule(struct napi_struct *n);
|
2014-10-29 09:05:13 +08:00
|
|
|
void __napi_schedule_irqoff(struct napi_struct *n);
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool napi_disable_pending(struct napi_struct *n)
|
2008-01-08 12:35:07 +08:00
|
|
|
{
|
|
|
|
return test_bit(NAPI_STATE_DISABLE, &n->state);
|
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* napi_schedule_prep - check if napi can be scheduled
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Test if NAPI routine is already running, and if not mark
|
|
|
|
* it as running. This is used as a condition variable
|
2008-01-08 12:35:07 +08:00
|
|
|
* insure only one NAPI poll instance runs. We also make
|
|
|
|
* sure there is no pending NAPI disable.
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool napi_schedule_prep(struct napi_struct *n)
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
{
|
2008-01-08 12:35:07 +08:00
|
|
|
return !napi_disable_pending(n) &&
|
|
|
|
!test_and_set_bit(NAPI_STATE_SCHED, &n->state);
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* napi_schedule - schedule NAPI poll
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Schedule NAPI poll routine to be called if it is not already
|
|
|
|
* running.
|
|
|
|
*/
|
|
|
|
static inline void napi_schedule(struct napi_struct *n)
|
|
|
|
{
|
|
|
|
if (napi_schedule_prep(n))
|
|
|
|
__napi_schedule(n);
|
|
|
|
}
|
|
|
|
|
2014-10-29 09:05:13 +08:00
|
|
|
/**
|
|
|
|
* napi_schedule_irqoff - schedule NAPI poll
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Variant of napi_schedule(), assuming hard irqs are masked.
|
|
|
|
*/
|
|
|
|
static inline void napi_schedule_irqoff(struct napi_struct *n)
|
|
|
|
{
|
|
|
|
if (napi_schedule_prep(n))
|
|
|
|
__napi_schedule_irqoff(n);
|
|
|
|
}
|
|
|
|
|
2007-10-10 06:47:37 +08:00
|
|
|
/* Try to reschedule poll. Called by dev->poll() after napi_complete(). */
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool napi_reschedule(struct napi_struct *napi)
|
2007-10-10 06:47:37 +08:00
|
|
|
{
|
|
|
|
if (napi_schedule_prep(napi)) {
|
|
|
|
__napi_schedule(napi);
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
return true;
|
2007-10-10 06:47:37 +08:00
|
|
|
}
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
return false;
|
2007-10-10 06:47:37 +08:00
|
|
|
}
|
|
|
|
|
net: gro: add a per device gro flush timer
Tuning coalescing parameters on NIC can be really hard.
Servers can handle both bulk and RPC like traffic, with conflicting
goals : bulk flows want as big GRO packets as possible, RPC want minimal
latencies.
To reach big GRO packets on 10Gbe NIC, one can use :
ethtool -C eth0 rx-usecs 4 rx-frames 44
But this penalizes rpc sessions, with an increase of latencies, up to
50% in some cases, as NICs generally do not force an interrupt when
a packet with TCP Push flag is received.
Some NICs do not have an absolute timer, only a timer rearmed for every
incoming packet.
This patch uses a different strategy : Let GRO stack decides what do do,
based on traffic pattern.
Packets with Push flag wont be delayed.
Packets without Push flag might be held in GRO engine, if we keep
receiving data.
This new mechanism is off by default, and shall be enabled by setting
/sys/class/net/ethX/gro_flush_timeout to a value in nanosecond.
To fully enable this mechanism, drivers should use napi_complete_done()
instead of napi_complete().
Tested:
Ran 200 netperf TCP_STREAM from A to B (10Gbe mlx4 link, 8 RX queues)
Without this feature, we send back about 305,000 ACK per second.
GRO aggregation ratio is low (811/305 = 2.65 segments per GRO packet)
Setting a timer of 2000 nsec is enough to increase GRO packet sizes
and reduce number of ACK packets. (811/19.2 = 42)
Receiver performs less calls to upper stacks, less wakes up.
This also reduces cpu usage on the sender, as it receives less ACK
packets.
Note that reducing number of wakes up increases cpu efficiency, but can
decrease QPS, as applications wont have the chance to warmup cpu caches
doing a partial read of RPC requests/answers if they fit in one skb.
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811269.80 305732.30 1199462.57 19705.72 0.00
0.00 0.50
B:~# echo 2000 >/sys/class/net/eth0/gro_flush_timeout
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811577.30 19230.80 1199916.51 1239.80 0.00
0.00 0.50
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-07 13:09:44 +08:00
|
|
|
void __napi_complete(struct napi_struct *n);
|
|
|
|
void napi_complete_done(struct napi_struct *n, int work_done);
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* napi_complete - NAPI processing complete
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Mark NAPI processing as complete.
|
net: gro: add a per device gro flush timer
Tuning coalescing parameters on NIC can be really hard.
Servers can handle both bulk and RPC like traffic, with conflicting
goals : bulk flows want as big GRO packets as possible, RPC want minimal
latencies.
To reach big GRO packets on 10Gbe NIC, one can use :
ethtool -C eth0 rx-usecs 4 rx-frames 44
But this penalizes rpc sessions, with an increase of latencies, up to
50% in some cases, as NICs generally do not force an interrupt when
a packet with TCP Push flag is received.
Some NICs do not have an absolute timer, only a timer rearmed for every
incoming packet.
This patch uses a different strategy : Let GRO stack decides what do do,
based on traffic pattern.
Packets with Push flag wont be delayed.
Packets without Push flag might be held in GRO engine, if we keep
receiving data.
This new mechanism is off by default, and shall be enabled by setting
/sys/class/net/ethX/gro_flush_timeout to a value in nanosecond.
To fully enable this mechanism, drivers should use napi_complete_done()
instead of napi_complete().
Tested:
Ran 200 netperf TCP_STREAM from A to B (10Gbe mlx4 link, 8 RX queues)
Without this feature, we send back about 305,000 ACK per second.
GRO aggregation ratio is low (811/305 = 2.65 segments per GRO packet)
Setting a timer of 2000 nsec is enough to increase GRO packet sizes
and reduce number of ACK packets. (811/19.2 = 42)
Receiver performs less calls to upper stacks, less wakes up.
This also reduces cpu usage on the sender, as it receives less ACK
packets.
Note that reducing number of wakes up increases cpu efficiency, but can
decrease QPS, as applications wont have the chance to warmup cpu caches
doing a partial read of RPC requests/answers if they fit in one skb.
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811269.80 305732.30 1199462.57 19705.72 0.00
0.00 0.50
B:~# echo 2000 >/sys/class/net/eth0/gro_flush_timeout
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811577.30 19230.80 1199916.51 1239.80 0.00
0.00 0.50
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-07 13:09:44 +08:00
|
|
|
* Consider using napi_complete_done() instead.
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
*/
|
net: gro: add a per device gro flush timer
Tuning coalescing parameters on NIC can be really hard.
Servers can handle both bulk and RPC like traffic, with conflicting
goals : bulk flows want as big GRO packets as possible, RPC want minimal
latencies.
To reach big GRO packets on 10Gbe NIC, one can use :
ethtool -C eth0 rx-usecs 4 rx-frames 44
But this penalizes rpc sessions, with an increase of latencies, up to
50% in some cases, as NICs generally do not force an interrupt when
a packet with TCP Push flag is received.
Some NICs do not have an absolute timer, only a timer rearmed for every
incoming packet.
This patch uses a different strategy : Let GRO stack decides what do do,
based on traffic pattern.
Packets with Push flag wont be delayed.
Packets without Push flag might be held in GRO engine, if we keep
receiving data.
This new mechanism is off by default, and shall be enabled by setting
/sys/class/net/ethX/gro_flush_timeout to a value in nanosecond.
To fully enable this mechanism, drivers should use napi_complete_done()
instead of napi_complete().
Tested:
Ran 200 netperf TCP_STREAM from A to B (10Gbe mlx4 link, 8 RX queues)
Without this feature, we send back about 305,000 ACK per second.
GRO aggregation ratio is low (811/305 = 2.65 segments per GRO packet)
Setting a timer of 2000 nsec is enough to increase GRO packet sizes
and reduce number of ACK packets. (811/19.2 = 42)
Receiver performs less calls to upper stacks, less wakes up.
This also reduces cpu usage on the sender, as it receives less ACK
packets.
Note that reducing number of wakes up increases cpu efficiency, but can
decrease QPS, as applications wont have the chance to warmup cpu caches
doing a partial read of RPC requests/answers if they fit in one skb.
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811269.80 305732.30 1199462.57 19705.72 0.00
0.00 0.50
B:~# echo 2000 >/sys/class/net/eth0/gro_flush_timeout
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811577.30 19230.80 1199916.51 1239.80 0.00
0.00 0.50
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-07 13:09:44 +08:00
|
|
|
static inline void napi_complete(struct napi_struct *n)
|
|
|
|
{
|
|
|
|
return napi_complete_done(n, 0);
|
|
|
|
}
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
|
2013-06-10 16:39:41 +08:00
|
|
|
/**
|
|
|
|
* napi_by_id - lookup a NAPI by napi_id
|
|
|
|
* @napi_id: hashed napi_id
|
|
|
|
*
|
|
|
|
* lookup @napi_id in napi_hash table
|
|
|
|
* must be called under rcu_read_lock()
|
|
|
|
*/
|
2013-09-27 05:48:15 +08:00
|
|
|
struct napi_struct *napi_by_id(unsigned int napi_id);
|
2013-06-10 16:39:41 +08:00
|
|
|
|
|
|
|
/**
|
|
|
|
* napi_hash_add - add a NAPI to global hashtable
|
|
|
|
* @napi: napi context
|
|
|
|
*
|
|
|
|
* generate a new napi_id and store a @napi under it in napi_hash
|
|
|
|
*/
|
2013-09-27 05:48:15 +08:00
|
|
|
void napi_hash_add(struct napi_struct *napi);
|
2013-06-10 16:39:41 +08:00
|
|
|
|
|
|
|
/**
|
|
|
|
* napi_hash_del - remove a NAPI from global table
|
|
|
|
* @napi: napi context
|
|
|
|
*
|
|
|
|
* Warning: caller must observe rcu grace period
|
|
|
|
* before freeing memory containing @napi
|
|
|
|
*/
|
2013-09-27 05:48:15 +08:00
|
|
|
void napi_hash_del(struct napi_struct *napi);
|
2013-06-10 16:39:41 +08:00
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* napi_disable - prevent NAPI from scheduling
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Stop NAPI from being scheduled on this context.
|
|
|
|
* Waits till any outstanding processing completes.
|
|
|
|
*/
|
net: gro: add a per device gro flush timer
Tuning coalescing parameters on NIC can be really hard.
Servers can handle both bulk and RPC like traffic, with conflicting
goals : bulk flows want as big GRO packets as possible, RPC want minimal
latencies.
To reach big GRO packets on 10Gbe NIC, one can use :
ethtool -C eth0 rx-usecs 4 rx-frames 44
But this penalizes rpc sessions, with an increase of latencies, up to
50% in some cases, as NICs generally do not force an interrupt when
a packet with TCP Push flag is received.
Some NICs do not have an absolute timer, only a timer rearmed for every
incoming packet.
This patch uses a different strategy : Let GRO stack decides what do do,
based on traffic pattern.
Packets with Push flag wont be delayed.
Packets without Push flag might be held in GRO engine, if we keep
receiving data.
This new mechanism is off by default, and shall be enabled by setting
/sys/class/net/ethX/gro_flush_timeout to a value in nanosecond.
To fully enable this mechanism, drivers should use napi_complete_done()
instead of napi_complete().
Tested:
Ran 200 netperf TCP_STREAM from A to B (10Gbe mlx4 link, 8 RX queues)
Without this feature, we send back about 305,000 ACK per second.
GRO aggregation ratio is low (811/305 = 2.65 segments per GRO packet)
Setting a timer of 2000 nsec is enough to increase GRO packet sizes
and reduce number of ACK packets. (811/19.2 = 42)
Receiver performs less calls to upper stacks, less wakes up.
This also reduces cpu usage on the sender, as it receives less ACK
packets.
Note that reducing number of wakes up increases cpu efficiency, but can
decrease QPS, as applications wont have the chance to warmup cpu caches
doing a partial read of RPC requests/answers if they fit in one skb.
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811269.80 305732.30 1199462.57 19705.72 0.00
0.00 0.50
B:~# echo 2000 >/sys/class/net/eth0/gro_flush_timeout
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811577.30 19230.80 1199916.51 1239.80 0.00
0.00 0.50
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-07 13:09:44 +08:00
|
|
|
void napi_disable(struct napi_struct *n);
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
|
|
|
|
/**
|
|
|
|
* napi_enable - enable NAPI scheduling
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Resume NAPI from being scheduled on this context.
|
|
|
|
* Must be paired with napi_disable.
|
|
|
|
*/
|
|
|
|
static inline void napi_enable(struct napi_struct *n)
|
|
|
|
{
|
|
|
|
BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
|
2014-03-18 01:06:10 +08:00
|
|
|
smp_mb__before_atomic();
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
clear_bit(NAPI_STATE_SCHED, &n->state);
|
|
|
|
}
|
|
|
|
|
2007-10-18 04:26:41 +08:00
|
|
|
#ifdef CONFIG_SMP
|
|
|
|
/**
|
|
|
|
* napi_synchronize - wait until NAPI is not running
|
|
|
|
* @n: napi context
|
|
|
|
*
|
|
|
|
* Wait until NAPI is done being scheduled on this context.
|
|
|
|
* Waits till any outstanding processing completes but
|
|
|
|
* does not disable future activations.
|
|
|
|
*/
|
|
|
|
static inline void napi_synchronize(const struct napi_struct *n)
|
|
|
|
{
|
|
|
|
while (test_bit(NAPI_STATE_SCHED, &n->state))
|
|
|
|
msleep(1);
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
# define napi_synchronize(n) barrier()
|
|
|
|
#endif
|
|
|
|
|
2009-11-05 01:50:58 +08:00
|
|
|
enum netdev_queue_state_t {
|
2011-11-29 00:32:44 +08:00
|
|
|
__QUEUE_STATE_DRV_XOFF,
|
|
|
|
__QUEUE_STATE_STACK_XOFF,
|
2008-08-01 07:58:50 +08:00
|
|
|
__QUEUE_STATE_FROZEN,
|
2008-07-09 14:14:46 +08:00
|
|
|
};
|
2014-04-03 02:52:57 +08:00
|
|
|
|
|
|
|
#define QUEUE_STATE_DRV_XOFF (1 << __QUEUE_STATE_DRV_XOFF)
|
|
|
|
#define QUEUE_STATE_STACK_XOFF (1 << __QUEUE_STATE_STACK_XOFF)
|
|
|
|
#define QUEUE_STATE_FROZEN (1 << __QUEUE_STATE_FROZEN)
|
|
|
|
|
|
|
|
#define QUEUE_STATE_ANY_XOFF (QUEUE_STATE_DRV_XOFF | QUEUE_STATE_STACK_XOFF)
|
|
|
|
#define QUEUE_STATE_ANY_XOFF_OR_FROZEN (QUEUE_STATE_ANY_XOFF | \
|
|
|
|
QUEUE_STATE_FROZEN)
|
|
|
|
#define QUEUE_STATE_DRV_XOFF_OR_FROZEN (QUEUE_STATE_DRV_XOFF | \
|
|
|
|
QUEUE_STATE_FROZEN)
|
|
|
|
|
2011-11-29 00:32:44 +08:00
|
|
|
/*
|
|
|
|
* __QUEUE_STATE_DRV_XOFF is used by drivers to stop the transmit queue. The
|
|
|
|
* netif_tx_* functions below are used to manipulate this flag. The
|
|
|
|
* __QUEUE_STATE_STACK_XOFF flag is used by the stack to stop the transmit
|
|
|
|
* queue independently. The netif_xmit_*stopped functions below are called
|
|
|
|
* to check if the queue has been stopped by the driver or stack (either
|
|
|
|
* of the XOFF bits are set in the state). Drivers should not need to call
|
|
|
|
* netif_xmit*stopped functions, they should only be using netif_tx_*.
|
|
|
|
*/
|
2008-07-09 14:14:46 +08:00
|
|
|
|
2008-07-09 07:55:56 +08:00
|
|
|
struct netdev_queue {
|
2009-04-28 19:43:42 +08:00
|
|
|
/*
|
|
|
|
* read mostly part
|
|
|
|
*/
|
2008-07-09 07:55:56 +08:00
|
|
|
struct net_device *dev;
|
2014-09-13 11:04:52 +08:00
|
|
|
struct Qdisc __rcu *qdisc;
|
2008-07-09 08:42:10 +08:00
|
|
|
struct Qdisc *qdisc_sleeping;
|
2011-11-16 20:15:10 +08:00
|
|
|
#ifdef CONFIG_SYSFS
|
2010-11-21 21:17:27 +08:00
|
|
|
struct kobject kobj;
|
|
|
|
#endif
|
2010-11-29 16:14:37 +08:00
|
|
|
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
|
|
|
|
int numa_node;
|
|
|
|
#endif
|
2009-04-28 19:43:42 +08:00
|
|
|
/*
|
|
|
|
* write mostly part
|
|
|
|
*/
|
|
|
|
spinlock_t _xmit_lock ____cacheline_aligned_in_smp;
|
|
|
|
int xmit_lock_owner;
|
2009-05-18 11:55:16 +08:00
|
|
|
/*
|
|
|
|
* please use this field instead of dev->trans_start
|
|
|
|
*/
|
|
|
|
unsigned long trans_start;
|
2011-11-16 20:15:10 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Number of TX timeouts for this queue
|
|
|
|
* (/sys/class/net/DEV/Q/trans_timeout)
|
|
|
|
*/
|
|
|
|
unsigned long trans_timeout;
|
2011-11-29 00:33:09 +08:00
|
|
|
|
|
|
|
unsigned long state;
|
|
|
|
|
|
|
|
#ifdef CONFIG_BQL
|
|
|
|
struct dql dql;
|
|
|
|
#endif
|
2015-03-18 20:57:33 +08:00
|
|
|
unsigned long tx_maxrate;
|
2008-07-17 15:34:19 +08:00
|
|
|
} ____cacheline_aligned_in_smp;
|
2008-07-09 07:55:56 +08:00
|
|
|
|
2010-11-29 16:14:37 +08:00
|
|
|
static inline int netdev_queue_numa_node_read(const struct netdev_queue *q)
|
|
|
|
{
|
|
|
|
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
|
|
|
|
return q->numa_node;
|
|
|
|
#else
|
2010-12-14 11:09:15 +08:00
|
|
|
return NUMA_NO_NODE;
|
2010-11-29 16:14:37 +08:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netdev_queue_numa_node_write(struct netdev_queue *q, int node)
|
|
|
|
{
|
|
|
|
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
|
|
|
|
q->numa_node = node;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2010-03-25 03:13:54 +08:00
|
|
|
#ifdef CONFIG_RPS
|
2010-03-16 16:03:29 +08:00
|
|
|
/*
|
|
|
|
* This structure holds an RPS map which can be of variable length. The
|
|
|
|
* map is an array of CPUs.
|
|
|
|
*/
|
|
|
|
struct rps_map {
|
|
|
|
unsigned int len;
|
|
|
|
struct rcu_head rcu;
|
|
|
|
u16 cpus[0];
|
|
|
|
};
|
2011-12-24 14:56:49 +08:00
|
|
|
#define RPS_MAP_SIZE(_num) (sizeof(struct rps_map) + ((_num) * sizeof(u16)))
|
2010-03-16 16:03:29 +08:00
|
|
|
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
/*
|
2011-01-19 19:03:53 +08:00
|
|
|
* The rps_dev_flow structure contains the mapping of a flow to a CPU, the
|
|
|
|
* tail pointer for that CPU's input queue at the time of last enqueue, and
|
|
|
|
* a hardware filter index.
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
*/
|
|
|
|
struct rps_dev_flow {
|
|
|
|
u16 cpu;
|
2011-01-19 19:03:53 +08:00
|
|
|
u16 filter;
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
unsigned int last_qtail;
|
|
|
|
};
|
2011-01-19 19:03:53 +08:00
|
|
|
#define RPS_NO_FILTER 0xffff
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The rps_dev_flow_table structure contains a table of flow mappings.
|
|
|
|
*/
|
|
|
|
struct rps_dev_flow_table {
|
|
|
|
unsigned int mask;
|
|
|
|
struct rcu_head rcu;
|
|
|
|
struct rps_dev_flow flows[0];
|
|
|
|
};
|
|
|
|
#define RPS_DEV_FLOW_TABLE_SIZE(_num) (sizeof(struct rps_dev_flow_table) + \
|
2011-12-24 14:56:49 +08:00
|
|
|
((_num) * sizeof(struct rps_dev_flow)))
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The rps_sock_flow_table contains mappings of flows to the last CPU
|
|
|
|
* on which they were processed by the application (set in recvmsg).
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
* Each entry is a 32bit value. Upper part is the high order bits
|
|
|
|
* of flow hash, lower part is cpu number.
|
|
|
|
* rps_cpu_mask is used to partition the space, depending on number of
|
|
|
|
* possible cpus : rps_cpu_mask = roundup_pow_of_two(nr_cpu_ids) - 1
|
|
|
|
* For example, if 64 cpus are possible, rps_cpu_mask = 0x3f,
|
|
|
|
* meaning we use 32-6=26 bits for the hash.
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
*/
|
|
|
|
struct rps_sock_flow_table {
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
u32 mask;
|
2015-02-09 12:39:13 +08:00
|
|
|
|
|
|
|
u32 ents[0] ____cacheline_aligned_in_smp;
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
};
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
#define RPS_SOCK_FLOW_TABLE_SIZE(_num) (offsetof(struct rps_sock_flow_table, ents[_num]))
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
|
|
|
|
#define RPS_NO_CPU 0xffff
|
|
|
|
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
extern u32 rps_cpu_mask;
|
|
|
|
extern struct rps_sock_flow_table __rcu *rps_sock_flow_table;
|
|
|
|
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
static inline void rps_record_sock_flow(struct rps_sock_flow_table *table,
|
|
|
|
u32 hash)
|
|
|
|
{
|
|
|
|
if (table && hash) {
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
unsigned int index = hash & table->mask;
|
|
|
|
u32 val = hash & ~rps_cpu_mask;
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
|
|
|
|
/* We only give a hint, preemption can change cpu under us */
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
val |= raw_smp_processor_id();
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
|
net: rfs: add hash collision detection
Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.
Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)
This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.
I use a 32bit value, and dynamically split it in two parts.
For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.
Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.
If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).
This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.
This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-07 04:59:01 +08:00
|
|
|
if (table->ents[index] != val)
|
|
|
|
table->ents[index] = val;
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-01-19 19:03:53 +08:00
|
|
|
#ifdef CONFIG_RFS_ACCEL
|
2013-09-27 05:48:15 +08:00
|
|
|
bool rps_may_expire_flow(struct net_device *dev, u16 rxq_index, u32 flow_id,
|
|
|
|
u16 filter_id);
|
2011-01-19 19:03:53 +08:00
|
|
|
#endif
|
2014-01-17 14:23:28 +08:00
|
|
|
#endif /* CONFIG_RPS */
|
2011-01-19 19:03:53 +08:00
|
|
|
|
2010-03-16 16:03:29 +08:00
|
|
|
/* This structure contains an instance of an RX queue. */
|
|
|
|
struct netdev_rx_queue {
|
2014-01-17 14:23:28 +08:00
|
|
|
#ifdef CONFIG_RPS
|
2010-10-25 11:02:02 +08:00
|
|
|
struct rps_map __rcu *rps_map;
|
|
|
|
struct rps_dev_flow_table __rcu *rps_flow_table;
|
2014-01-17 14:23:28 +08:00
|
|
|
#endif
|
2010-10-25 11:02:02 +08:00
|
|
|
struct kobject kobj;
|
2010-11-09 18:47:38 +08:00
|
|
|
struct net_device *dev;
|
2010-03-16 16:03:29 +08:00
|
|
|
} ____cacheline_aligned_in_smp;
|
2014-01-17 14:23:28 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* RX queue sysfs structures and functions.
|
|
|
|
*/
|
|
|
|
struct rx_queue_attribute {
|
|
|
|
struct attribute attr;
|
|
|
|
ssize_t (*show)(struct netdev_rx_queue *queue,
|
|
|
|
struct rx_queue_attribute *attr, char *buf);
|
|
|
|
ssize_t (*store)(struct netdev_rx_queue *queue,
|
|
|
|
struct rx_queue_attribute *attr, const char *buf, size_t len);
|
|
|
|
};
|
2008-11-20 13:32:24 +08:00
|
|
|
|
2010-11-26 16:36:09 +08:00
|
|
|
#ifdef CONFIG_XPS
|
|
|
|
/*
|
|
|
|
* This structure holds an XPS map which can be of variable length. The
|
|
|
|
* map is an array of queues.
|
|
|
|
*/
|
|
|
|
struct xps_map {
|
|
|
|
unsigned int len;
|
|
|
|
unsigned int alloc_len;
|
|
|
|
struct rcu_head rcu;
|
|
|
|
u16 queues[0];
|
|
|
|
};
|
2011-12-24 14:56:49 +08:00
|
|
|
#define XPS_MAP_SIZE(_num) (sizeof(struct xps_map) + ((_num) * sizeof(u16)))
|
2010-11-26 16:36:09 +08:00
|
|
|
#define XPS_MIN_MAP_ALLOC ((L1_CACHE_BYTES - sizeof(struct xps_map)) \
|
|
|
|
/ sizeof(u16))
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This structure holds all XPS maps for device. Maps are indexed by CPU.
|
|
|
|
*/
|
|
|
|
struct xps_dev_maps {
|
|
|
|
struct rcu_head rcu;
|
2010-11-29 05:43:02 +08:00
|
|
|
struct xps_map __rcu *cpu_map[0];
|
2010-11-26 16:36:09 +08:00
|
|
|
};
|
|
|
|
#define XPS_DEV_MAPS_SIZE (sizeof(struct xps_dev_maps) + \
|
|
|
|
(nr_cpu_ids * sizeof(struct xps_map *)))
|
|
|
|
#endif /* CONFIG_XPS */
|
|
|
|
|
2011-01-17 16:06:04 +08:00
|
|
|
#define TC_MAX_QUEUE 16
|
|
|
|
#define TC_BITMASK 15
|
|
|
|
/* HW offloaded queuing disciplines txq count and offset maps */
|
|
|
|
struct netdev_tc_txq {
|
|
|
|
u16 count;
|
|
|
|
u16 offset;
|
|
|
|
};
|
|
|
|
|
2012-01-05 04:23:39 +08:00
|
|
|
#if defined(CONFIG_FCOE) || defined(CONFIG_FCOE_MODULE)
|
|
|
|
/*
|
|
|
|
* This structure is to hold information about the device
|
|
|
|
* configured to run FCoE protocol stack.
|
|
|
|
*/
|
|
|
|
struct netdev_fcoe_hbainfo {
|
|
|
|
char manufacturer[64];
|
|
|
|
char serial_number[64];
|
|
|
|
char hardware_version[64];
|
|
|
|
char driver_version[64];
|
|
|
|
char optionrom_version[64];
|
|
|
|
char firmware_version[64];
|
|
|
|
char model[256];
|
|
|
|
char model_description[256];
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
|
2014-11-28 21:34:16 +08:00
|
|
|
#define MAX_PHYS_ITEM_ID_LEN 32
|
2013-07-30 00:16:49 +08:00
|
|
|
|
2014-11-28 21:34:16 +08:00
|
|
|
/* This structure holds a unique identifier to identify some
|
|
|
|
* physical item (port for example) used by a netdevice.
|
2013-07-30 00:16:49 +08:00
|
|
|
*/
|
2014-11-28 21:34:16 +08:00
|
|
|
struct netdev_phys_item_id {
|
|
|
|
unsigned char id[MAX_PHYS_ITEM_ID_LEN];
|
2013-07-30 00:16:49 +08:00
|
|
|
unsigned char id_len;
|
|
|
|
};
|
|
|
|
|
2015-07-19 09:24:49 +08:00
|
|
|
static inline bool netdev_phys_item_id_same(struct netdev_phys_item_id *a,
|
|
|
|
struct netdev_phys_item_id *b)
|
|
|
|
{
|
|
|
|
return a->id_len == b->id_len &&
|
|
|
|
memcmp(a->id, b->id, a->id_len) == 0;
|
|
|
|
}
|
|
|
|
|
2014-02-16 22:55:20 +08:00
|
|
|
typedef u16 (*select_queue_fallback_t)(struct net_device *dev,
|
|
|
|
struct sk_buff *skb);
|
|
|
|
|
2008-11-20 13:32:24 +08:00
|
|
|
/*
|
|
|
|
* This structure defines the management hooks for network devices.
|
2008-11-21 12:14:53 +08:00
|
|
|
* The following hooks can be defined; unless noted otherwise, they are
|
|
|
|
* optional and can be filled with a null pointer.
|
2008-11-20 13:32:24 +08:00
|
|
|
*
|
|
|
|
* int (*ndo_init)(struct net_device *dev);
|
|
|
|
* This function is called once when network device is registered.
|
|
|
|
* The network device can use this to any late stage initializaton
|
|
|
|
* or semantic validattion. It can fail with an error code which will
|
|
|
|
* be propogated back to register_netdev
|
|
|
|
*
|
|
|
|
* void (*ndo_uninit)(struct net_device *dev);
|
|
|
|
* This function is called when device is unregistered or when registration
|
|
|
|
* fails. It is not called if init fails.
|
|
|
|
*
|
|
|
|
* int (*ndo_open)(struct net_device *dev);
|
|
|
|
* This function is called when network device transistions to the up
|
|
|
|
* state.
|
|
|
|
*
|
|
|
|
* int (*ndo_stop)(struct net_device *dev);
|
|
|
|
* This function is called when network device transistions to the down
|
|
|
|
* state.
|
|
|
|
*
|
2009-09-01 03:50:40 +08:00
|
|
|
* netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb,
|
|
|
|
* struct net_device *dev);
|
2008-11-21 12:14:53 +08:00
|
|
|
* Called when a packet needs to be transmitted.
|
2015-04-03 19:47:17 +08:00
|
|
|
* Returns NETDEV_TX_OK. Can return NETDEV_TX_BUSY, but you should stop
|
|
|
|
* the queue before that can happen; it's for obsolete devices and weird
|
|
|
|
* corner cases, but the stack really does a non-trivial amount
|
|
|
|
* of useless work if you return NETDEV_TX_BUSY.
|
2009-09-01 03:50:40 +08:00
|
|
|
* (can also return NETDEV_TX_LOCKED iff NETIF_F_LLTX)
|
2008-11-21 12:14:53 +08:00
|
|
|
* Required can not be NULL.
|
|
|
|
*
|
2014-01-10 16:18:26 +08:00
|
|
|
* u16 (*ndo_select_queue)(struct net_device *dev, struct sk_buff *skb,
|
2014-02-16 22:55:20 +08:00
|
|
|
* void *accel_priv, select_queue_fallback_t fallback);
|
2008-11-21 12:14:53 +08:00
|
|
|
* Called to decide which queue to when device supports multiple
|
|
|
|
* transmit queues.
|
|
|
|
*
|
2008-11-20 13:32:24 +08:00
|
|
|
* void (*ndo_change_rx_flags)(struct net_device *dev, int flags);
|
|
|
|
* This function is called to allow device receiver to make
|
|
|
|
* changes to configuration when multicast or promiscious is enabled.
|
|
|
|
*
|
|
|
|
* void (*ndo_set_rx_mode)(struct net_device *dev);
|
|
|
|
* This function is called device changes address list filtering.
|
2011-08-16 14:29:00 +08:00
|
|
|
* If driver handles unicast address filtering, it should set
|
|
|
|
* IFF_UNICAST_FLT to its priv_flags.
|
2008-11-20 13:32:24 +08:00
|
|
|
*
|
|
|
|
* int (*ndo_set_mac_address)(struct net_device *dev, void *addr);
|
|
|
|
* This function is called when the Media Access Control address
|
2009-04-27 20:45:54 +08:00
|
|
|
* needs to be changed. If this interface is not defined, the
|
2008-11-20 13:32:24 +08:00
|
|
|
* mac address can not be changed.
|
|
|
|
*
|
|
|
|
* int (*ndo_validate_addr)(struct net_device *dev);
|
|
|
|
* Test if Media Access Control address is valid for the device.
|
|
|
|
*
|
|
|
|
* int (*ndo_do_ioctl)(struct net_device *dev, struct ifreq *ifr, int cmd);
|
|
|
|
* Called when a user request an ioctl which can't be handled by
|
|
|
|
* the generic interface code. If not defined ioctl's return
|
|
|
|
* not supported error code.
|
|
|
|
*
|
|
|
|
* int (*ndo_set_config)(struct net_device *dev, struct ifmap *map);
|
|
|
|
* Used to set network devices bus interface parameters. This interface
|
|
|
|
* is retained for legacy reason, new devices should use the bus
|
|
|
|
* interface (PCI) for low level management.
|
|
|
|
*
|
|
|
|
* int (*ndo_change_mtu)(struct net_device *dev, int new_mtu);
|
|
|
|
* Called when a user wants to change the Maximum Transfer Unit
|
|
|
|
* of a device. If not defined, any request to change MTU will
|
|
|
|
* will return an error.
|
|
|
|
*
|
2008-11-21 12:14:53 +08:00
|
|
|
* void (*ndo_tx_timeout)(struct net_device *dev);
|
2008-11-20 13:32:24 +08:00
|
|
|
* Callback uses when the transmitter has not made any progress
|
|
|
|
* for dev->watchdog ticks.
|
|
|
|
*
|
2010-07-09 17:11:52 +08:00
|
|
|
* struct rtnl_link_stats64* (*ndo_get_stats64)(struct net_device *dev,
|
2010-07-08 05:58:56 +08:00
|
|
|
* struct rtnl_link_stats64 *storage);
|
2009-10-08 04:53:11 +08:00
|
|
|
* struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);
|
2008-11-20 13:32:24 +08:00
|
|
|
* Called when a user wants to get the network device usage
|
2010-06-08 15:19:54 +08:00
|
|
|
* statistics. Drivers must do one of the following:
|
2010-07-09 17:11:52 +08:00
|
|
|
* 1. Define @ndo_get_stats64 to fill in a zero-initialised
|
|
|
|
* rtnl_link_stats64 structure passed by the caller.
|
2010-06-16 06:08:48 +08:00
|
|
|
* 2. Define @ndo_get_stats to update a net_device_stats structure
|
2010-06-08 15:19:54 +08:00
|
|
|
* (which should normally be dev->stats) and return a pointer to
|
|
|
|
* it. The structure may be changed asynchronously only if each
|
|
|
|
* field is written atomically.
|
|
|
|
* 3. Update dev->stats asynchronously and atomically, and define
|
|
|
|
* neither operation.
|
2008-11-20 13:32:24 +08:00
|
|
|
*
|
2015-01-12 17:16:25 +08:00
|
|
|
* int (*ndo_vlan_rx_add_vid)(struct net_device *dev, __be16 proto, u16 vid);
|
2013-04-19 10:04:28 +08:00
|
|
|
* If device support VLAN filtering this function is called when a
|
|
|
|
* VLAN id is registered.
|
2008-11-20 13:32:24 +08:00
|
|
|
*
|
2015-01-12 17:16:25 +08:00
|
|
|
* int (*ndo_vlan_rx_kill_vid)(struct net_device *dev, __be16 proto, u16 vid);
|
2013-04-19 10:04:28 +08:00
|
|
|
* If device support VLAN filtering this function is called when a
|
|
|
|
* VLAN id is unregistered.
|
2008-11-20 13:32:24 +08:00
|
|
|
*
|
|
|
|
* void (*ndo_poll_controller)(struct net_device *dev);
|
2010-02-10 09:43:46 +08:00
|
|
|
*
|
|
|
|
* SR-IOV management functions.
|
|
|
|
* int (*ndo_set_vf_mac)(struct net_device *dev, int vf, u8* mac);
|
|
|
|
* int (*ndo_set_vf_vlan)(struct net_device *dev, int vf, u16 vlan, u8 qos);
|
net-next:v4: Add support to configure SR-IOV VF minimum and maximum Tx rate through ip tool.
o min_tx_rate puts lower limit on the VF bandwidth. VF is guaranteed
to have a bandwidth of at least this value.
max_tx_rate puts cap on the VF bandwidth. VF can have a bandwidth
of up to this value.
o A new handler set_vf_rate for attr IFLA_VF_RATE has been introduced
which takes 4 arguments:
netdev, VF number, min_tx_rate, max_tx_rate
o ndo_set_vf_rate replaces ndo_set_vf_tx_rate handler.
o Drivers that currently implement ndo_set_vf_tx_rate should now call
ndo_set_vf_rate instead and reject attempt to set a minimum bandwidth
greater than 0 for IFLA_VF_TX_RATE when IFLA_VF_RATE is not yet
implemented by driver.
o If user enters only one of either min_tx_rate or max_tx_rate, then,
userland should read back the other value from driver and set both
for IFLA_VF_RATE.
Drivers that have not yet implemented IFLA_VF_RATE should always
return min_tx_rate as 0 when read from ip tool.
o If both IFLA_VF_TX_RATE and IFLA_VF_RATE options are specified, then
IFLA_VF_RATE should override.
o Idea is to have consistent display of rate values to user.
o Usage example: -
./ip link set p4p1 vf 0 rate 900
./ip link show p4p1
32: p4p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
link/ether 00:0e:1e:08:b0:f0 brd ff:ff:ff:ff:ff:ff
vf 0 MAC 3e:a0:ca:bd:ae:5a, tx rate 900 (Mbps), max_tx_rate 900Mbps
vf 1 MAC f6:c6:7c:3f:3d:6c
vf 2 MAC 56:32:43:98:d7:71
vf 3 MAC d6:be:c3:b5:85:ff
vf 4 MAC ee:a9:9a:1e:19:14
vf 5 MAC 4a:d0:4c:07:52:18
vf 6 MAC 3a:76:44:93:62:f9
vf 7 MAC 82:e9:e7:e3:15:1a
./ip link set p4p1 vf 0 max_tx_rate 300 min_tx_rate 200
./ip link show p4p1
32: p4p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
link/ether 00:0e:1e:08:b0:f0 brd ff:ff:ff:ff:ff:ff
vf 0 MAC 3e:a0:ca:bd:ae:5a, tx rate 300 (Mbps), max_tx_rate 300Mbps,
min_tx_rate 200Mbps
vf 1 MAC f6:c6:7c:3f:3d:6c
vf 2 MAC 56:32:43:98:d7:71
vf 3 MAC d6:be:c3:b5:85:ff
vf 4 MAC ee:a9:9a:1e:19:14
vf 5 MAC 4a:d0:4c:07:52:18
vf 6 MAC 3a:76:44:93:62:f9
vf 7 MAC 82:e9:e7:e3:15:1a
./ip link set p4p1 vf 0 max_tx_rate 600 rate 300
./ip link show p4p1
32: p4p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
link/ether 00:0e:1e:08:b0:f brd ff:ff:ff:ff:ff:ff
vf 0 MAC 3e:a0:ca:bd:ae:5, tx rate 600 (Mbps), max_tx_rate 600Mbps,
min_tx_rate 200Mbps
vf 1 MAC f6:c6:7c:3f:3d:6c
vf 2 MAC 56:32:43:98:d7:71
vf 3 MAC d6:be:c3:b5:85:ff
vf 4 MAC ee:a9:9a:1e:19:14
vf 5 MAC 4a:d0:4c:07:52:18
vf 6 MAC 3a:76:44:93:62:f9
vf 7 MAC 82:e9:e7:e3:15:1a
Signed-off-by: Sucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-05-22 21:59:05 +08:00
|
|
|
* int (*ndo_set_vf_rate)(struct net_device *dev, int vf, int min_tx_rate,
|
|
|
|
* int max_tx_rate);
|
2011-10-08 11:05:24 +08:00
|
|
|
* int (*ndo_set_vf_spoofchk)(struct net_device *dev, int vf, bool setting);
|
2010-02-10 09:43:46 +08:00
|
|
|
* int (*ndo_get_vf_config)(struct net_device *dev,
|
|
|
|
* int vf, struct ifla_vf_info *ivf);
|
2013-06-13 18:19:10 +08:00
|
|
|
* int (*ndo_set_vf_link_state)(struct net_device *dev, int vf, int link_state);
|
net: Add netlink support for virtual port management (was iovnl)
Add new netdev ops ndo_{set|get}_vf_port to allow setting of
port-profile on a netdev interface. Extends netlink socket RTM_SETLINK/
RTM_GETLINK with two new sub msgs called IFLA_VF_PORTS and IFLA_PORT_SELF
(added to end of IFLA_cmd list). These are both nested atrtibutes
using this layout:
[IFLA_NUM_VF]
[IFLA_VF_PORTS]
[IFLA_VF_PORT]
[IFLA_PORT_*], ...
[IFLA_VF_PORT]
[IFLA_PORT_*], ...
...
[IFLA_PORT_SELF]
[IFLA_PORT_*], ...
These attributes are design to be set and get symmetrically. VF_PORTS
is a list of VF_PORTs, one for each VF, when dealing with an SR-IOV
device. PORT_SELF is for the PF of the SR-IOV device, in case it wants
to also have a port-profile, or for the case where the VF==PF, like in
enic patch 2/2 of this patch set.
A port-profile is used to configure/enable the external switch virtual port
backing the netdev interface, not to configure the host-facing side of the
netdev. A port-profile is an identifier known to the switch. How port-
profiles are installed on the switch or how available port-profiles are
made know to the host is outside the scope of this patch.
There are two types of port-profiles specs in the netlink msg. The first spec
is for 802.1Qbg (pre-)standard, VDP protocol. The second spec is for devices
that run a similar protocol as VDP but in firmware, thus hiding the protocol
details. In either case, the specs have much in common and makes sense to
define the netlink msg as the union of the two specs. For example, both specs
have a notition of associating/deassociating a port-profile. And both specs
require some information from the hypervisor manager, such as client port
instance ID.
The general flow is the port-profile is applied to a host netdev interface
using RTM_SETLINK, the receiver of the RTM_SETLINK msg communicates with the
switch, and the switch virtual port backing the host netdev interface is
configured/enabled based on the settings defined by the port-profile. What
those settings comprise, and how those settings are managed is again
outside the scope of this patch, since this patch only deals with the
first step in the flow.
Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: Roopa Prabhu <roprabhu@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-18 13:49:55 +08:00
|
|
|
* int (*ndo_set_vf_port)(struct net_device *dev, int vf,
|
|
|
|
* struct nlattr *port[]);
|
2015-03-31 02:35:23 +08:00
|
|
|
*
|
|
|
|
* Enable or disable the VF ability to query its RSS Redirection Table and
|
|
|
|
* Hash Key. This is needed since on some devices VF share this information
|
|
|
|
* with PF and querying it may adduce a theoretical security risk.
|
|
|
|
* int (*ndo_set_vf_rss_query_en)(struct net_device *dev, int vf, bool setting);
|
net: Add netlink support for virtual port management (was iovnl)
Add new netdev ops ndo_{set|get}_vf_port to allow setting of
port-profile on a netdev interface. Extends netlink socket RTM_SETLINK/
RTM_GETLINK with two new sub msgs called IFLA_VF_PORTS and IFLA_PORT_SELF
(added to end of IFLA_cmd list). These are both nested atrtibutes
using this layout:
[IFLA_NUM_VF]
[IFLA_VF_PORTS]
[IFLA_VF_PORT]
[IFLA_PORT_*], ...
[IFLA_VF_PORT]
[IFLA_PORT_*], ...
...
[IFLA_PORT_SELF]
[IFLA_PORT_*], ...
These attributes are design to be set and get symmetrically. VF_PORTS
is a list of VF_PORTs, one for each VF, when dealing with an SR-IOV
device. PORT_SELF is for the PF of the SR-IOV device, in case it wants
to also have a port-profile, or for the case where the VF==PF, like in
enic patch 2/2 of this patch set.
A port-profile is used to configure/enable the external switch virtual port
backing the netdev interface, not to configure the host-facing side of the
netdev. A port-profile is an identifier known to the switch. How port-
profiles are installed on the switch or how available port-profiles are
made know to the host is outside the scope of this patch.
There are two types of port-profiles specs in the netlink msg. The first spec
is for 802.1Qbg (pre-)standard, VDP protocol. The second spec is for devices
that run a similar protocol as VDP but in firmware, thus hiding the protocol
details. In either case, the specs have much in common and makes sense to
define the netlink msg as the union of the two specs. For example, both specs
have a notition of associating/deassociating a port-profile. And both specs
require some information from the hypervisor manager, such as client port
instance ID.
The general flow is the port-profile is applied to a host netdev interface
using RTM_SETLINK, the receiver of the RTM_SETLINK msg communicates with the
switch, and the switch virtual port backing the host netdev interface is
configured/enabled based on the settings defined by the port-profile. What
those settings comprise, and how those settings are managed is again
outside the scope of this patch, since this patch only deals with the
first step in the flow.
Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: Roopa Prabhu <roprabhu@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-18 13:49:55 +08:00
|
|
|
* int (*ndo_get_vf_port)(struct net_device *dev, int vf, struct sk_buff *skb);
|
2011-01-17 16:06:04 +08:00
|
|
|
* int (*ndo_setup_tc)(struct net_device *dev, u8 tc)
|
|
|
|
* Called to setup 'tc' number of traffic classes in the net device. This
|
|
|
|
* is always called from the stack with the rtnl lock held and netif tx
|
|
|
|
* queues stopped. This allows the netdevice to perform queue management
|
|
|
|
* safely.
|
2011-01-19 19:03:53 +08:00
|
|
|
*
|
2011-03-09 16:48:03 +08:00
|
|
|
* Fiber Channel over Ethernet (FCoE) offload functions.
|
|
|
|
* int (*ndo_fcoe_enable)(struct net_device *dev);
|
|
|
|
* Called when the FCoE protocol stack wants to start using LLD for FCoE
|
|
|
|
* so the underlying device can perform whatever needed configuration or
|
|
|
|
* initialization to support acceleration of FCoE traffic.
|
|
|
|
*
|
|
|
|
* int (*ndo_fcoe_disable)(struct net_device *dev);
|
|
|
|
* Called when the FCoE protocol stack wants to stop using LLD for FCoE
|
|
|
|
* so the underlying device can perform whatever needed clean-ups to
|
|
|
|
* stop supporting acceleration of FCoE traffic.
|
|
|
|
*
|
|
|
|
* int (*ndo_fcoe_ddp_setup)(struct net_device *dev, u16 xid,
|
|
|
|
* struct scatterlist *sgl, unsigned int sgc);
|
|
|
|
* Called when the FCoE Initiator wants to initialize an I/O that
|
|
|
|
* is a possible candidate for Direct Data Placement (DDP). The LLD can
|
|
|
|
* perform necessary setup and returns 1 to indicate the device is set up
|
|
|
|
* successfully to perform DDP on this I/O, otherwise this returns 0.
|
|
|
|
*
|
|
|
|
* int (*ndo_fcoe_ddp_done)(struct net_device *dev, u16 xid);
|
|
|
|
* Called when the FCoE Initiator/Target is done with the DDPed I/O as
|
|
|
|
* indicated by the FC exchange id 'xid', so the underlying device can
|
|
|
|
* clean up and reuse resources for later DDP requests.
|
|
|
|
*
|
|
|
|
* int (*ndo_fcoe_ddp_target)(struct net_device *dev, u16 xid,
|
|
|
|
* struct scatterlist *sgl, unsigned int sgc);
|
|
|
|
* Called when the FCoE Target wants to initialize an I/O that
|
|
|
|
* is a possible candidate for Direct Data Placement (DDP). The LLD can
|
|
|
|
* perform necessary setup and returns 1 to indicate the device is set up
|
|
|
|
* successfully to perform DDP on this I/O, otherwise this returns 0.
|
|
|
|
*
|
2012-01-05 04:23:39 +08:00
|
|
|
* int (*ndo_fcoe_get_hbainfo)(struct net_device *dev,
|
|
|
|
* struct netdev_fcoe_hbainfo *hbainfo);
|
|
|
|
* Called when the FCoE Protocol stack wants information on the underlying
|
|
|
|
* device. This information is utilized by the FCoE protocol stack to
|
|
|
|
* register attributes with Fiber Channel management service as per the
|
|
|
|
* FC-GS Fabric Device Management Information(FDMI) specification.
|
|
|
|
*
|
2011-03-09 16:48:03 +08:00
|
|
|
* int (*ndo_fcoe_get_wwn)(struct net_device *dev, u64 *wwn, int type);
|
|
|
|
* Called when the underlying device wants to override default World Wide
|
|
|
|
* Name (WWN) generation mechanism in FCoE protocol stack to pass its own
|
|
|
|
* World Wide Port Name (WWPN) or World Wide Node Name (WWNN) to the FCoE
|
|
|
|
* protocol stack to use.
|
|
|
|
*
|
2011-01-19 19:03:53 +08:00
|
|
|
* RFS acceleration.
|
|
|
|
* int (*ndo_rx_flow_steer)(struct net_device *dev, const struct sk_buff *skb,
|
|
|
|
* u16 rxq_index, u32 flow_id);
|
|
|
|
* Set hardware filter for RFS. rxq_index is the target queue index;
|
|
|
|
* flow_id is a flow ID to be passed to rps_may_expire_flow() later.
|
|
|
|
* Return the filter ID on success, or a negative error code.
|
2011-02-13 18:15:37 +08:00
|
|
|
*
|
2013-01-04 06:49:02 +08:00
|
|
|
* Slave management functions (for bridge, bonding, etc).
|
2011-02-13 18:15:37 +08:00
|
|
|
* int (*ndo_add_slave)(struct net_device *dev, struct net_device *slave_dev);
|
|
|
|
* Called to make another netdev an underling.
|
|
|
|
*
|
|
|
|
* int (*ndo_del_slave)(struct net_device *dev, struct net_device *slave_dev);
|
|
|
|
* Called to release previously enslaved netdev.
|
2011-02-16 00:59:17 +08:00
|
|
|
*
|
|
|
|
* Feature/offload setting functions.
|
2011-11-15 23:29:55 +08:00
|
|
|
* netdev_features_t (*ndo_fix_features)(struct net_device *dev,
|
|
|
|
* netdev_features_t features);
|
2011-02-16 00:59:17 +08:00
|
|
|
* Adjusts the requested feature flags according to device-specific
|
|
|
|
* constraints, and returns the resulting flags. Must not modify
|
|
|
|
* the device state.
|
|
|
|
*
|
2011-11-15 23:29:55 +08:00
|
|
|
* int (*ndo_set_features)(struct net_device *dev, netdev_features_t features);
|
2011-02-16 00:59:17 +08:00
|
|
|
* Called to update device configuration to new features. Passed
|
|
|
|
* feature set might be less than what was returned by ndo_fix_features()).
|
|
|
|
* Must return >0 or -errno if it changed dev->features itself.
|
|
|
|
*
|
2012-10-01 20:32:33 +08:00
|
|
|
* int (*ndo_fdb_add)(struct ndmsg *ndm, struct nlattr *tb[],
|
|
|
|
* struct net_device *dev,
|
2014-11-28 21:34:15 +08:00
|
|
|
* const unsigned char *addr, u16 vid, u16 flags)
|
2012-04-15 14:43:56 +08:00
|
|
|
* Adds an FDB entry to dev for addr.
|
2013-02-13 20:00:18 +08:00
|
|
|
* int (*ndo_fdb_del)(struct ndmsg *ndm, struct nlattr *tb[],
|
|
|
|
* struct net_device *dev,
|
2014-11-28 21:34:15 +08:00
|
|
|
* const unsigned char *addr, u16 vid)
|
2012-04-15 14:43:56 +08:00
|
|
|
* Deletes the FDB entry from dev coresponding to addr.
|
|
|
|
* int (*ndo_fdb_dump)(struct sk_buff *skb, struct netlink_callback *cb,
|
2014-07-10 19:01:58 +08:00
|
|
|
* struct net_device *dev, struct net_device *filter_dev,
|
|
|
|
* int idx)
|
2012-04-15 14:43:56 +08:00
|
|
|
* Used to add FDB entries to dump requests. Implementers should add
|
|
|
|
* entries to skb and update idx with the number of entries.
|
2012-10-24 16:12:57 +08:00
|
|
|
*
|
2015-03-17 18:16:00 +08:00
|
|
|
* int (*ndo_bridge_setlink)(struct net_device *dev, struct nlmsghdr *nlh,
|
|
|
|
* u16 flags)
|
2012-10-24 16:12:57 +08:00
|
|
|
* int (*ndo_bridge_getlink)(struct sk_buff *skb, u32 pid, u32 seq,
|
2015-04-29 00:33:49 +08:00
|
|
|
* struct net_device *dev, u32 filter_mask,
|
|
|
|
* int nlflags)
|
2015-03-17 18:16:00 +08:00
|
|
|
* int (*ndo_bridge_dellink)(struct net_device *dev, struct nlmsghdr *nlh,
|
|
|
|
* u16 flags);
|
2012-12-28 07:49:37 +08:00
|
|
|
*
|
|
|
|
* int (*ndo_change_carrier)(struct net_device *dev, bool new_carrier);
|
|
|
|
* Called to change device carrier. Soft-devices (like dummy, team, etc)
|
|
|
|
* which do not represent real hardware may define this to allow their
|
|
|
|
* userspace components to manage their virtual carrier state. Devices
|
|
|
|
* that determine carrier state from physical hardware properties (eg
|
|
|
|
* network cables) or protocol-dependent mechanisms (eg
|
|
|
|
* USB_CDC_NOTIFY_NETWORK_CONNECTION) should NOT implement this function.
|
2013-07-30 00:16:49 +08:00
|
|
|
*
|
|
|
|
* int (*ndo_get_phys_port_id)(struct net_device *dev,
|
2014-11-28 21:34:16 +08:00
|
|
|
* struct netdev_phys_item_id *ppid);
|
2013-07-30 00:16:49 +08:00
|
|
|
* Called to get ID of physical port of this device. If driver does
|
|
|
|
* not implement this, it is assumed that the hw is not able to have
|
|
|
|
* multiple net devices on single physical port.
|
2013-09-04 17:13:38 +08:00
|
|
|
*
|
|
|
|
* void (*ndo_add_vxlan_port)(struct net_device *dev,
|
2013-09-13 22:34:13 +08:00
|
|
|
* sa_family_t sa_family, __be16 port);
|
2013-09-04 17:13:38 +08:00
|
|
|
* Called by vxlan to notiy a driver about the UDP port and socket
|
|
|
|
* address family that vxlan is listnening to. It is called only when
|
|
|
|
* a new port starts listening. The operation is protected by the
|
|
|
|
* vxlan_net->sock_lock.
|
|
|
|
*
|
|
|
|
* void (*ndo_del_vxlan_port)(struct net_device *dev,
|
2013-09-13 22:34:13 +08:00
|
|
|
* sa_family_t sa_family, __be16 port);
|
2013-09-04 17:13:38 +08:00
|
|
|
* Called by vxlan to notify the driver about a UDP port and socket
|
|
|
|
* address family that vxlan is not listening to anymore. The operation
|
|
|
|
* is protected by the vxlan_net->sock_lock.
|
2013-11-07 01:54:46 +08:00
|
|
|
*
|
|
|
|
* void* (*ndo_dfwd_add_station)(struct net_device *pdev,
|
|
|
|
* struct net_device *dev)
|
|
|
|
* Called by upper layer devices to accelerate switching or other
|
|
|
|
* station functionality into hardware. 'pdev is the lowerdev
|
|
|
|
* to use for the offload and 'dev' is the net device that will
|
|
|
|
* back the offload. Returns a pointer to the private structure
|
|
|
|
* the upper layer will maintain.
|
|
|
|
* void (*ndo_dfwd_del_station)(struct net_device *pdev, void *priv)
|
|
|
|
* Called by upper layer device to delete the station created
|
|
|
|
* by 'ndo_dfwd_add_station'. 'pdev' is the net device backing
|
|
|
|
* the station and priv is the structure returned by the add
|
|
|
|
* operation.
|
|
|
|
* netdev_tx_t (*ndo_dfwd_start_xmit)(struct sk_buff *skb,
|
|
|
|
* struct net_device *dev,
|
|
|
|
* void *priv);
|
|
|
|
* Callback to use for xmit over the accelerated station. This
|
|
|
|
* is used in place of ndo_start_xmit on accelerated net
|
|
|
|
* devices.
|
2014-12-24 14:37:26 +08:00
|
|
|
* netdev_features_t (*ndo_features_check) (struct sk_buff *skb,
|
|
|
|
* struct net_device *dev
|
|
|
|
* netdev_features_t features);
|
2014-10-15 06:19:06 +08:00
|
|
|
* Called by core transmit path to determine if device is capable of
|
2014-12-24 14:37:26 +08:00
|
|
|
* performing offload operations on a given packet. This is to give
|
|
|
|
* the device an opportunity to implement any restrictions that cannot
|
|
|
|
* be otherwise expressed by feature flags. The check is called with
|
|
|
|
* the set of features that the stack has calculated and it returns
|
|
|
|
* those the driver believes to be appropriate.
|
2015-03-18 20:57:33 +08:00
|
|
|
* int (*ndo_set_tx_maxrate)(struct net_device *dev,
|
|
|
|
* int queue_index, u32 maxrate);
|
|
|
|
* Called when a user wants to set a max-rate limitation of specific
|
|
|
|
* TX queue.
|
2015-04-02 23:07:00 +08:00
|
|
|
* int (*ndo_get_iflink)(const struct net_device *dev);
|
|
|
|
* Called to get the iflink value of this device.
|
2015-07-15 04:43:19 +08:00
|
|
|
* void (*ndo_change_proto_down)(struct net_device *dev,
|
|
|
|
* bool proto_down);
|
|
|
|
* This function is used to pass protocol port error state information
|
|
|
|
* to the switch driver. The switch driver can react to the proto_down
|
|
|
|
* by doing a phys down on the associated switch port.
|
|
|
|
*
|
2008-11-20 13:32:24 +08:00
|
|
|
*/
|
|
|
|
struct net_device_ops {
|
|
|
|
int (*ndo_init)(struct net_device *dev);
|
|
|
|
void (*ndo_uninit)(struct net_device *dev);
|
|
|
|
int (*ndo_open)(struct net_device *dev);
|
|
|
|
int (*ndo_stop)(struct net_device *dev);
|
2009-09-01 03:50:40 +08:00
|
|
|
netdev_tx_t (*ndo_start_xmit) (struct sk_buff *skb,
|
2008-11-21 12:14:53 +08:00
|
|
|
struct net_device *dev);
|
|
|
|
u16 (*ndo_select_queue)(struct net_device *dev,
|
2014-01-10 16:18:26 +08:00
|
|
|
struct sk_buff *skb,
|
2014-02-16 22:55:20 +08:00
|
|
|
void *accel_priv,
|
|
|
|
select_queue_fallback_t fallback);
|
2008-11-20 13:32:24 +08:00
|
|
|
void (*ndo_change_rx_flags)(struct net_device *dev,
|
|
|
|
int flags);
|
|
|
|
void (*ndo_set_rx_mode)(struct net_device *dev);
|
|
|
|
int (*ndo_set_mac_address)(struct net_device *dev,
|
|
|
|
void *addr);
|
|
|
|
int (*ndo_validate_addr)(struct net_device *dev);
|
|
|
|
int (*ndo_do_ioctl)(struct net_device *dev,
|
|
|
|
struct ifreq *ifr, int cmd);
|
|
|
|
int (*ndo_set_config)(struct net_device *dev,
|
|
|
|
struct ifmap *map);
|
2008-11-21 12:14:53 +08:00
|
|
|
int (*ndo_change_mtu)(struct net_device *dev,
|
|
|
|
int new_mtu);
|
|
|
|
int (*ndo_neigh_setup)(struct net_device *dev,
|
|
|
|
struct neigh_parms *);
|
2008-11-20 13:32:24 +08:00
|
|
|
void (*ndo_tx_timeout) (struct net_device *dev);
|
|
|
|
|
2010-07-08 05:58:56 +08:00
|
|
|
struct rtnl_link_stats64* (*ndo_get_stats64)(struct net_device *dev,
|
|
|
|
struct rtnl_link_stats64 *storage);
|
2008-11-20 13:32:24 +08:00
|
|
|
struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);
|
|
|
|
|
2011-12-09 08:52:37 +08:00
|
|
|
int (*ndo_vlan_rx_add_vid)(struct net_device *dev,
|
2013-04-19 10:04:28 +08:00
|
|
|
__be16 proto, u16 vid);
|
2011-12-09 08:52:37 +08:00
|
|
|
int (*ndo_vlan_rx_kill_vid)(struct net_device *dev,
|
2013-04-19 10:04:28 +08:00
|
|
|
__be16 proto, u16 vid);
|
2008-11-20 13:32:24 +08:00
|
|
|
#ifdef CONFIG_NET_POLL_CONTROLLER
|
|
|
|
void (*ndo_poll_controller)(struct net_device *dev);
|
2010-06-11 00:12:47 +08:00
|
|
|
int (*ndo_netpoll_setup)(struct net_device *dev,
|
2014-03-28 06:36:38 +08:00
|
|
|
struct netpoll_info *info);
|
2010-05-06 15:47:21 +08:00
|
|
|
void (*ndo_netpoll_cleanup)(struct net_device *dev);
|
2013-06-10 16:39:50 +08:00
|
|
|
#endif
|
2013-08-01 11:10:25 +08:00
|
|
|
#ifdef CONFIG_NET_RX_BUSY_POLL
|
2013-07-10 22:13:26 +08:00
|
|
|
int (*ndo_busy_poll)(struct napi_struct *dev);
|
2008-11-20 13:32:24 +08:00
|
|
|
#endif
|
2010-02-10 09:43:46 +08:00
|
|
|
int (*ndo_set_vf_mac)(struct net_device *dev,
|
|
|
|
int queue, u8 *mac);
|
|
|
|
int (*ndo_set_vf_vlan)(struct net_device *dev,
|
|
|
|
int queue, u16 vlan, u8 qos);
|
net-next:v4: Add support to configure SR-IOV VF minimum and maximum Tx rate through ip tool.
o min_tx_rate puts lower limit on the VF bandwidth. VF is guaranteed
to have a bandwidth of at least this value.
max_tx_rate puts cap on the VF bandwidth. VF can have a bandwidth
of up to this value.
o A new handler set_vf_rate for attr IFLA_VF_RATE has been introduced
which takes 4 arguments:
netdev, VF number, min_tx_rate, max_tx_rate
o ndo_set_vf_rate replaces ndo_set_vf_tx_rate handler.
o Drivers that currently implement ndo_set_vf_tx_rate should now call
ndo_set_vf_rate instead and reject attempt to set a minimum bandwidth
greater than 0 for IFLA_VF_TX_RATE when IFLA_VF_RATE is not yet
implemented by driver.
o If user enters only one of either min_tx_rate or max_tx_rate, then,
userland should read back the other value from driver and set both
for IFLA_VF_RATE.
Drivers that have not yet implemented IFLA_VF_RATE should always
return min_tx_rate as 0 when read from ip tool.
o If both IFLA_VF_TX_RATE and IFLA_VF_RATE options are specified, then
IFLA_VF_RATE should override.
o Idea is to have consistent display of rate values to user.
o Usage example: -
./ip link set p4p1 vf 0 rate 900
./ip link show p4p1
32: p4p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
link/ether 00:0e:1e:08:b0:f0 brd ff:ff:ff:ff:ff:ff
vf 0 MAC 3e:a0:ca:bd:ae:5a, tx rate 900 (Mbps), max_tx_rate 900Mbps
vf 1 MAC f6:c6:7c:3f:3d:6c
vf 2 MAC 56:32:43:98:d7:71
vf 3 MAC d6:be:c3:b5:85:ff
vf 4 MAC ee:a9:9a:1e:19:14
vf 5 MAC 4a:d0:4c:07:52:18
vf 6 MAC 3a:76:44:93:62:f9
vf 7 MAC 82:e9:e7:e3:15:1a
./ip link set p4p1 vf 0 max_tx_rate 300 min_tx_rate 200
./ip link show p4p1
32: p4p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
link/ether 00:0e:1e:08:b0:f0 brd ff:ff:ff:ff:ff:ff
vf 0 MAC 3e:a0:ca:bd:ae:5a, tx rate 300 (Mbps), max_tx_rate 300Mbps,
min_tx_rate 200Mbps
vf 1 MAC f6:c6:7c:3f:3d:6c
vf 2 MAC 56:32:43:98:d7:71
vf 3 MAC d6:be:c3:b5:85:ff
vf 4 MAC ee:a9:9a:1e:19:14
vf 5 MAC 4a:d0:4c:07:52:18
vf 6 MAC 3a:76:44:93:62:f9
vf 7 MAC 82:e9:e7:e3:15:1a
./ip link set p4p1 vf 0 max_tx_rate 600 rate 300
./ip link show p4p1
32: p4p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
link/ether 00:0e:1e:08:b0:f brd ff:ff:ff:ff:ff:ff
vf 0 MAC 3e:a0:ca:bd:ae:5, tx rate 600 (Mbps), max_tx_rate 600Mbps,
min_tx_rate 200Mbps
vf 1 MAC f6:c6:7c:3f:3d:6c
vf 2 MAC 56:32:43:98:d7:71
vf 3 MAC d6:be:c3:b5:85:ff
vf 4 MAC ee:a9:9a:1e:19:14
vf 5 MAC 4a:d0:4c:07:52:18
vf 6 MAC 3a:76:44:93:62:f9
vf 7 MAC 82:e9:e7:e3:15:1a
Signed-off-by: Sucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-05-22 21:59:05 +08:00
|
|
|
int (*ndo_set_vf_rate)(struct net_device *dev,
|
|
|
|
int vf, int min_tx_rate,
|
|
|
|
int max_tx_rate);
|
2011-10-08 11:05:24 +08:00
|
|
|
int (*ndo_set_vf_spoofchk)(struct net_device *dev,
|
|
|
|
int vf, bool setting);
|
2010-02-10 09:43:46 +08:00
|
|
|
int (*ndo_get_vf_config)(struct net_device *dev,
|
|
|
|
int vf,
|
|
|
|
struct ifla_vf_info *ivf);
|
2013-06-13 18:19:10 +08:00
|
|
|
int (*ndo_set_vf_link_state)(struct net_device *dev,
|
|
|
|
int vf, int link_state);
|
2015-06-15 22:59:07 +08:00
|
|
|
int (*ndo_get_vf_stats)(struct net_device *dev,
|
|
|
|
int vf,
|
|
|
|
struct ifla_vf_stats
|
|
|
|
*vf_stats);
|
net: Add netlink support for virtual port management (was iovnl)
Add new netdev ops ndo_{set|get}_vf_port to allow setting of
port-profile on a netdev interface. Extends netlink socket RTM_SETLINK/
RTM_GETLINK with two new sub msgs called IFLA_VF_PORTS and IFLA_PORT_SELF
(added to end of IFLA_cmd list). These are both nested atrtibutes
using this layout:
[IFLA_NUM_VF]
[IFLA_VF_PORTS]
[IFLA_VF_PORT]
[IFLA_PORT_*], ...
[IFLA_VF_PORT]
[IFLA_PORT_*], ...
...
[IFLA_PORT_SELF]
[IFLA_PORT_*], ...
These attributes are design to be set and get symmetrically. VF_PORTS
is a list of VF_PORTs, one for each VF, when dealing with an SR-IOV
device. PORT_SELF is for the PF of the SR-IOV device, in case it wants
to also have a port-profile, or for the case where the VF==PF, like in
enic patch 2/2 of this patch set.
A port-profile is used to configure/enable the external switch virtual port
backing the netdev interface, not to configure the host-facing side of the
netdev. A port-profile is an identifier known to the switch. How port-
profiles are installed on the switch or how available port-profiles are
made know to the host is outside the scope of this patch.
There are two types of port-profiles specs in the netlink msg. The first spec
is for 802.1Qbg (pre-)standard, VDP protocol. The second spec is for devices
that run a similar protocol as VDP but in firmware, thus hiding the protocol
details. In either case, the specs have much in common and makes sense to
define the netlink msg as the union of the two specs. For example, both specs
have a notition of associating/deassociating a port-profile. And both specs
require some information from the hypervisor manager, such as client port
instance ID.
The general flow is the port-profile is applied to a host netdev interface
using RTM_SETLINK, the receiver of the RTM_SETLINK msg communicates with the
switch, and the switch virtual port backing the host netdev interface is
configured/enabled based on the settings defined by the port-profile. What
those settings comprise, and how those settings are managed is again
outside the scope of this patch, since this patch only deals with the
first step in the flow.
Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: Roopa Prabhu <roprabhu@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-18 13:49:55 +08:00
|
|
|
int (*ndo_set_vf_port)(struct net_device *dev,
|
|
|
|
int vf,
|
|
|
|
struct nlattr *port[]);
|
|
|
|
int (*ndo_get_vf_port)(struct net_device *dev,
|
|
|
|
int vf, struct sk_buff *skb);
|
2015-03-31 02:35:23 +08:00
|
|
|
int (*ndo_set_vf_rss_query_en)(
|
|
|
|
struct net_device *dev,
|
|
|
|
int vf, bool setting);
|
2011-01-17 16:06:04 +08:00
|
|
|
int (*ndo_setup_tc)(struct net_device *dev, u8 tc);
|
2011-11-25 22:40:26 +08:00
|
|
|
#if IS_ENABLED(CONFIG_FCOE)
|
2009-08-31 20:31:36 +08:00
|
|
|
int (*ndo_fcoe_enable)(struct net_device *dev);
|
|
|
|
int (*ndo_fcoe_disable)(struct net_device *dev);
|
2009-02-28 06:06:59 +08:00
|
|
|
int (*ndo_fcoe_ddp_setup)(struct net_device *dev,
|
|
|
|
u16 xid,
|
|
|
|
struct scatterlist *sgl,
|
|
|
|
unsigned int sgc);
|
|
|
|
int (*ndo_fcoe_ddp_done)(struct net_device *dev,
|
|
|
|
u16 xid);
|
2011-02-01 15:22:06 +08:00
|
|
|
int (*ndo_fcoe_ddp_target)(struct net_device *dev,
|
|
|
|
u16 xid,
|
|
|
|
struct scatterlist *sgl,
|
|
|
|
unsigned int sgc);
|
2012-01-05 04:23:39 +08:00
|
|
|
int (*ndo_fcoe_get_hbainfo)(struct net_device *dev,
|
|
|
|
struct netdev_fcoe_hbainfo *hbainfo);
|
2011-08-26 17:45:41 +08:00
|
|
|
#endif
|
|
|
|
|
2011-11-25 22:40:26 +08:00
|
|
|
#if IS_ENABLED(CONFIG_LIBFCOE)
|
2009-10-29 02:24:35 +08:00
|
|
|
#define NETDEV_FCOE_WWNN 0
|
|
|
|
#define NETDEV_FCOE_WWPN 1
|
|
|
|
int (*ndo_fcoe_get_wwn)(struct net_device *dev,
|
|
|
|
u64 *wwn, int type);
|
2009-02-28 06:06:59 +08:00
|
|
|
#endif
|
2011-08-26 17:45:41 +08:00
|
|
|
|
2011-01-19 19:03:53 +08:00
|
|
|
#ifdef CONFIG_RFS_ACCEL
|
|
|
|
int (*ndo_rx_flow_steer)(struct net_device *dev,
|
|
|
|
const struct sk_buff *skb,
|
|
|
|
u16 rxq_index,
|
|
|
|
u32 flow_id);
|
|
|
|
#endif
|
2011-02-13 18:15:37 +08:00
|
|
|
int (*ndo_add_slave)(struct net_device *dev,
|
|
|
|
struct net_device *slave_dev);
|
|
|
|
int (*ndo_del_slave)(struct net_device *dev,
|
|
|
|
struct net_device *slave_dev);
|
2011-11-15 23:29:55 +08:00
|
|
|
netdev_features_t (*ndo_fix_features)(struct net_device *dev,
|
|
|
|
netdev_features_t features);
|
2011-02-16 00:59:17 +08:00
|
|
|
int (*ndo_set_features)(struct net_device *dev,
|
2011-11-15 23:29:55 +08:00
|
|
|
netdev_features_t features);
|
2011-07-25 08:01:38 +08:00
|
|
|
int (*ndo_neigh_construct)(struct neighbour *n);
|
2011-12-20 04:04:41 +08:00
|
|
|
void (*ndo_neigh_destroy)(struct neighbour *n);
|
2012-04-15 14:43:56 +08:00
|
|
|
|
|
|
|
int (*ndo_fdb_add)(struct ndmsg *ndm,
|
2012-10-01 20:32:33 +08:00
|
|
|
struct nlattr *tb[],
|
2012-04-15 14:43:56 +08:00
|
|
|
struct net_device *dev,
|
2012-09-17 18:03:26 +08:00
|
|
|
const unsigned char *addr,
|
2014-11-28 21:34:15 +08:00
|
|
|
u16 vid,
|
2012-04-15 14:43:56 +08:00
|
|
|
u16 flags);
|
|
|
|
int (*ndo_fdb_del)(struct ndmsg *ndm,
|
2013-02-13 20:00:18 +08:00
|
|
|
struct nlattr *tb[],
|
2012-04-15 14:43:56 +08:00
|
|
|
struct net_device *dev,
|
2014-11-28 21:34:15 +08:00
|
|
|
const unsigned char *addr,
|
|
|
|
u16 vid);
|
2012-04-15 14:43:56 +08:00
|
|
|
int (*ndo_fdb_dump)(struct sk_buff *skb,
|
|
|
|
struct netlink_callback *cb,
|
|
|
|
struct net_device *dev,
|
2014-07-10 19:01:58 +08:00
|
|
|
struct net_device *filter_dev,
|
2012-04-15 14:43:56 +08:00
|
|
|
int idx);
|
2012-10-24 16:12:57 +08:00
|
|
|
|
|
|
|
int (*ndo_bridge_setlink)(struct net_device *dev,
|
2015-01-30 14:40:12 +08:00
|
|
|
struct nlmsghdr *nlh,
|
|
|
|
u16 flags);
|
2012-10-24 16:12:57 +08:00
|
|
|
int (*ndo_bridge_getlink)(struct sk_buff *skb,
|
|
|
|
u32 pid, u32 seq,
|
2013-02-13 20:00:13 +08:00
|
|
|
struct net_device *dev,
|
2015-04-29 00:33:49 +08:00
|
|
|
u32 filter_mask,
|
|
|
|
int nlflags);
|
2013-02-13 20:00:12 +08:00
|
|
|
int (*ndo_bridge_dellink)(struct net_device *dev,
|
2015-01-30 14:40:12 +08:00
|
|
|
struct nlmsghdr *nlh,
|
|
|
|
u16 flags);
|
2012-12-28 07:49:37 +08:00
|
|
|
int (*ndo_change_carrier)(struct net_device *dev,
|
|
|
|
bool new_carrier);
|
2013-07-30 00:16:49 +08:00
|
|
|
int (*ndo_get_phys_port_id)(struct net_device *dev,
|
2014-11-28 21:34:16 +08:00
|
|
|
struct netdev_phys_item_id *ppid);
|
2015-03-18 10:23:15 +08:00
|
|
|
int (*ndo_get_phys_port_name)(struct net_device *dev,
|
|
|
|
char *name, size_t len);
|
2013-09-04 17:13:38 +08:00
|
|
|
void (*ndo_add_vxlan_port)(struct net_device *dev,
|
|
|
|
sa_family_t sa_family,
|
2013-09-13 22:34:13 +08:00
|
|
|
__be16 port);
|
2013-09-04 17:13:38 +08:00
|
|
|
void (*ndo_del_vxlan_port)(struct net_device *dev,
|
|
|
|
sa_family_t sa_family,
|
2013-09-13 22:34:13 +08:00
|
|
|
__be16 port);
|
2013-11-07 01:54:46 +08:00
|
|
|
|
|
|
|
void* (*ndo_dfwd_add_station)(struct net_device *pdev,
|
|
|
|
struct net_device *dev);
|
|
|
|
void (*ndo_dfwd_del_station)(struct net_device *pdev,
|
|
|
|
void *priv);
|
|
|
|
|
|
|
|
netdev_tx_t (*ndo_dfwd_start_xmit) (struct sk_buff *skb,
|
|
|
|
struct net_device *dev,
|
|
|
|
void *priv);
|
2014-05-17 05:04:54 +08:00
|
|
|
int (*ndo_get_lock_subclass)(struct net_device *dev);
|
2014-12-24 14:37:26 +08:00
|
|
|
netdev_features_t (*ndo_features_check) (struct sk_buff *skb,
|
|
|
|
struct net_device *dev,
|
|
|
|
netdev_features_t features);
|
2015-03-18 20:57:33 +08:00
|
|
|
int (*ndo_set_tx_maxrate)(struct net_device *dev,
|
|
|
|
int queue_index,
|
|
|
|
u32 maxrate);
|
2015-04-02 23:07:00 +08:00
|
|
|
int (*ndo_get_iflink)(const struct net_device *dev);
|
2015-07-15 04:43:19 +08:00
|
|
|
int (*ndo_change_proto_down)(struct net_device *dev,
|
|
|
|
bool proto_down);
|
2008-11-20 13:32:24 +08:00
|
|
|
};
|
|
|
|
|
2014-02-26 09:15:13 +08:00
|
|
|
/**
|
|
|
|
* enum net_device_priv_flags - &struct net_device priv_flags
|
|
|
|
*
|
|
|
|
* These are the &struct net_device, they are only set internally
|
|
|
|
* by drivers and used in the kernel. These flags are invisible to
|
|
|
|
* userspace, this means that the order of these flags can change
|
|
|
|
* during any kernel release.
|
|
|
|
*
|
|
|
|
* You should have a pretty good reason to be extending these flags.
|
|
|
|
*
|
|
|
|
* @IFF_802_1Q_VLAN: 802.1Q VLAN device
|
|
|
|
* @IFF_EBRIDGE: Ethernet bridging device
|
|
|
|
* @IFF_BONDING: bonding master or slave
|
|
|
|
* @IFF_ISATAP: ISATAP interface (RFC4214)
|
|
|
|
* @IFF_WAN_HDLC: WAN HDLC device
|
|
|
|
* @IFF_XMIT_DST_RELEASE: dev_hard_start_xmit() is allowed to
|
|
|
|
* release skb->dst
|
|
|
|
* @IFF_DONT_BRIDGE: disallow bridging this ether dev
|
|
|
|
* @IFF_DISABLE_NETPOLL: disable netpoll at run-time
|
|
|
|
* @IFF_MACVLAN_PORT: device used as macvlan port
|
|
|
|
* @IFF_BRIDGE_PORT: device used as bridge port
|
|
|
|
* @IFF_OVS_DATAPATH: device used as Open vSwitch datapath port
|
|
|
|
* @IFF_TX_SKB_SHARING: The interface supports sharing skbs on transmit
|
|
|
|
* @IFF_UNICAST_FLT: Supports unicast filtering
|
|
|
|
* @IFF_TEAM_PORT: device used as team port
|
|
|
|
* @IFF_SUPP_NOFCS: device supports sending custom FCS
|
|
|
|
* @IFF_LIVE_ADDR_CHANGE: device supports hardware address
|
|
|
|
* change when it's running
|
|
|
|
* @IFF_MACVLAN: Macvlan device
|
2015-08-17 00:26:49 +08:00
|
|
|
* @IFF_VRF_MASTER: device is a VRF master
|
2015-08-14 01:01:06 +08:00
|
|
|
* @IFF_NO_QUEUE: device can run without qdisc attached
|
2015-08-27 15:31:20 +08:00
|
|
|
* @IFF_OPENVSWITCH: device is a Open vSwitch master
|
2014-02-26 09:15:13 +08:00
|
|
|
*/
|
|
|
|
enum netdev_priv_flags {
|
|
|
|
IFF_802_1Q_VLAN = 1<<0,
|
|
|
|
IFF_EBRIDGE = 1<<1,
|
2015-08-27 15:31:21 +08:00
|
|
|
IFF_BONDING = 1<<2,
|
|
|
|
IFF_ISATAP = 1<<3,
|
|
|
|
IFF_WAN_HDLC = 1<<4,
|
|
|
|
IFF_XMIT_DST_RELEASE = 1<<5,
|
|
|
|
IFF_DONT_BRIDGE = 1<<6,
|
|
|
|
IFF_DISABLE_NETPOLL = 1<<7,
|
|
|
|
IFF_MACVLAN_PORT = 1<<8,
|
|
|
|
IFF_BRIDGE_PORT = 1<<9,
|
|
|
|
IFF_OVS_DATAPATH = 1<<10,
|
|
|
|
IFF_TX_SKB_SHARING = 1<<11,
|
|
|
|
IFF_UNICAST_FLT = 1<<12,
|
|
|
|
IFF_TEAM_PORT = 1<<13,
|
|
|
|
IFF_SUPP_NOFCS = 1<<14,
|
|
|
|
IFF_LIVE_ADDR_CHANGE = 1<<15,
|
|
|
|
IFF_MACVLAN = 1<<16,
|
|
|
|
IFF_XMIT_DST_RELEASE_PERM = 1<<17,
|
|
|
|
IFF_IPVLAN_MASTER = 1<<18,
|
|
|
|
IFF_IPVLAN_SLAVE = 1<<19,
|
|
|
|
IFF_VRF_MASTER = 1<<20,
|
|
|
|
IFF_NO_QUEUE = 1<<21,
|
|
|
|
IFF_OPENVSWITCH = 1<<22,
|
2014-02-26 09:15:13 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
#define IFF_802_1Q_VLAN IFF_802_1Q_VLAN
|
|
|
|
#define IFF_EBRIDGE IFF_EBRIDGE
|
|
|
|
#define IFF_BONDING IFF_BONDING
|
|
|
|
#define IFF_ISATAP IFF_ISATAP
|
|
|
|
#define IFF_WAN_HDLC IFF_WAN_HDLC
|
|
|
|
#define IFF_XMIT_DST_RELEASE IFF_XMIT_DST_RELEASE
|
|
|
|
#define IFF_DONT_BRIDGE IFF_DONT_BRIDGE
|
|
|
|
#define IFF_DISABLE_NETPOLL IFF_DISABLE_NETPOLL
|
|
|
|
#define IFF_MACVLAN_PORT IFF_MACVLAN_PORT
|
|
|
|
#define IFF_BRIDGE_PORT IFF_BRIDGE_PORT
|
|
|
|
#define IFF_OVS_DATAPATH IFF_OVS_DATAPATH
|
|
|
|
#define IFF_TX_SKB_SHARING IFF_TX_SKB_SHARING
|
|
|
|
#define IFF_UNICAST_FLT IFF_UNICAST_FLT
|
|
|
|
#define IFF_TEAM_PORT IFF_TEAM_PORT
|
|
|
|
#define IFF_SUPP_NOFCS IFF_SUPP_NOFCS
|
|
|
|
#define IFF_LIVE_ADDR_CHANGE IFF_LIVE_ADDR_CHANGE
|
|
|
|
#define IFF_MACVLAN IFF_MACVLAN
|
2014-10-06 09:38:35 +08:00
|
|
|
#define IFF_XMIT_DST_RELEASE_PERM IFF_XMIT_DST_RELEASE_PERM
|
2014-11-24 15:07:46 +08:00
|
|
|
#define IFF_IPVLAN_MASTER IFF_IPVLAN_MASTER
|
|
|
|
#define IFF_IPVLAN_SLAVE IFF_IPVLAN_SLAVE
|
2015-08-14 04:59:00 +08:00
|
|
|
#define IFF_VRF_MASTER IFF_VRF_MASTER
|
2015-08-14 01:01:06 +08:00
|
|
|
#define IFF_NO_QUEUE IFF_NO_QUEUE
|
2015-08-27 15:31:20 +08:00
|
|
|
#define IFF_OPENVSWITCH IFF_OPENVSWITCH
|
2014-02-26 09:15:13 +08:00
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
/**
|
|
|
|
* struct net_device - The DEVICE structure.
|
|
|
|
* Actually, this whole structure is a big mistake. It mixes I/O
|
|
|
|
* data with strictly "high-level" data, and it has to know about
|
|
|
|
* almost every data structure used in the INET module.
|
|
|
|
*
|
|
|
|
* @name: This is the first field of the "visible" part of this structure
|
|
|
|
* (i.e. as seen by users in the "Space.c" file). It is the name
|
|
|
|
* of the interface.
|
|
|
|
*
|
|
|
|
* @name_hlist: Device name hash chain, please keep it close to name[]
|
|
|
|
* @ifalias: SNMP alias
|
|
|
|
* @mem_end: Shared memory end
|
|
|
|
* @mem_start: Shared memory start
|
|
|
|
* @base_addr: Device I/O address
|
|
|
|
* @irq: Device IRQ number
|
|
|
|
*
|
2015-04-10 21:52:38 +08:00
|
|
|
* @carrier_changes: Stats to monitor carrier on<->off transitions
|
|
|
|
*
|
2014-07-31 02:27:36 +08:00
|
|
|
* @state: Generic network queuing layer state, see netdev_state_t
|
|
|
|
* @dev_list: The global list of network devices
|
|
|
|
* @napi_list: List entry, that is used for polling napi devices
|
|
|
|
* @unreg_list: List entry, that is used, when we are unregistering the
|
|
|
|
* device, see the function unregister_netdev
|
|
|
|
* @close_list: List entry, that is used, when we are closing the device
|
|
|
|
*
|
|
|
|
* @adj_list: Directly linked devices, like slaves for bonding
|
|
|
|
* @all_adj_list: All linked devices, *including* neighbours
|
|
|
|
* @features: Currently active device features
|
|
|
|
* @hw_features: User-changeable features
|
|
|
|
*
|
|
|
|
* @wanted_features: User-requested features
|
|
|
|
* @vlan_features: Mask of features inheritable by VLAN devices
|
|
|
|
*
|
|
|
|
* @hw_enc_features: Mask of features inherited by encapsulating devices
|
|
|
|
* This field indicates what encapsulation
|
|
|
|
* offloads the hardware is capable of doing,
|
|
|
|
* and drivers will need to set them appropriately.
|
|
|
|
*
|
|
|
|
* @mpls_features: Mask of features inheritable by MPLS
|
|
|
|
*
|
|
|
|
* @ifindex: interface index
|
2015-04-07 18:10:16 +08:00
|
|
|
* @group: The group, that the device belongs to
|
2014-07-31 02:27:36 +08:00
|
|
|
*
|
|
|
|
* @stats: Statistics struct, which was left as a legacy, use
|
|
|
|
* rtnl_link_stats64 instead
|
|
|
|
*
|
|
|
|
* @rx_dropped: Dropped packets by core network,
|
|
|
|
* do not use this in drivers
|
|
|
|
* @tx_dropped: Dropped packets by core network,
|
|
|
|
* do not use this in drivers
|
|
|
|
*
|
|
|
|
* @wireless_handlers: List of functions to handle Wireless Extensions,
|
|
|
|
* instead of ioctl,
|
|
|
|
* see <net/iw_handler.h> for details.
|
|
|
|
* @wireless_data: Instance data managed by the core of wireless extensions
|
|
|
|
*
|
|
|
|
* @netdev_ops: Includes several pointers to callbacks,
|
|
|
|
* if one wants to override the ndo_*() functions
|
|
|
|
* @ethtool_ops: Management operations
|
2015-03-02 14:11:09 +08:00
|
|
|
* @header_ops: Includes callbacks for creating,parsing,caching,etc
|
2014-07-31 02:27:36 +08:00
|
|
|
* of Layer 2 headers.
|
|
|
|
*
|
|
|
|
* @flags: Interface flags (a la BSD)
|
|
|
|
* @priv_flags: Like 'flags' but invisible to userspace,
|
|
|
|
* see if.h for the definitions
|
|
|
|
* @gflags: Global flags ( kept as legacy )
|
|
|
|
* @padded: How much padding added by alloc_netdev()
|
|
|
|
* @operstate: RFC2863 operstate
|
|
|
|
* @link_mode: Mapping policy to operstate
|
|
|
|
* @if_port: Selectable AUI, TP, ...
|
|
|
|
* @dma: DMA channel
|
|
|
|
* @mtu: Interface MTU value
|
|
|
|
* @type: Interface hardware type
|
|
|
|
* @hard_header_len: Hardware header length
|
|
|
|
*
|
|
|
|
* @needed_headroom: Extra headroom the hardware may need, but not in all
|
|
|
|
* cases can this be guaranteed
|
|
|
|
* @needed_tailroom: Extra tailroom the hardware may need, but not in all
|
|
|
|
* cases can this be guaranteed. Some cases also use
|
|
|
|
* LL_MAX_HEADER instead to allocate the skb
|
|
|
|
*
|
|
|
|
* interface address info:
|
|
|
|
*
|
|
|
|
* @perm_addr: Permanent hw address
|
|
|
|
* @addr_assign_type: Hw address assignment type
|
|
|
|
* @addr_len: Hardware address length
|
|
|
|
* @neigh_priv_len; Used in neigh_alloc(),
|
|
|
|
* initialized only in atm/clip.c
|
|
|
|
* @dev_id: Used to differentiate devices that share
|
|
|
|
* the same link layer address
|
|
|
|
* @dev_port: Used to differentiate devices that share
|
|
|
|
* the same function
|
|
|
|
* @addr_list_lock: XXX: need comments on this one
|
|
|
|
* @uc_promisc: Counter, that indicates, that promiscuous mode
|
|
|
|
* has been enabled due to the need to listen to
|
|
|
|
* additional unicast addresses in a device that
|
|
|
|
* does not implement ndo_set_rx_mode()
|
2015-04-10 21:52:38 +08:00
|
|
|
* @uc: unicast mac addresses
|
|
|
|
* @mc: multicast mac addresses
|
|
|
|
* @dev_addrs: list of device hw addresses
|
|
|
|
* @queues_kset: Group of all Kobjects in the Tx and RX queues
|
2014-07-31 02:27:36 +08:00
|
|
|
* @promiscuity: Number of times, the NIC is told to work in
|
|
|
|
* Promiscuous mode, if it becomes 0 the NIC will
|
|
|
|
* exit from working in Promiscuous mode
|
|
|
|
* @allmulti: Counter, enables or disables allmulticast mode
|
|
|
|
*
|
|
|
|
* @vlan_info: VLAN info
|
|
|
|
* @dsa_ptr: dsa specific data
|
|
|
|
* @tipc_ptr: TIPC specific data
|
|
|
|
* @atalk_ptr: AppleTalk link
|
|
|
|
* @ip_ptr: IPv4 specific data
|
|
|
|
* @dn_ptr: DECnet specific data
|
|
|
|
* @ip6_ptr: IPv6 specific data
|
|
|
|
* @ax25_ptr: AX.25 specific data
|
2015-08-14 04:59:00 +08:00
|
|
|
* @vrf_ptr: VRF specific data
|
2014-07-31 02:27:36 +08:00
|
|
|
* @ieee80211_ptr: IEEE 802.11 specific data, assign before registering
|
|
|
|
*
|
|
|
|
* @last_rx: Time of last Rx
|
|
|
|
* @dev_addr: Hw address (before bcast,
|
|
|
|
* because most packets are unicast)
|
|
|
|
*
|
|
|
|
* @_rx: Array of RX queues
|
|
|
|
* @num_rx_queues: Number of RX queues
|
|
|
|
* allocated at register_netdev() time
|
|
|
|
* @real_num_rx_queues: Number of RX queues currently active in device
|
|
|
|
*
|
|
|
|
* @rx_handler: handler for received packets
|
|
|
|
* @rx_handler_data: XXX: need comments on this one
|
|
|
|
* @ingress_queue: XXX: need comments on this one
|
|
|
|
* @broadcast: hw bcast address
|
|
|
|
*
|
2015-04-10 21:52:38 +08:00
|
|
|
* @rx_cpu_rmap: CPU reverse-mapping for RX completion interrupts,
|
|
|
|
* indexed by RX queue number. Assigned by driver.
|
|
|
|
* This must only be set if the ndo_rx_flow_steer
|
|
|
|
* operation is defined
|
|
|
|
* @index_hlist: Device index hash chain
|
|
|
|
*
|
2014-07-31 02:27:36 +08:00
|
|
|
* @_tx: Array of TX queues
|
|
|
|
* @num_tx_queues: Number of TX queues allocated at alloc_netdev_mq() time
|
|
|
|
* @real_num_tx_queues: Number of TX queues currently active in device
|
|
|
|
* @qdisc: Root qdisc from userspace point of view
|
|
|
|
* @tx_queue_len: Max frames per queue allowed
|
|
|
|
* @tx_global_lock: XXX: need comments on this one
|
|
|
|
*
|
|
|
|
* @xps_maps: XXX: need comments on this one
|
|
|
|
*
|
2015-07-19 09:24:48 +08:00
|
|
|
* @offload_fwd_mark: Offload device fwding mark
|
|
|
|
*
|
2014-07-31 02:27:36 +08:00
|
|
|
* @trans_start: Time (in jiffies) of last Tx
|
|
|
|
* @watchdog_timeo: Represents the timeout that is used by
|
|
|
|
* the watchdog ( see dev_watchdog() )
|
|
|
|
* @watchdog_timer: List of timers
|
|
|
|
*
|
|
|
|
* @pcpu_refcnt: Number of references to this device
|
|
|
|
* @todo_list: Delayed register/unregister
|
|
|
|
* @link_watch_list: XXX: need comments on this one
|
|
|
|
*
|
|
|
|
* @reg_state: Register/unregister state machine
|
|
|
|
* @dismantle: Device is going to be freed
|
|
|
|
* @rtnl_link_state: This enum represents the phases of creating
|
|
|
|
* a new link
|
|
|
|
*
|
|
|
|
* @destructor: Called from unregister,
|
|
|
|
* can be used to call free_netdev
|
|
|
|
* @npinfo: XXX: need comments on this one
|
|
|
|
* @nd_net: Network namespace this network device is inside
|
|
|
|
*
|
|
|
|
* @ml_priv: Mid-layer private
|
|
|
|
* @lstats: Loopback statistics
|
|
|
|
* @tstats: Tunnel statistics
|
|
|
|
* @dstats: Dummy statistics
|
|
|
|
* @vstats: Virtual ethernet statistics
|
|
|
|
*
|
|
|
|
* @garp_port: GARP
|
|
|
|
* @mrp_port: MRP
|
|
|
|
*
|
|
|
|
* @dev: Class/net/name entry
|
|
|
|
* @sysfs_groups: Space for optional device, statistics and wireless
|
|
|
|
* sysfs groups
|
|
|
|
*
|
|
|
|
* @sysfs_rx_queue_group: Space for optional per-rx queue attributes
|
|
|
|
* @rtnl_link_ops: Rtnl_link_ops
|
|
|
|
*
|
|
|
|
* @gso_max_size: Maximum size of generic segmentation offload
|
|
|
|
* @gso_max_segs: Maximum number of segments that can be passed to the
|
|
|
|
* NIC for GSO
|
2014-10-06 01:11:27 +08:00
|
|
|
* @gso_min_segs: Minimum number of segments that can be passed to the
|
|
|
|
* NIC for GSO
|
2014-07-31 02:27:36 +08:00
|
|
|
*
|
|
|
|
* @dcbnl_ops: Data Center Bridging netlink ops
|
|
|
|
* @num_tc: Number of traffic classes in the net device
|
|
|
|
* @tc_to_txq: XXX: need comments on this one
|
|
|
|
* @prio_tc_map XXX: need comments on this one
|
|
|
|
*
|
|
|
|
* @fcoe_ddp_xid: Max exchange id for FCoE LRO by ddp
|
|
|
|
*
|
|
|
|
* @priomap: XXX: need comments on this one
|
|
|
|
* @phydev: Physical device may attach itself
|
|
|
|
* for hardware timestamping
|
|
|
|
*
|
|
|
|
* @qdisc_tx_busylock: XXX: need comments on this one
|
|
|
|
*
|
2015-07-15 04:43:19 +08:00
|
|
|
* @proto_down: protocol port state information can be sent to the
|
|
|
|
* switch driver and used to set the phys state of the
|
|
|
|
* switch port.
|
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* FIXME: cleanup struct net_device such that network protocol info
|
|
|
|
* moves out.
|
|
|
|
*/
|
|
|
|
|
2009-11-05 01:50:58 +08:00
|
|
|
struct net_device {
|
2005-04-17 06:20:36 +08:00
|
|
|
char name[IFNAMSIZ];
|
[NET]: Reorder some hot fields of struct net_device
Place them on separate cache lines in SMP to lower memory bouncing
between multiple CPU accessing the device.
- One part is mostly used on receive path (including
eth_type_trans()) (poll_list, poll, quota, weight, last_rx,
dev_addr, broadcast)
- One part is mostly used on queue transmit path (qdisc)
(queue_lock, qdisc, qdisc_sleeping, qdisc_list, tx_queue_len)
- One part is mostly used on xmit path (device)
(xmit_lock, xmit_lock_owner, priv, hard_start_xmit, trans_start)
'features' is placed outside of these hot points, in a location that
may be shared by all cpus (because mostly read)
name_hlist is moved close to name[IFNAMSIZ] to speedup __dev_get_by_name()
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-28 06:23:16 +08:00
|
|
|
struct hlist_node name_hlist;
|
2008-09-23 12:28:11 +08:00
|
|
|
char *ifalias;
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
* I/O specific fields
|
|
|
|
* FIXME: Merge these and struct ifmap into one
|
|
|
|
*/
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned long mem_end;
|
|
|
|
unsigned long mem_start;
|
|
|
|
unsigned long base_addr;
|
|
|
|
int irq;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2015-04-10 21:52:38 +08:00
|
|
|
atomic_t carrier_changes;
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
2014-07-31 02:27:36 +08:00
|
|
|
* Some hardware also needs these fields (state,dev_list,
|
|
|
|
* napi_list,unreg_list,close_list) but they are not
|
2005-04-17 06:20:36 +08:00
|
|
|
* part of the usual set specified in Space.c.
|
|
|
|
*/
|
|
|
|
|
|
|
|
unsigned long state;
|
|
|
|
|
2007-05-04 06:13:45 +08:00
|
|
|
struct list_head dev_list;
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
struct list_head napi_list;
|
2009-10-27 15:03:04 +08:00
|
|
|
struct list_head unreg_list;
|
2013-10-06 10:26:05 +08:00
|
|
|
struct list_head close_list;
|
2015-01-28 03:35:48 +08:00
|
|
|
struct list_head ptype_all;
|
|
|
|
struct list_head ptype_specific;
|
net: add adj_list to save only neighbours
Currently, we distinguish neighbours (first-level linked devices) from
non-neighbours by the neighbour bool in the netdev_adjacent. This could be
quite time-consuming in case we would like to traverse *only* through
neighbours - cause we'd have to traverse through all devices and check for
this flag, and in a (quite common) scenario where we have lots of vlans on
top of bridge, which is on top of a bond - the bonding would have to go
through all those vlans to get its upper neighbour linked devices.
This situation is really unpleasant, cause there are already a lot of cases
when a device with slaves needs to go through them in hot path.
To fix this, introduce a new upper/lower device lists structure -
adj_list, which contains only the neighbours. It works always in
pair with the all_adj_list structure (renamed from upper/lower_dev_list),
i.e. both of them contain the same links, only that all_adj_list contains
also non-neighbour device links. It's really a small change visible,
currently, only for __netdev_adjacent_dev_insert/remove(), and doesn't
change the main linked logic at all.
Also, add some comments a fix a name collision in
netdev_for_each_upper_dev_rcu() and rework the naming by the following
rules:
netdev_(all_)(upper|lower)_*
If "all_" is present, then we work with the whole list of upper/lower
devices, otherwise - only with direct neighbours. Uninline functions - to
get better stack traces.
CC: "David S. Miller" <davem@davemloft.net>
CC: Eric Dumazet <edumazet@google.com>
CC: Jiri Pirko <jiri@resnulli.us>
CC: Alexander Duyck <alexander.h.duyck@intel.com>
CC: Cong Wang <amwang@redhat.com>
Signed-off-by: Veaceslav Falico <vfalico@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-09-25 15:20:07 +08:00
|
|
|
|
|
|
|
struct {
|
|
|
|
struct list_head upper;
|
|
|
|
struct list_head lower;
|
|
|
|
} adj_list;
|
|
|
|
|
|
|
|
struct {
|
|
|
|
struct list_head upper;
|
|
|
|
struct list_head lower;
|
|
|
|
} all_adj_list;
|
net: reorder some fields of net_device
As time passed, some fields were added in net_device, and not
at sensible offsets.
Lets reorder some fields to reduce number of cache lines in RX path.
Fields not used in data path should be moved out of this critical cache
line.
In particular, move broadcast[] to the end of the rx section,
as it is less used, and ethernet uses only the beginning of the 32bytes
field.
Before patch :
offsetof(struct net_device,dev_addr)=0x258
offsetof(struct net_device,rx_handler)=0x2b8
offsetof(struct net_device,ingress_queue)=0x2c8
offsetof(struct net_device,broadcast)=0x278
After :
offsetof(struct net_device,dev_addr)=0x280
offsetof(struct net_device,rx_handler)=0x298
offsetof(struct net_device,ingress_queue)=0x2a8
offsetof(struct net_device,broadcast)=0x2b0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-30 14:31:03 +08:00
|
|
|
|
2011-11-15 23:29:55 +08:00
|
|
|
netdev_features_t features;
|
|
|
|
netdev_features_t hw_features;
|
|
|
|
netdev_features_t wanted_features;
|
|
|
|
netdev_features_t vlan_features;
|
2012-12-07 22:14:14 +08:00
|
|
|
netdev_features_t hw_enc_features;
|
2013-05-24 05:02:52 +08:00
|
|
|
netdev_features_t mpls_features;
|
2011-01-25 07:32:47 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
int ifindex;
|
2015-04-02 23:07:09 +08:00
|
|
|
int group;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-03-29 05:29:08 +08:00
|
|
|
struct net_device_stats stats;
|
2014-03-27 23:45:56 +08:00
|
|
|
|
|
|
|
atomic_long_t rx_dropped;
|
|
|
|
atomic_long_t tx_dropped;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-04-27 11:48:23 +08:00
|
|
|
#ifdef CONFIG_WIRELESS_EXT
|
2005-04-17 06:20:36 +08:00
|
|
|
const struct iw_handler_def * wireless_handlers;
|
|
|
|
struct iw_public_data * wireless_data;
|
2007-04-27 11:48:23 +08:00
|
|
|
#endif
|
2008-11-20 13:32:24 +08:00
|
|
|
const struct net_device_ops *netdev_ops;
|
2006-09-09 02:16:13 +08:00
|
|
|
const struct ethtool_ops *ethtool_ops;
|
2015-03-16 12:07:14 +08:00
|
|
|
#ifdef CONFIG_NET_SWITCHDEV
|
2015-05-11 00:47:47 +08:00
|
|
|
const struct switchdev_ops *switchdev_ops;
|
2015-03-16 12:07:14 +08:00
|
|
|
#endif
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-10-09 16:40:57 +08:00
|
|
|
const struct header_ops *header_ops;
|
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned int flags;
|
|
|
|
unsigned int priv_flags;
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
unsigned short gflags;
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned short padded;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned char operstate;
|
|
|
|
unsigned char link_mode;
|
2006-03-21 09:09:11 +08:00
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned char if_port;
|
|
|
|
unsigned char dma;
|
2011-05-10 01:42:46 +08:00
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned int mtu;
|
|
|
|
unsigned short type;
|
|
|
|
unsigned short hard_header_len;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-05-13 11:48:31 +08:00
|
|
|
unsigned short needed_headroom;
|
|
|
|
unsigned short needed_tailroom;
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Interface address info. */
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned char perm_addr[MAX_ADDR_LEN];
|
|
|
|
unsigned char addr_assign_type;
|
|
|
|
unsigned char addr_len;
|
2013-12-12 17:15:59 +08:00
|
|
|
unsigned short neigh_priv_len;
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned short dev_id;
|
|
|
|
unsigned short dev_port;
|
2009-05-23 07:22:17 +08:00
|
|
|
spinlock_t addr_list_lock;
|
2015-04-10 21:52:38 +08:00
|
|
|
unsigned char name_assign_type;
|
|
|
|
bool uc_promisc;
|
2014-07-31 02:27:36 +08:00
|
|
|
struct netdev_hw_addr_list uc;
|
|
|
|
struct netdev_hw_addr_list mc;
|
|
|
|
struct netdev_hw_addr_list dev_addrs;
|
|
|
|
|
net: reorder some fields of net_device
As time passed, some fields were added in net_device, and not
at sensible offsets.
Lets reorder some fields to reduce number of cache lines in RX path.
Fields not used in data path should be moved out of this critical cache
line.
In particular, move broadcast[] to the end of the rx section,
as it is less used, and ethernet uses only the beginning of the 32bytes
field.
Before patch :
offsetof(struct net_device,dev_addr)=0x258
offsetof(struct net_device,rx_handler)=0x2b8
offsetof(struct net_device,ingress_queue)=0x2c8
offsetof(struct net_device,broadcast)=0x278
After :
offsetof(struct net_device,dev_addr)=0x280
offsetof(struct net_device,rx_handler)=0x298
offsetof(struct net_device,ingress_queue)=0x2a8
offsetof(struct net_device,broadcast)=0x2b0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-30 14:31:03 +08:00
|
|
|
#ifdef CONFIG_SYSFS
|
|
|
|
struct kset *queues_kset;
|
|
|
|
#endif
|
2008-06-18 12:12:48 +08:00
|
|
|
unsigned int promiscuity;
|
|
|
|
unsigned int allmulti;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
|
|
|
/* Protocol specific pointers */
|
2010-10-20 21:56:05 +08:00
|
|
|
|
2011-11-25 22:40:26 +08:00
|
|
|
#if IS_ENABLED(CONFIG_VLAN_8021Q)
|
2014-07-31 02:27:36 +08:00
|
|
|
struct vlan_info __rcu *vlan_info;
|
2010-10-20 21:56:05 +08:00
|
|
|
#endif
|
2011-11-25 22:38:38 +08:00
|
|
|
#if IS_ENABLED(CONFIG_NET_DSA)
|
2014-07-31 02:27:36 +08:00
|
|
|
struct dsa_switch_tree *dsa_ptr;
|
tipc: remove TIPC usage of field af_packet_priv in struct net_device
TIPC is currently using the field 'af_packet_priv' in struct net_device
as a handle to find the bearer instance associated to the given network
device. But, by doing so it is blocking other networking cleanups, such
as the one discussed here:
http://patchwork.ozlabs.org/patch/178044/
This commit removes this usage from TIPC. Instead, we introduce a new
field, 'tipc_ptr', to the net_device structure, to serve this purpose.
When TIPC bearer is enabled, the bearer object is associated to
'tipc_ptr'. When a TIPC packet arrives in the recv_msg() upcall
from a networking device, the bearer object can now be obtained from
'tipc_ptr'. When a bearer is disabled, the bearer object is detached
from its underlying network device by setting 'tipc_ptr' to NULL.
Additionally, an RCU lock is used to protect the new pointer.
Henceforth, the existing tipc_net_lock is used in write mode to
serialize write accesses to this pointer, while the new RCU lock is
applied on the read side to ensure that the pointer is 100% valid
within its wrapped area for all readers.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Cc: Patrick McHardy <kaber@trash.net>
Reviewed-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-11 12:45:41 +08:00
|
|
|
#endif
|
|
|
|
#if IS_ENABLED(CONFIG_TIPC)
|
2014-07-31 02:27:36 +08:00
|
|
|
struct tipc_bearer __rcu *tipc_ptr;
|
net: Distributed Switch Architecture protocol support
Distributed Switch Architecture is a protocol for managing hardware
switch chips. It consists of a set of MII management registers and
commands to configure the switch, and an ethernet header format to
signal which of the ports of the switch a packet was received from
or is intended to be sent to.
The switches that this driver supports are typically embedded in
access points and routers, and a typical setup with a DSA switch
looks something like this:
+-----------+ +-----------+
| | RGMII | |
| +-------+ +------ 1000baseT MDI ("WAN")
| | | 6-port +------ 1000baseT MDI ("LAN1")
| CPU | | ethernet +------ 1000baseT MDI ("LAN2")
| |MIImgmt| switch +------ 1000baseT MDI ("LAN3")
| +-------+ w/5 PHYs +------ 1000baseT MDI ("LAN4")
| | | |
+-----------+ +-----------+
The switch driver presents each port on the switch as a separate
network interface to Linux, polls the switch to maintain software
link state of those ports, forwards MII management interface
accesses to those network interfaces (e.g. as done by ethtool) to
the switch, and exposes the switch's hardware statistics counters
via the appropriate Linux kernel interfaces.
This initial patch supports the MII management interface register
layout of the Marvell 88E6123, 88E6161 and 88E6165 switch chips, and
supports the "Ethertype DSA" packet tagging format.
(There is no officially registered ethertype for the Ethertype DSA
packet format, so we just grab a random one. The ethertype to use
is programmed into the switch, and the switch driver uses the value
of ETH_P_EDSA for this, so this define can be changed at any time in
the future if the one we chose is allocated to another protocol or
if Ethertype DSA gets its own officially registered ethertype, and
everything will continue to work.)
Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Tested-by: Nicolas Pitre <nico@marvell.com>
Tested-by: Byron Bradley <byron.bbradley@gmail.com>
Tested-by: Tim Ellis <tim.ellis@mac.com>
Tested-by: Peter van Valderen <linux@ddcrew.com>
Tested-by: Dirk Teurlings <dirk@upexia.nl>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-07 21:44:02 +08:00
|
|
|
#endif
|
2014-07-31 02:27:36 +08:00
|
|
|
void *atalk_ptr;
|
|
|
|
struct in_device __rcu *ip_ptr;
|
|
|
|
struct dn_dev __rcu *dn_ptr;
|
|
|
|
struct inet6_dev __rcu *ip6_ptr;
|
|
|
|
void *ax25_ptr;
|
2015-08-14 04:59:00 +08:00
|
|
|
struct net_vrf_dev __rcu *vrf_ptr;
|
2014-07-31 02:27:36 +08:00
|
|
|
struct wireless_dev *ieee80211_ptr;
|
2014-11-02 13:44:54 +08:00
|
|
|
struct wpan_dev *ieee802154_ptr;
|
2015-04-22 18:14:37 +08:00
|
|
|
#if IS_ENABLED(CONFIG_MPLS_ROUTING)
|
|
|
|
struct mpls_dev __rcu *mpls_ptr;
|
|
|
|
#endif
|
2005-04-17 06:20:36 +08:00
|
|
|
|
[NET]: Reorder some hot fields of struct net_device
Place them on separate cache lines in SMP to lower memory bouncing
between multiple CPU accessing the device.
- One part is mostly used on receive path (including
eth_type_trans()) (poll_list, poll, quota, weight, last_rx,
dev_addr, broadcast)
- One part is mostly used on queue transmit path (qdisc)
(queue_lock, qdisc, qdisc_sleeping, qdisc_list, tx_queue_len)
- One part is mostly used on xmit path (device)
(xmit_lock, xmit_lock_owner, priv, hard_start_xmit, trans_start)
'features' is placed outside of these hot points, in a location that
may be shared by all cpus (because mostly read)
name_hlist is moved close to name[IFNAMSIZ] to speedup __dev_get_by_name()
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-28 06:23:16 +08:00
|
|
|
/*
|
2010-09-16 10:58:13 +08:00
|
|
|
* Cache lines mostly used on receive path (including eth_type_trans())
|
[NET]: Reorder some hot fields of struct net_device
Place them on separate cache lines in SMP to lower memory bouncing
between multiple CPU accessing the device.
- One part is mostly used on receive path (including
eth_type_trans()) (poll_list, poll, quota, weight, last_rx,
dev_addr, broadcast)
- One part is mostly used on queue transmit path (qdisc)
(queue_lock, qdisc, qdisc_sleeping, qdisc_list, tx_queue_len)
- One part is mostly used on xmit path (device)
(xmit_lock, xmit_lock_owner, priv, hard_start_xmit, trans_start)
'features' is placed outside of these hot points, in a location that
may be shared by all cpus (because mostly read)
name_hlist is moved close to name[IFNAMSIZ] to speedup __dev_get_by_name()
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-28 06:23:16 +08:00
|
|
|
*/
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned long last_rx;
|
2010-08-31 15:40:16 +08:00
|
|
|
|
[NET]: Reorder some hot fields of struct net_device
Place them on separate cache lines in SMP to lower memory bouncing
between multiple CPU accessing the device.
- One part is mostly used on receive path (including
eth_type_trans()) (poll_list, poll, quota, weight, last_rx,
dev_addr, broadcast)
- One part is mostly used on queue transmit path (qdisc)
(queue_lock, qdisc, qdisc_sleeping, qdisc_list, tx_queue_len)
- One part is mostly used on xmit path (device)
(xmit_lock, xmit_lock_owner, priv, hard_start_xmit, trans_start)
'features' is placed outside of these hot points, in a location that
may be shared by all cpus (because mostly read)
name_hlist is moved close to name[IFNAMSIZ] to speedup __dev_get_by_name()
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-28 06:23:16 +08:00
|
|
|
/* Interface address info used in eth_type_trans() */
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned char *dev_addr;
|
2009-05-05 10:48:28 +08:00
|
|
|
|
2010-03-16 16:03:29 +08:00
|
|
|
|
2014-01-17 14:23:28 +08:00
|
|
|
#ifdef CONFIG_SYSFS
|
2010-03-16 16:03:29 +08:00
|
|
|
struct netdev_rx_queue *_rx;
|
|
|
|
|
|
|
|
unsigned int num_rx_queues;
|
2010-09-27 16:24:33 +08:00
|
|
|
unsigned int real_num_rx_queues;
|
2011-01-19 19:03:53 +08:00
|
|
|
|
2010-03-25 03:13:54 +08:00
|
|
|
#endif
|
2010-03-16 16:03:29 +08:00
|
|
|
|
net: gro: add a per device gro flush timer
Tuning coalescing parameters on NIC can be really hard.
Servers can handle both bulk and RPC like traffic, with conflicting
goals : bulk flows want as big GRO packets as possible, RPC want minimal
latencies.
To reach big GRO packets on 10Gbe NIC, one can use :
ethtool -C eth0 rx-usecs 4 rx-frames 44
But this penalizes rpc sessions, with an increase of latencies, up to
50% in some cases, as NICs generally do not force an interrupt when
a packet with TCP Push flag is received.
Some NICs do not have an absolute timer, only a timer rearmed for every
incoming packet.
This patch uses a different strategy : Let GRO stack decides what do do,
based on traffic pattern.
Packets with Push flag wont be delayed.
Packets without Push flag might be held in GRO engine, if we keep
receiving data.
This new mechanism is off by default, and shall be enabled by setting
/sys/class/net/ethX/gro_flush_timeout to a value in nanosecond.
To fully enable this mechanism, drivers should use napi_complete_done()
instead of napi_complete().
Tested:
Ran 200 netperf TCP_STREAM from A to B (10Gbe mlx4 link, 8 RX queues)
Without this feature, we send back about 305,000 ACK per second.
GRO aggregation ratio is low (811/305 = 2.65 segments per GRO packet)
Setting a timer of 2000 nsec is enough to increase GRO packet sizes
and reduce number of ACK packets. (811/19.2 = 42)
Receiver performs less calls to upper stacks, less wakes up.
This also reduces cpu usage on the sender, as it receives less ACK
packets.
Note that reducing number of wakes up increases cpu efficiency, but can
decrease QPS, as applications wont have the chance to warmup cpu caches
doing a partial read of RPC requests/answers if they fit in one skb.
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811269.80 305732.30 1199462.57 19705.72 0.00
0.00 0.50
B:~# echo 2000 >/sys/class/net/eth0/gro_flush_timeout
B:~# sar -n DEV 1 10 | grep eth0 | tail -1
Average: eth0 811577.30 19230.80 1199916.51 1239.80 0.00
0.00 0.50
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-07 13:09:44 +08:00
|
|
|
unsigned long gro_flush_timeout;
|
2010-11-15 14:38:12 +08:00
|
|
|
rx_handler_func_t __rcu *rx_handler;
|
|
|
|
void __rcu *rx_handler_data;
|
2008-07-17 15:34:19 +08:00
|
|
|
|
2015-05-12 01:28:49 +08:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
net: sched: further simplify handle_ing
Ingress qdisc has no other purpose than calling into tc_classify()
that executes attached classifier(s) and action(s).
It has a 1:1 relationship to dev->ingress_queue. After having commit
087c1a601ad7 ("net: sched: run ingress qdisc without locks") removed
the central ingress lock, one major contention point is gone.
The extra indirection layers however, are not necessary for calling
into ingress qdisc. pktgen calling locally into netif_receive_skb()
with a dummy u32, single CPU result on a Supermicro X10SLM-F, Xeon
E3-1240: before ~21,1 Mpps, after patch ~22,9 Mpps.
We can redirect the private classifier list to the netdev directly,
without changing any classifier API bits (!) and execute on that from
handle_ing() side. The __QDISC_STATE_DEACTIVATE test can be removed,
ingress qdisc doesn't have a queue and thus dev_deactivate_queue()
is also not applicable, ingress_cl_list provides similar behaviour.
In other words, ingress qdisc acts like TCQ_F_BUILTIN qdisc.
One next possible step is the removal of the dev's ingress (dummy)
netdev_queue, and to only have the list member in the netdevice
itself.
Note, the filter chain is RCU protected and individual filter elements
are being kfree'd by sched subsystem after RCU grace period. RCU read
lock is being held by __netif_receive_skb_core().
Joint work with Alexei Starovoitov.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-10 04:51:32 +08:00
|
|
|
struct tcf_proto __rcu *ingress_cl_list;
|
|
|
|
#endif
|
2010-10-02 14:11:55 +08:00
|
|
|
struct netdev_queue __rcu *ingress_queue;
|
netfilter: add netfilter ingress hook after handle_ing() under unique static key
This patch adds the Netfilter ingress hook just after the existing tc ingress
hook, that seems to be the consensus solution for this.
Note that the Netfilter hook resides under the global static key that enables
ingress filtering. Nonetheless, Netfilter still also has its own static key for
minimal impact on the existing handle_ing().
* Without this patch:
Result: OK: 6216490(c6216338+d152) usec, 100000000 (60byte,0frags)
16086246pps 7721Mb/sec (7721398080bps) errors: 100000000
42.46% kpktgend_0 [kernel.kallsyms] [k] __netif_receive_skb_core
25.92% kpktgend_0 [kernel.kallsyms] [k] kfree_skb
7.81% kpktgend_0 [pktgen] [k] pktgen_thread_worker
5.62% kpktgend_0 [kernel.kallsyms] [k] ip_rcv
2.70% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_internal
2.34% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_sk
1.44% kpktgend_0 [kernel.kallsyms] [k] __build_skb
* With this patch:
Result: OK: 6214833(c6214731+d101) usec, 100000000 (60byte,0frags)
16090536pps 7723Mb/sec (7723457280bps) errors: 100000000
41.23% kpktgend_0 [kernel.kallsyms] [k] __netif_receive_skb_core
26.57% kpktgend_0 [kernel.kallsyms] [k] kfree_skb
7.72% kpktgend_0 [pktgen] [k] pktgen_thread_worker
5.55% kpktgend_0 [kernel.kallsyms] [k] ip_rcv
2.78% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_internal
2.06% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_sk
1.43% kpktgend_0 [kernel.kallsyms] [k] __build_skb
* Without this patch + tc ingress:
tc filter add dev eth4 parent ffff: protocol ip prio 1 \
u32 match ip dst 4.3.2.1/32
Result: OK: 9269001(c9268821+d179) usec, 100000000 (60byte,0frags)
10788648pps 5178Mb/sec (5178551040bps) errors: 100000000
40.99% kpktgend_0 [kernel.kallsyms] [k] __netif_receive_skb_core
17.50% kpktgend_0 [kernel.kallsyms] [k] kfree_skb
11.77% kpktgend_0 [cls_u32] [k] u32_classify
5.62% kpktgend_0 [kernel.kallsyms] [k] tc_classify_compat
5.18% kpktgend_0 [pktgen] [k] pktgen_thread_worker
3.23% kpktgend_0 [kernel.kallsyms] [k] tc_classify
2.97% kpktgend_0 [kernel.kallsyms] [k] ip_rcv
1.83% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_internal
1.50% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_sk
0.99% kpktgend_0 [kernel.kallsyms] [k] __build_skb
* With this patch + tc ingress:
tc filter add dev eth4 parent ffff: protocol ip prio 1 \
u32 match ip dst 4.3.2.1/32
Result: OK: 9308218(c9308091+d126) usec, 100000000 (60byte,0frags)
10743194pps 5156Mb/sec (5156733120bps) errors: 100000000
42.01% kpktgend_0 [kernel.kallsyms] [k] __netif_receive_skb_core
17.78% kpktgend_0 [kernel.kallsyms] [k] kfree_skb
11.70% kpktgend_0 [cls_u32] [k] u32_classify
5.46% kpktgend_0 [kernel.kallsyms] [k] tc_classify_compat
5.16% kpktgend_0 [pktgen] [k] pktgen_thread_worker
2.98% kpktgend_0 [kernel.kallsyms] [k] ip_rcv
2.84% kpktgend_0 [kernel.kallsyms] [k] tc_classify
1.96% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_internal
1.57% kpktgend_0 [kernel.kallsyms] [k] netif_receive_skb_sk
Note that the results are very similar before and after.
I can see gcc gets the code under the ingress static key out of the hot path.
Then, on that cold branch, it generates the code to accomodate the netfilter
ingress static key. My explanation for this is that this reduces the pressure
on the instruction cache for non-users as the new code is out of the hot path,
and it comes with minimal impact for tc ingress users.
Using gcc version 4.8.4 on:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 8
[...]
L1d cache: 16K
L1i cache: 64K
L2 cache: 2048K
L3 cache: 8192K
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acked-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-14 00:19:38 +08:00
|
|
|
#ifdef CONFIG_NETFILTER_INGRESS
|
|
|
|
struct list_head nf_hooks_ingress;
|
|
|
|
#endif
|
net: sched: further simplify handle_ing
Ingress qdisc has no other purpose than calling into tc_classify()
that executes attached classifier(s) and action(s).
It has a 1:1 relationship to dev->ingress_queue. After having commit
087c1a601ad7 ("net: sched: run ingress qdisc without locks") removed
the central ingress lock, one major contention point is gone.
The extra indirection layers however, are not necessary for calling
into ingress qdisc. pktgen calling locally into netif_receive_skb()
with a dummy u32, single CPU result on a Supermicro X10SLM-F, Xeon
E3-1240: before ~21,1 Mpps, after patch ~22,9 Mpps.
We can redirect the private classifier list to the netdev directly,
without changing any classifier API bits (!) and execute on that from
handle_ing() side. The __QDISC_STATE_DEACTIVATE test can be removed,
ingress qdisc doesn't have a queue and thus dev_deactivate_queue()
is also not applicable, ingress_cl_list provides similar behaviour.
In other words, ingress qdisc acts like TCQ_F_BUILTIN qdisc.
One next possible step is the removal of the dev's ingress (dummy)
netdev_queue, and to only have the list member in the netdevice
itself.
Note, the filter chain is RCU protected and individual filter elements
are being kfree'd by sched subsystem after RCU grace period. RCU read
lock is being held by __netif_receive_skb_core().
Joint work with Alexei Starovoitov.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-10 04:51:32 +08:00
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned char broadcast[MAX_ADDR_LEN];
|
2015-04-10 21:52:38 +08:00
|
|
|
#ifdef CONFIG_RFS_ACCEL
|
|
|
|
struct cpu_rmap *rx_cpu_rmap;
|
|
|
|
#endif
|
|
|
|
struct hlist_node index_hlist;
|
2010-09-16 10:58:13 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Cache lines mostly used on transmit path
|
|
|
|
*/
|
2008-07-17 15:34:19 +08:00
|
|
|
struct netdev_queue *_tx ____cacheline_aligned_in_smp;
|
|
|
|
unsigned int num_tx_queues;
|
2008-07-17 16:56:23 +08:00
|
|
|
unsigned int real_num_tx_queues;
|
2009-09-04 14:41:18 +08:00
|
|
|
struct Qdisc *qdisc;
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned long tx_queue_len;
|
2008-08-01 07:58:50 +08:00
|
|
|
spinlock_t tx_global_lock;
|
2015-04-10 21:52:38 +08:00
|
|
|
int watchdog_timeo;
|
2010-09-16 10:58:13 +08:00
|
|
|
|
2010-11-26 16:36:09 +08:00
|
|
|
#ifdef CONFIG_XPS
|
2010-11-29 05:43:02 +08:00
|
|
|
struct xps_dev_maps __rcu *xps_maps;
|
2010-11-26 16:36:09 +08:00
|
|
|
#endif
|
2010-11-21 21:17:27 +08:00
|
|
|
|
2015-07-19 09:24:48 +08:00
|
|
|
#ifdef CONFIG_NET_SWITCHDEV
|
|
|
|
u32 offload_fwd_mark;
|
|
|
|
#endif
|
|
|
|
|
[NET]: Reorder some hot fields of struct net_device
Place them on separate cache lines in SMP to lower memory bouncing
between multiple CPU accessing the device.
- One part is mostly used on receive path (including
eth_type_trans()) (poll_list, poll, quota, weight, last_rx,
dev_addr, broadcast)
- One part is mostly used on queue transmit path (qdisc)
(queue_lock, qdisc, qdisc_sleeping, qdisc_list, tx_queue_len)
- One part is mostly used on xmit path (device)
(xmit_lock, xmit_lock_owner, priv, hard_start_xmit, trans_start)
'features' is placed outside of these hot points, in a location that
may be shared by all cpus (because mostly read)
name_hlist is moved close to name[IFNAMSIZ] to speedup __dev_get_by_name()
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-28 06:23:16 +08:00
|
|
|
/* These may be needed for future network-power-down code. */
|
2009-05-18 11:55:16 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* trans_start here is expensive for high speed devices on SMP,
|
|
|
|
* please use netdev_queue->trans_start instead.
|
|
|
|
*/
|
2014-07-31 02:27:36 +08:00
|
|
|
unsigned long trans_start;
|
[NET]: Reorder some hot fields of struct net_device
Place them on separate cache lines in SMP to lower memory bouncing
between multiple CPU accessing the device.
- One part is mostly used on receive path (including
eth_type_trans()) (poll_list, poll, quota, weight, last_rx,
dev_addr, broadcast)
- One part is mostly used on queue transmit path (qdisc)
(queue_lock, qdisc, qdisc_sleeping, qdisc_list, tx_queue_len)
- One part is mostly used on xmit path (device)
(xmit_lock, xmit_lock_owner, priv, hard_start_xmit, trans_start)
'features' is placed outside of these hot points, in a location that
may be shared by all cpus (because mostly read)
name_hlist is moved close to name[IFNAMSIZ] to speedup __dev_get_by_name()
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-28 06:23:16 +08:00
|
|
|
|
|
|
|
struct timer_list watchdog_timer;
|
|
|
|
|
net: percpu net_device refcount
We tried very hard to remove all possible dev_hold()/dev_put() pairs in
network stack, using RCU conversions.
There is still an unavoidable device refcount change for every dst we
create/destroy, and this can slow down some workloads (routers or some
app servers, mmap af_packet)
We can switch to a percpu refcount implementation, now dynamic per_cpu
infrastructure is mature. On a 64 cpus machine, this consumes 256 bytes
per device.
On x86, dev_hold(dev) code :
before
lock incl 0x280(%ebx)
after:
movl 0x260(%ebx),%eax
incl fs:(%eax)
Stress bench :
(Sending 160.000.000 UDP frames,
IP route cache disabled, dual E5540 @2.53GHz,
32bit kernel, FIB_TRIE)
Before:
real 1m1.662s
user 0m14.373s
sys 12m55.960s
After:
real 0m51.179s
user 0m15.329s
sys 10m15.942s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-10-11 18:22:12 +08:00
|
|
|
int __percpu *pcpu_refcnt;
|
2005-04-17 06:20:36 +08:00
|
|
|
struct list_head todo_list;
|
|
|
|
|
linkwatch: linkwatch_forget_dev() to speedup device dismantle
Herbert Xu a écrit :
> On Tue, Nov 17, 2009 at 04:26:04AM -0800, David Miller wrote:
>> Really, the link watch stuff is just due for a redesign. I don't
>> think a simple hack is going to cut it this time, sorry Eric :-)
>
> I have no objections against any redesigns, but since the only
> caller of linkwatch_forget_dev runs in process context with the
> RTNL, it could also legally emit those events.
Thanks guys, here an updated version then, before linkwatch surgery ?
In this version, I force the event to be sent synchronously.
[PATCH net-next-2.6] linkwatch: linkwatch_forget_dev() to speedup device dismantle
time ip link del eth3.103 ; time ip link del eth3.104 ; time ip link del eth3.105
real 0m0.266s
user 0m0.000s
sys 0m0.001s
real 0m0.770s
user 0m0.000s
sys 0m0.000s
real 0m1.022s
user 0m0.000s
sys 0m0.000s
One problem of current schem in vlan dismantle phase is the
holding of device done by following chain :
vlan_dev_stop() ->
netif_carrier_off(dev) ->
linkwatch_fire_event(dev) ->
dev_hold() ...
And __linkwatch_run_queue() runs up to one second later...
A generic fix to this problem is to add a linkwatch_forget_dev() method
to unlink the device from the list of watched devices.
dev->link_watch_next becomes dev->link_watch_list (and use a bit more memory),
to be able to unlink device in O(1).
After patch :
time ip link del eth3.103 ; time ip link del eth3.104 ; time ip link del eth3.105
real 0m0.024s
user 0m0.000s
sys 0m0.000s
real 0m0.032s
user 0m0.000s
sys 0m0.001s
real 0m0.033s
user 0m0.000s
sys 0m0.000s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-11-17 13:59:21 +08:00
|
|
|
struct list_head link_watch_list;
|
2007-05-09 09:34:17 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
enum { NETREG_UNINITIALIZED=0,
|
2006-05-11 04:21:17 +08:00
|
|
|
NETREG_REGISTERED, /* completed register_netdevice */
|
2005-04-17 06:20:36 +08:00
|
|
|
NETREG_UNREGISTERING, /* called unregister_netdevice */
|
|
|
|
NETREG_UNREGISTERED, /* completed unregister todo */
|
|
|
|
NETREG_RELEASED, /* called free_netdev */
|
2009-01-15 13:05:05 +08:00
|
|
|
NETREG_DUMMY, /* dummy device for NAPI poll */
|
2011-05-19 20:24:16 +08:00
|
|
|
} reg_state:8;
|
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
bool dismantle;
|
2010-02-26 14:34:51 +08:00
|
|
|
|
|
|
|
enum {
|
|
|
|
RTNL_LINK_INITIALIZED,
|
|
|
|
RTNL_LINK_INITIALIZING,
|
|
|
|
} rtnl_link_state:16;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-11-20 13:32:24 +08:00
|
|
|
void (*destructor)(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
#ifdef CONFIG_NETPOLL
|
2013-01-23 05:29:39 +08:00
|
|
|
struct netpoll_info __rcu *npinfo;
|
2005-04-17 06:20:36 +08:00
|
|
|
#endif
|
2008-07-15 18:03:33 +08:00
|
|
|
|
2015-03-12 12:06:44 +08:00
|
|
|
possible_net_t nd_net;
|
2007-09-12 17:56:32 +08:00
|
|
|
|
2008-05-12 18:29:11 +08:00
|
|
|
/* mid-layer private */
|
2010-09-24 07:51:51 +08:00
|
|
|
union {
|
2014-07-31 02:27:36 +08:00
|
|
|
void *ml_priv;
|
|
|
|
struct pcpu_lstats __percpu *lstats;
|
2014-01-04 13:57:59 +08:00
|
|
|
struct pcpu_sw_netstats __percpu *tstats;
|
2014-07-31 02:27:36 +08:00
|
|
|
struct pcpu_dstats __percpu *dstats;
|
|
|
|
struct pcpu_vstats __percpu *vstats;
|
2010-09-24 07:51:51 +08:00
|
|
|
};
|
2014-07-31 02:27:36 +08:00
|
|
|
|
2010-10-25 05:32:36 +08:00
|
|
|
struct garp_port __rcu *garp_port;
|
2013-02-09 01:17:06 +08:00
|
|
|
struct mrp_port __rcu *mrp_port;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2014-07-31 02:27:36 +08:00
|
|
|
struct device dev;
|
2009-10-29 22:18:21 +08:00
|
|
|
const struct attribute_group *sysfs_groups[4];
|
2014-01-17 14:23:28 +08:00
|
|
|
const struct attribute_group *sysfs_rx_queue_group;
|
2007-06-14 03:03:51 +08:00
|
|
|
|
|
|
|
const struct rtnl_link_ops *rtnl_link_ops;
|
2007-07-07 04:36:20 +08:00
|
|
|
|
[NET]: Add per-connection option to set max TSO frame size
Update: My mailer ate one of Jarek's feedback mails... Fixed the
parameter in netif_set_gso_max_size() to be u32, not u16. Fixed the
whitespace issue due to a patch import botch. Changed the types from
u32 to unsigned int to be more consistent with other variables in the
area. Also brought the patch up to the latest net-2.6.26 tree.
Update: Made gso_max_size container 32 bits, not 16. Moved the
location of gso_max_size within netdev to be less hotpath. Made more
consistent names between the sock and netdev layers, and added a
define for the max GSO size.
Update: Respun for net-2.6.26 tree.
Update: changed max_gso_frame_size and sk_gso_max_size from signed to
unsigned - thanks Stephen!
This patch adds the ability for device drivers to control the size of
the TSO frames being sent to them, per TCP connection. By setting the
netdevice's gso_max_size value, the socket layer will set the GSO
frame size based on that value. This will propogate into the TCP
layer, and send TSO's of that size to the hardware.
This can be desirable to help tune the bursty nature of TSO on a
per-adapter basis, where one may have 1 GbE and 10 GbE devices
coexisting in a system, one running multiqueue and the other not, etc.
This can also be desirable for devices that cannot support full 64 KB
TSO's, but still want to benefit from some level of segmentation
offloading.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-21 18:43:19 +08:00
|
|
|
/* for setting kernel sock attribute on TCP connection setup */
|
|
|
|
#define GSO_MAX_SIZE 65536
|
|
|
|
unsigned int gso_max_size;
|
2012-07-30 23:57:00 +08:00
|
|
|
#define GSO_MAX_SEGS 65535
|
|
|
|
u16 gso_max_segs;
|
2014-10-06 01:11:27 +08:00
|
|
|
u16 gso_min_segs;
|
2008-11-25 17:02:08 +08:00
|
|
|
#ifdef CONFIG_DCB
|
2009-10-05 14:01:03 +08:00
|
|
|
const struct dcbnl_rtnl_ops *dcbnl_ops;
|
2008-11-21 12:52:10 +08:00
|
|
|
#endif
|
2011-01-17 16:06:04 +08:00
|
|
|
u8 num_tc;
|
|
|
|
struct netdev_tc_txq tc_to_txq[TC_MAX_QUEUE];
|
|
|
|
u8 prio_tc_map[TC_BITMASK + 1];
|
2008-11-21 12:52:10 +08:00
|
|
|
|
2011-11-25 22:40:26 +08:00
|
|
|
#if IS_ENABLED(CONFIG_FCOE)
|
2009-02-28 06:06:59 +08:00
|
|
|
unsigned int fcoe_ddp_xid;
|
2011-11-22 13:10:51 +08:00
|
|
|
#endif
|
2013-12-30 00:27:11 +08:00
|
|
|
#if IS_ENABLED(CONFIG_CGROUP_NET_PRIO)
|
2011-11-22 13:10:51 +08:00
|
|
|
struct netprio_map __rcu *priomap;
|
2009-02-28 06:06:59 +08:00
|
|
|
#endif
|
2010-07-17 16:49:36 +08:00
|
|
|
struct phy_device *phydev;
|
2012-09-05 09:02:56 +08:00
|
|
|
struct lock_class_key *qdisc_tx_busylock;
|
2015-07-15 04:43:19 +08:00
|
|
|
bool proto_down;
|
2005-04-17 06:20:36 +08:00
|
|
|
};
|
2002-04-10 03:14:34 +08:00
|
|
|
#define to_net_dev(d) container_of(d, struct net_device, dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
#define NETDEV_ALIGN 32
|
|
|
|
|
2011-01-17 16:06:04 +08:00
|
|
|
static inline
|
|
|
|
int netdev_get_prio_tc_map(const struct net_device *dev, u32 prio)
|
|
|
|
{
|
|
|
|
return dev->prio_tc_map[prio & TC_BITMASK];
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline
|
|
|
|
int netdev_set_prio_tc_map(struct net_device *dev, u8 prio, u8 tc)
|
|
|
|
{
|
|
|
|
if (tc >= dev->num_tc)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
dev->prio_tc_map[prio & TC_BITMASK] = tc & TC_BITMASK;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline
|
|
|
|
void netdev_reset_tc(struct net_device *dev)
|
|
|
|
{
|
|
|
|
dev->num_tc = 0;
|
|
|
|
memset(dev->tc_to_txq, 0, sizeof(dev->tc_to_txq));
|
|
|
|
memset(dev->prio_tc_map, 0, sizeof(dev->prio_tc_map));
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline
|
|
|
|
int netdev_set_tc_queue(struct net_device *dev, u8 tc, u16 count, u16 offset)
|
|
|
|
{
|
|
|
|
if (tc >= dev->num_tc)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
dev->tc_to_txq[tc].count = count;
|
|
|
|
dev->tc_to_txq[tc].offset = offset;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline
|
|
|
|
int netdev_set_num_tc(struct net_device *dev, u8 num_tc)
|
|
|
|
{
|
|
|
|
if (num_tc > TC_MAX_QUEUE)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
dev->num_tc = num_tc;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline
|
|
|
|
int netdev_get_num_tc(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->num_tc;
|
|
|
|
}
|
|
|
|
|
2008-07-17 15:34:19 +08:00
|
|
|
static inline
|
|
|
|
struct netdev_queue *netdev_get_tx_queue(const struct net_device *dev,
|
|
|
|
unsigned int index)
|
|
|
|
{
|
|
|
|
return &dev->_tx[index];
|
|
|
|
}
|
|
|
|
|
2014-08-27 17:11:27 +08:00
|
|
|
static inline struct netdev_queue *skb_get_tx_queue(const struct net_device *dev,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
return netdev_get_tx_queue(dev, skb_get_queue_mapping(skb));
|
|
|
|
}
|
|
|
|
|
2008-07-17 15:34:19 +08:00
|
|
|
static inline void netdev_for_each_tx_queue(struct net_device *dev,
|
|
|
|
void (*f)(struct net_device *,
|
|
|
|
struct netdev_queue *,
|
|
|
|
void *),
|
|
|
|
void *arg)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
for (i = 0; i < dev->num_tx_queues; i++)
|
|
|
|
f(dev, &dev->_tx[i], arg);
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
struct netdev_queue *netdev_pick_tx(struct net_device *dev,
|
2014-01-10 16:18:26 +08:00
|
|
|
struct sk_buff *skb,
|
|
|
|
void *accel_priv);
|
2012-09-18 04:16:31 +08:00
|
|
|
|
2008-03-25 20:47:49 +08:00
|
|
|
/*
|
|
|
|
* Net namespace inlines
|
|
|
|
*/
|
|
|
|
static inline
|
|
|
|
struct net *dev_net(const struct net_device *dev)
|
|
|
|
{
|
2010-06-01 14:51:19 +08:00
|
|
|
return read_pnet(&dev->nd_net);
|
2008-03-25 20:47:49 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline
|
2008-03-26 15:48:17 +08:00
|
|
|
void dev_net_set(struct net_device *dev, struct net *net)
|
2008-03-25 20:47:49 +08:00
|
|
|
{
|
2015-03-12 12:06:44 +08:00
|
|
|
write_pnet(&dev->nd_net, net);
|
2008-03-25 20:47:49 +08:00
|
|
|
}
|
|
|
|
|
2014-08-28 08:04:46 +08:00
|
|
|
static inline bool netdev_uses_dsa(struct net_device *dev)
|
2008-10-07 21:45:02 +08:00
|
|
|
{
|
2014-09-13 14:12:46 +08:00
|
|
|
#if IS_ENABLED(CONFIG_NET_DSA)
|
2014-08-28 08:04:52 +08:00
|
|
|
if (dev->dsa_ptr != NULL)
|
|
|
|
return dsa_uses_tagged_protocol(dev->dsa_ptr);
|
2008-10-07 21:46:07 +08:00
|
|
|
#endif
|
2014-08-28 08:04:52 +08:00
|
|
|
return false;
|
2008-10-07 21:46:07 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netdev_priv - access network device private data
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Get network device private data
|
|
|
|
*/
|
2007-06-14 03:03:21 +08:00
|
|
|
static inline void *netdev_priv(const struct net_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2009-05-27 12:42:37 +08:00
|
|
|
return (char *)dev + ALIGN(sizeof(struct net_device), NETDEV_ALIGN);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Set the sysfs physical device reference for the network logical device
|
|
|
|
* if set prior to registration will cause a symlink during initialization.
|
|
|
|
*/
|
2002-04-10 03:14:34 +08:00
|
|
|
#define SET_NETDEV_DEV(net, pdev) ((net)->dev.parent = (pdev))
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-09-01 05:08:19 +08:00
|
|
|
/* Set the sysfs device type for the network logical device to allow
|
2013-10-12 07:29:46 +08:00
|
|
|
* fine-grained identification of different network device types. For
|
2009-09-01 05:08:19 +08:00
|
|
|
* example Ethernet, Wirelss LAN, Bluetooth, WiMAX etc.
|
|
|
|
*/
|
|
|
|
#define SET_NETDEV_DEVTYPE(net, devtype) ((net)->dev.type = (devtype))
|
|
|
|
|
2013-03-05 23:57:22 +08:00
|
|
|
/* Default NAPI poll() weight
|
|
|
|
* Device drivers are strongly advised to not use bigger value
|
|
|
|
*/
|
|
|
|
#define NAPI_POLL_WEIGHT 64
|
|
|
|
|
2007-11-01 17:21:47 +08:00
|
|
|
/**
|
|
|
|
* netif_napi_add - initialize a napi context
|
|
|
|
* @dev: network device
|
|
|
|
* @napi: napi context
|
|
|
|
* @poll: polling function
|
|
|
|
* @weight: default weight
|
|
|
|
*
|
|
|
|
* netif_napi_add() must be used to initialize a napi context prior to calling
|
|
|
|
* *any* of the other napi related functions.
|
|
|
|
*/
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
|
|
|
|
int (*poll)(struct napi_struct *, int), int weight);
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
|
2008-07-09 06:13:05 +08:00
|
|
|
/**
|
|
|
|
* netif_napi_del - remove a napi context
|
|
|
|
* @napi: napi context
|
|
|
|
*
|
|
|
|
* netif_napi_del() removes a napi context from the network device napi list
|
|
|
|
*/
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
void netif_napi_del(struct napi_struct *napi);
|
|
|
|
|
|
|
|
struct napi_gro_cb {
|
2009-05-27 02:50:21 +08:00
|
|
|
/* Virtual address of skb_shinfo(skb)->frags[0].page + offset. */
|
|
|
|
void *frag0;
|
|
|
|
|
2009-05-27 02:50:27 +08:00
|
|
|
/* Length of frag0. */
|
|
|
|
unsigned int frag0_len;
|
|
|
|
|
2009-01-29 22:19:50 +08:00
|
|
|
/* This indicates where we are processing relative to skb->data. */
|
|
|
|
int data_offset;
|
|
|
|
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
/* This is non-zero if the packet cannot be merged with the new skb. */
|
net-gre-gro: Add GRE support to the GRO stack
This patch built on top of Commit 299603e8370a93dd5d8e8d800f0dff1ce2c53d36
("net-gro: Prepare GRO stack for the upcoming tunneling support") to add
the support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO
stack. It also serves as an example for supporting other encapsulation
protocols in the GRO stack in the future.
The patch supports version 0 and all the flags (key, csum, seq#) but
will flush any pkt with the S (seq#) flag. This is because the S flag
is not support by GSO, and a GRO pkt may end up in the forwarding path,
thus requiring GSO support to break it up correctly.
Currently the "packet_offload" structure only contains L3 (ETH_P_IP/
ETH_P_IPV6) GRO offload support so the encapped pkts are limited to
IP pkts (i.e., w/o L2 hdr). But support for other protocol type can
be easily added, so is the support for GRE variations like NVGRE.
The patch also support csum offload. Specifically if the csum flag is on
and the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),
the code will take advantage of the csum computed by the h/w when
validating the GRE csum.
Note that commit 60769a5dcd8755715c7143b4571d5c44f01796f1 "ipv4: gre:
add GRO capability" already introduces GRO capability to IPv4 GRE
tunnels, using the gro_cells infrastructure. But GRO is done after
GRE hdr has been removed (i.e., decapped). The following patch applies
GRO when pkts first come in (before hitting the GRE tunnel code). There
is some performance advantage for applying GRO as early as possible.
Also this approach is transparent to other subsystem like Open vSwitch
where GRE decap is handled outside of the IP stack hence making it
harder for the gro_cells stuff to apply. On the other hand, some NICs
are still not capable of hashing on the inner hdr of a GRE pkt (RSS).
In that case the GRO processing of pkts from the same remote host will
all happen on the same CPU and the performance may be suboptimal.
I'm including some rough preliminary performance numbers below. Note
that the performance will be highly dependent on traffic load, mix as
usual. Moreover it also depends on NIC offload features hence the
following is by no means a comprehesive study. Local testing and tuning
will be needed to decide the best setting.
All tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.
(super_netperf 50 -H 192.168.1.18 -l 30)
An IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1
mode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)
is configured.
The GRO support for pkts AFTER decap are controlled through the device
feature of the GRE device (e.g., ethtool -K gre1 gro on/off).
1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off
thruput: 9.16Gbps
CPU utilization: 19%
1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off
thruput: 5.9Gbps
CPU utilization: 15%
1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 12-13%
1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 10%
The following tests were performed on a different NIC that is capable of
csum offload. I.e., the h/w is capable of computing IP payload csum
(CHECKSUM_COMPLETE).
2.1 ethtool -K gre1 gro on (hence will use gro_cells)
2.1.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 8.53Gbps
CPU utilization: 9%
2.1.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 8.97Gbps
CPU utilization: 7-8%
2.1.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 8.83Gbps
CPU utilization: 5-6%
2.1.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.98Gbps
CPU utilization: 5%
2.2 ethtool -K gre1 gro off
2.2.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 5.93Gbps
CPU utilization: 9%
2.2.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 5.62Gbps
CPU utilization: 8%
2.2.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 7.69Gbps
CPU utilization: 8%
2.2.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.96Gbps
CPU utilization: 5-6%
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-01-08 02:23:19 +08:00
|
|
|
u16 flush;
|
|
|
|
|
|
|
|
/* Save the IP ID here and check when we get to the transport layer */
|
|
|
|
u16 flush_id;
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
|
|
|
|
/* Number of segments aggregated. */
|
2012-10-06 16:08:49 +08:00
|
|
|
u16 count;
|
|
|
|
|
2015-02-11 08:30:31 +08:00
|
|
|
/* Start offset for remote checksum offload */
|
|
|
|
u16 gro_remcsum_start;
|
|
|
|
|
2012-10-06 16:08:49 +08:00
|
|
|
/* jiffies when first packet was created/queued */
|
|
|
|
unsigned long age;
|
2012-10-09 03:38:50 +08:00
|
|
|
|
2014-09-18 03:25:57 +08:00
|
|
|
/* Used in ipv6_gro_receive() and foo-over-udp */
|
2014-01-20 19:59:19 +08:00
|
|
|
u16 proto;
|
|
|
|
|
2015-02-11 08:30:30 +08:00
|
|
|
/* This is non-zero if the packet may be of the same flow. */
|
|
|
|
u8 same_flow:1;
|
|
|
|
|
2014-01-20 19:59:19 +08:00
|
|
|
/* Used in udp_gro_receive */
|
2014-08-23 04:33:47 +08:00
|
|
|
u8 udp_mark:1;
|
|
|
|
|
|
|
|
/* GRO checksum is valid */
|
|
|
|
u8 csum_valid:1;
|
|
|
|
|
2014-08-28 12:26:56 +08:00
|
|
|
/* Number of checksums via CHECKSUM_UNNECESSARY */
|
|
|
|
u8 csum_cnt:3;
|
2012-12-06 21:54:59 +08:00
|
|
|
|
2015-02-11 08:30:30 +08:00
|
|
|
/* Free the skb? */
|
|
|
|
u8 free:2;
|
|
|
|
#define NAPI_GRO_FREE 1
|
|
|
|
#define NAPI_GRO_FREE_STOLEN_HEAD 2
|
|
|
|
|
2014-10-04 06:48:08 +08:00
|
|
|
/* Used in foo-over-udp, set in udp[46]_gro_receive */
|
|
|
|
u8 is_ipv6:1;
|
|
|
|
|
2015-02-11 08:30:30 +08:00
|
|
|
/* 7 bit hole */
|
|
|
|
|
net-gre-gro: Add GRE support to the GRO stack
This patch built on top of Commit 299603e8370a93dd5d8e8d800f0dff1ce2c53d36
("net-gro: Prepare GRO stack for the upcoming tunneling support") to add
the support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO
stack. It also serves as an example for supporting other encapsulation
protocols in the GRO stack in the future.
The patch supports version 0 and all the flags (key, csum, seq#) but
will flush any pkt with the S (seq#) flag. This is because the S flag
is not support by GSO, and a GRO pkt may end up in the forwarding path,
thus requiring GSO support to break it up correctly.
Currently the "packet_offload" structure only contains L3 (ETH_P_IP/
ETH_P_IPV6) GRO offload support so the encapped pkts are limited to
IP pkts (i.e., w/o L2 hdr). But support for other protocol type can
be easily added, so is the support for GRE variations like NVGRE.
The patch also support csum offload. Specifically if the csum flag is on
and the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),
the code will take advantage of the csum computed by the h/w when
validating the GRE csum.
Note that commit 60769a5dcd8755715c7143b4571d5c44f01796f1 "ipv4: gre:
add GRO capability" already introduces GRO capability to IPv4 GRE
tunnels, using the gro_cells infrastructure. But GRO is done after
GRE hdr has been removed (i.e., decapped). The following patch applies
GRO when pkts first come in (before hitting the GRE tunnel code). There
is some performance advantage for applying GRO as early as possible.
Also this approach is transparent to other subsystem like Open vSwitch
where GRE decap is handled outside of the IP stack hence making it
harder for the gro_cells stuff to apply. On the other hand, some NICs
are still not capable of hashing on the inner hdr of a GRE pkt (RSS).
In that case the GRO processing of pkts from the same remote host will
all happen on the same CPU and the performance may be suboptimal.
I'm including some rough preliminary performance numbers below. Note
that the performance will be highly dependent on traffic load, mix as
usual. Moreover it also depends on NIC offload features hence the
following is by no means a comprehesive study. Local testing and tuning
will be needed to decide the best setting.
All tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.
(super_netperf 50 -H 192.168.1.18 -l 30)
An IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1
mode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)
is configured.
The GRO support for pkts AFTER decap are controlled through the device
feature of the GRE device (e.g., ethtool -K gre1 gro on/off).
1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off
thruput: 9.16Gbps
CPU utilization: 19%
1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off
thruput: 5.9Gbps
CPU utilization: 15%
1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 12-13%
1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 10%
The following tests were performed on a different NIC that is capable of
csum offload. I.e., the h/w is capable of computing IP payload csum
(CHECKSUM_COMPLETE).
2.1 ethtool -K gre1 gro on (hence will use gro_cells)
2.1.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 8.53Gbps
CPU utilization: 9%
2.1.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 8.97Gbps
CPU utilization: 7-8%
2.1.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 8.83Gbps
CPU utilization: 5-6%
2.1.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.98Gbps
CPU utilization: 5%
2.2 ethtool -K gre1 gro off
2.2.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 5.93Gbps
CPU utilization: 9%
2.2.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 5.62Gbps
CPU utilization: 8%
2.2.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 7.69Gbps
CPU utilization: 8%
2.2.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.96Gbps
CPU utilization: 5-6%
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-01-08 02:23:19 +08:00
|
|
|
/* used to support CHECKSUM_COMPLETE for tunneling protocols */
|
|
|
|
__wsum csum;
|
|
|
|
|
2012-12-06 21:54:59 +08:00
|
|
|
/* used in skb_gro_receive() slow path */
|
|
|
|
struct sk_buff *last;
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
#define NAPI_GRO_CB(skb) ((struct napi_gro_cb *)(skb)->cb)
|
2008-07-09 06:13:05 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
struct packet_type {
|
2005-08-10 10:34:12 +08:00
|
|
|
__be16 type; /* This is really htons(ether_type). */
|
|
|
|
struct net_device *dev; /* NULL is wildcarded here */
|
|
|
|
int (*func) (struct sk_buff *,
|
|
|
|
struct net_device *,
|
|
|
|
struct packet_type *,
|
|
|
|
struct net_device *);
|
2012-08-17 06:02:58 +08:00
|
|
|
bool (*id_match)(struct packet_type *ptype,
|
|
|
|
struct sock *sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
void *af_packet_priv;
|
|
|
|
struct list_head list;
|
|
|
|
};
|
|
|
|
|
2012-11-15 16:49:23 +08:00
|
|
|
struct offload_callbacks {
|
2006-06-28 04:22:38 +08:00
|
|
|
struct sk_buff *(*gso_segment)(struct sk_buff *skb,
|
2011-11-15 23:29:55 +08:00
|
|
|
netdev_features_t features);
|
net: Add Generic Receive Offload infrastructure
This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.
For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically. When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.
Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.
In addition to the packet, gro_receive will get a list of currently held
packets. Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet. For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.
Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it. In this case gro_receive should return the pointer to
the existing skb in gro_list. Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.
Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow. Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.
If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.
Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.
Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries. In future, this may be
expanded to use a hash table to allow more flows to be held for merging.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-16 15:38:52 +08:00
|
|
|
struct sk_buff **(*gro_receive)(struct sk_buff **head,
|
2015-01-13 09:00:37 +08:00
|
|
|
struct sk_buff *skb);
|
net-gro: Prepare GRO stack for the upcoming tunneling support
This patch modifies the GRO stack to avoid the use of "network_header"
and associated macros like ip_hdr() and ipv6_hdr() in order to allow
an arbitary number of IP hdrs (v4 or v6) to be used in the
encapsulation chain. This lays the foundation for various IP
tunneling support (IP-in-IP, GRE, VXLAN, SIT,...) to be added later.
With this patch, the GRO stack traversing now is mostly based on
skb_gro_offset rather than special hdr offsets saved in skb (e.g.,
skb->network_header). As a result all but the top layer (i.e., the
the transport layer) must have hdrs of the same length in order for
a pkt to be considered for aggregation. Therefore when adding a new
encap layer (e.g., for tunneling), one must check and skip flows
(e.g., by setting NAPI_GRO_CB(p)->same_flow to 0) that have a
different hdr length.
Note that unlike the network header, the transport header can and
will continue to be set by the GRO code since there will be at
most one "transport layer" in the encap chain.
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Suggested-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 12:53:45 +08:00
|
|
|
int (*gro_complete)(struct sk_buff *skb, int nhoff);
|
2012-11-15 16:49:23 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
struct packet_offload {
|
|
|
|
__be16 type; /* This is really htons(ether_type). */
|
2015-06-02 05:56:09 +08:00
|
|
|
u16 priority;
|
2012-11-15 16:49:23 +08:00
|
|
|
struct offload_callbacks callbacks;
|
|
|
|
struct list_head list;
|
2005-04-17 06:20:36 +08:00
|
|
|
};
|
|
|
|
|
2015-01-13 09:00:37 +08:00
|
|
|
struct udp_offload;
|
|
|
|
|
|
|
|
struct udp_offload_callbacks {
|
|
|
|
struct sk_buff **(*gro_receive)(struct sk_buff **head,
|
|
|
|
struct sk_buff *skb,
|
|
|
|
struct udp_offload *uoff);
|
|
|
|
int (*gro_complete)(struct sk_buff *skb,
|
|
|
|
int nhoff,
|
|
|
|
struct udp_offload *uoff);
|
|
|
|
};
|
|
|
|
|
2014-01-20 19:59:19 +08:00
|
|
|
struct udp_offload {
|
|
|
|
__be16 port;
|
2014-09-18 03:25:57 +08:00
|
|
|
u8 ipproto;
|
2015-01-13 09:00:37 +08:00
|
|
|
struct udp_offload_callbacks callbacks;
|
2014-01-20 19:59:19 +08:00
|
|
|
};
|
|
|
|
|
2014-01-04 13:57:59 +08:00
|
|
|
/* often modified stats are per cpu, other are shared (netdev->stats) */
|
|
|
|
struct pcpu_sw_netstats {
|
|
|
|
u64 rx_packets;
|
|
|
|
u64 rx_bytes;
|
|
|
|
u64 tx_packets;
|
|
|
|
u64 tx_bytes;
|
|
|
|
struct u64_stats_sync syncp;
|
|
|
|
};
|
|
|
|
|
2014-02-14 03:46:28 +08:00
|
|
|
#define netdev_alloc_pcpu_stats(type) \
|
|
|
|
({ \
|
2014-03-11 00:41:46 +08:00
|
|
|
typeof(type) __percpu *pcpu_stats = alloc_percpu(type); \
|
2014-02-14 03:46:28 +08:00
|
|
|
if (pcpu_stats) { \
|
2015-04-23 18:06:30 +08:00
|
|
|
int __cpu; \
|
|
|
|
for_each_possible_cpu(__cpu) { \
|
2014-02-14 03:46:28 +08:00
|
|
|
typeof(type) *stat; \
|
2015-04-23 18:06:30 +08:00
|
|
|
stat = per_cpu_ptr(pcpu_stats, __cpu); \
|
2014-02-14 03:46:28 +08:00
|
|
|
u64_stats_init(&stat->syncp); \
|
|
|
|
} \
|
|
|
|
} \
|
|
|
|
pcpu_stats; \
|
|
|
|
})
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <linux/notifier.h>
|
|
|
|
|
2011-07-26 08:13:09 +08:00
|
|
|
/* netdevice notifier chain. Please remember to update the rtnetlink
|
|
|
|
* notification exclusion list in rtnetlink_event() when adding new
|
|
|
|
* types.
|
|
|
|
*/
|
|
|
|
#define NETDEV_UP 0x0001 /* For now you can't veto a device up/down */
|
|
|
|
#define NETDEV_DOWN 0x0002
|
|
|
|
#define NETDEV_REBOOT 0x0003 /* Tell a protocol stack a network interface
|
|
|
|
detected a hardware crash and restarted
|
|
|
|
- we can use this eg to kick tcp sessions
|
|
|
|
once done */
|
|
|
|
#define NETDEV_CHANGE 0x0004 /* Notify device state change */
|
|
|
|
#define NETDEV_REGISTER 0x0005
|
|
|
|
#define NETDEV_UNREGISTER 0x0006
|
2014-01-16 07:02:18 +08:00
|
|
|
#define NETDEV_CHANGEMTU 0x0007 /* notify after mtu change happened */
|
2011-07-26 08:13:09 +08:00
|
|
|
#define NETDEV_CHANGEADDR 0x0008
|
|
|
|
#define NETDEV_GOING_DOWN 0x0009
|
|
|
|
#define NETDEV_CHANGENAME 0x000A
|
|
|
|
#define NETDEV_FEAT_CHANGE 0x000B
|
|
|
|
#define NETDEV_BONDING_FAILOVER 0x000C
|
|
|
|
#define NETDEV_PRE_UP 0x000D
|
|
|
|
#define NETDEV_PRE_TYPE_CHANGE 0x000E
|
|
|
|
#define NETDEV_POST_TYPE_CHANGE 0x000F
|
|
|
|
#define NETDEV_POST_INIT 0x0010
|
2012-08-23 01:19:46 +08:00
|
|
|
#define NETDEV_UNREGISTER_FINAL 0x0011
|
2011-07-26 08:13:09 +08:00
|
|
|
#define NETDEV_RELEASE 0x0012
|
|
|
|
#define NETDEV_NOTIFY_PEERS 0x0013
|
|
|
|
#define NETDEV_JOIN 0x0014
|
2013-05-25 12:12:10 +08:00
|
|
|
#define NETDEV_CHANGEUPPER 0x0015
|
2013-07-20 18:13:53 +08:00
|
|
|
#define NETDEV_RESEND_IGMP 0x0016
|
2014-01-16 07:02:18 +08:00
|
|
|
#define NETDEV_PRECHANGEMTU 0x0017 /* notify before mtu change happened */
|
2014-08-19 22:02:12 +08:00
|
|
|
#define NETDEV_CHANGEINFODATA 0x0018
|
2015-02-03 22:48:29 +08:00
|
|
|
#define NETDEV_BONDING_INFO 0x0019
|
2011-07-26 08:13:09 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int register_netdevice_notifier(struct notifier_block *nb);
|
|
|
|
int unregister_netdevice_notifier(struct notifier_block *nb);
|
2013-05-28 09:30:21 +08:00
|
|
|
|
|
|
|
struct netdev_notifier_info {
|
|
|
|
struct net_device *dev;
|
|
|
|
};
|
|
|
|
|
2013-05-28 09:30:22 +08:00
|
|
|
struct netdev_notifier_change_info {
|
|
|
|
struct netdev_notifier_info info; /* must be first */
|
|
|
|
unsigned int flags_changed;
|
|
|
|
};
|
|
|
|
|
2015-08-27 15:31:18 +08:00
|
|
|
struct netdev_notifier_changeupper_info {
|
|
|
|
struct netdev_notifier_info info; /* must be first */
|
|
|
|
struct net_device *upper_dev; /* new upper dev */
|
|
|
|
bool master; /* is upper dev master */
|
|
|
|
bool linking; /* is the nofication for link or unlink */
|
|
|
|
};
|
|
|
|
|
2013-05-29 11:30:50 +08:00
|
|
|
static inline void netdev_notifier_info_init(struct netdev_notifier_info *info,
|
|
|
|
struct net_device *dev)
|
|
|
|
{
|
|
|
|
info->dev = dev;
|
|
|
|
}
|
|
|
|
|
2013-05-28 09:30:21 +08:00
|
|
|
static inline struct net_device *
|
|
|
|
netdev_notifier_info_to_dev(const struct netdev_notifier_info *info)
|
|
|
|
{
|
|
|
|
return info->dev;
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int call_netdevice_notifiers(unsigned long val, struct net_device *dev);
|
2011-07-26 08:13:09 +08:00
|
|
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
extern rwlock_t dev_base_lock; /* Device list lock */
|
|
|
|
|
2007-09-18 02:56:21 +08:00
|
|
|
#define for_each_netdev(net, d) \
|
|
|
|
list_for_each_entry(d, &(net)->dev_base_head, dev_list)
|
2009-11-30 06:25:26 +08:00
|
|
|
#define for_each_netdev_reverse(net, d) \
|
|
|
|
list_for_each_entry_reverse(d, &(net)->dev_base_head, dev_list)
|
2009-11-04 21:43:23 +08:00
|
|
|
#define for_each_netdev_rcu(net, d) \
|
|
|
|
list_for_each_entry_rcu(d, &(net)->dev_base_head, dev_list)
|
2007-09-18 02:56:21 +08:00
|
|
|
#define for_each_netdev_safe(net, d, n) \
|
|
|
|
list_for_each_entry_safe(d, n, &(net)->dev_base_head, dev_list)
|
|
|
|
#define for_each_netdev_continue(net, d) \
|
|
|
|
list_for_each_entry_continue(d, &(net)->dev_base_head, dev_list)
|
2009-11-10 15:54:47 +08:00
|
|
|
#define for_each_netdev_continue_rcu(net, d) \
|
|
|
|
list_for_each_entry_continue_rcu(d, &(net)->dev_base_head, dev_list)
|
2013-03-12 10:49:01 +08:00
|
|
|
#define for_each_netdev_in_bond_rcu(bond, slave) \
|
|
|
|
for_each_netdev_rcu(&init_net, slave) \
|
2015-01-14 15:52:35 +08:00
|
|
|
if (netdev_master_upper_dev_get_rcu(slave) == (bond))
|
2007-09-18 02:56:21 +08:00
|
|
|
#define net_device_entry(lh) list_entry(lh, struct net_device, dev_list)
|
2007-05-04 06:13:45 +08:00
|
|
|
|
2007-09-12 20:57:09 +08:00
|
|
|
static inline struct net_device *next_net_device(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct list_head *lh;
|
|
|
|
struct net *net;
|
|
|
|
|
2008-03-25 20:47:49 +08:00
|
|
|
net = dev_net(dev);
|
2007-09-12 20:57:09 +08:00
|
|
|
lh = dev->dev_list.next;
|
|
|
|
return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
|
|
|
|
}
|
|
|
|
|
2009-11-12 01:34:30 +08:00
|
|
|
static inline struct net_device *next_net_device_rcu(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct list_head *lh;
|
|
|
|
struct net *net;
|
|
|
|
|
|
|
|
net = dev_net(dev);
|
2011-01-27 02:08:02 +08:00
|
|
|
lh = rcu_dereference(list_next_rcu(&dev->dev_list));
|
2009-11-12 01:34:30 +08:00
|
|
|
return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
|
|
|
|
}
|
|
|
|
|
2007-09-12 20:57:09 +08:00
|
|
|
static inline struct net_device *first_net_device(struct net *net)
|
|
|
|
{
|
|
|
|
return list_empty(&net->dev_base_head) ? NULL :
|
|
|
|
net_device_entry(net->dev_base_head.next);
|
|
|
|
}
|
2007-05-04 06:13:45 +08:00
|
|
|
|
2011-01-27 02:08:02 +08:00
|
|
|
static inline struct net_device *first_net_device_rcu(struct net *net)
|
|
|
|
{
|
|
|
|
struct list_head *lh = rcu_dereference(list_next_rcu(&net->dev_base_head));
|
|
|
|
|
|
|
|
return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int netdev_boot_setup_check(struct net_device *dev);
|
|
|
|
unsigned long netdev_boot_base(const char *prefix, int unit);
|
|
|
|
struct net_device *dev_getbyhwaddr_rcu(struct net *net, unsigned short type,
|
|
|
|
const char *hwaddr);
|
|
|
|
struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type);
|
|
|
|
struct net_device *__dev_getfirstbyhwtype(struct net *net, unsigned short type);
|
|
|
|
void dev_add_pack(struct packet_type *pt);
|
|
|
|
void dev_remove_pack(struct packet_type *pt);
|
|
|
|
void __dev_remove_pack(struct packet_type *pt);
|
|
|
|
void dev_add_offload(struct packet_offload *po);
|
|
|
|
void dev_remove_offload(struct packet_offload *po);
|
|
|
|
|
2015-04-02 23:07:00 +08:00
|
|
|
int dev_get_iflink(const struct net_device *dev);
|
2014-09-12 06:35:09 +08:00
|
|
|
struct net_device *__dev_get_by_flags(struct net *net, unsigned short flags,
|
|
|
|
unsigned short mask);
|
2013-09-27 05:48:15 +08:00
|
|
|
struct net_device *dev_get_by_name(struct net *net, const char *name);
|
|
|
|
struct net_device *dev_get_by_name_rcu(struct net *net, const char *name);
|
|
|
|
struct net_device *__dev_get_by_name(struct net *net, const char *name);
|
|
|
|
int dev_alloc_name(struct net_device *dev, const char *name);
|
|
|
|
int dev_open(struct net_device *dev);
|
|
|
|
int dev_close(struct net_device *dev);
|
2015-03-19 10:52:33 +08:00
|
|
|
int dev_close_many(struct list_head *head, bool unlink);
|
2013-09-27 05:48:15 +08:00
|
|
|
void dev_disable_lro(struct net_device *dev);
|
2015-04-06 10:19:04 +08:00
|
|
|
int dev_loopback_xmit(struct sock *sk, struct sk_buff *newskb);
|
|
|
|
int dev_queue_xmit_sk(struct sock *sk, struct sk_buff *skb);
|
|
|
|
static inline int dev_queue_xmit(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
return dev_queue_xmit_sk(skb->sk, skb);
|
|
|
|
}
|
2014-01-10 16:18:26 +08:00
|
|
|
int dev_queue_xmit_accel(struct sk_buff *skb, void *accel_priv);
|
2013-09-27 05:48:15 +08:00
|
|
|
int register_netdevice(struct net_device *dev);
|
|
|
|
void unregister_netdevice_queue(struct net_device *dev, struct list_head *head);
|
|
|
|
void unregister_netdevice_many(struct list_head *head);
|
2009-10-27 15:03:04 +08:00
|
|
|
static inline void unregister_netdevice(struct net_device *dev)
|
|
|
|
{
|
|
|
|
unregister_netdevice_queue(dev, NULL);
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int netdev_refcnt_read(const struct net_device *dev);
|
|
|
|
void free_netdev(struct net_device *dev);
|
2013-10-31 04:10:44 +08:00
|
|
|
void netdev_freemem(struct net_device *dev);
|
2013-09-27 05:48:15 +08:00
|
|
|
void synchronize_net(void);
|
|
|
|
int init_dummy_netdev(struct net_device *dev);
|
2009-01-15 13:05:05 +08:00
|
|
|
|
2015-04-01 23:07:44 +08:00
|
|
|
DECLARE_PER_CPU(int, xmit_recursion);
|
|
|
|
static inline int dev_recursion_level(void)
|
|
|
|
{
|
|
|
|
return this_cpu_read(xmit_recursion);
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
struct net_device *dev_get_by_index(struct net *net, int ifindex);
|
|
|
|
struct net_device *__dev_get_by_index(struct net *net, int ifindex);
|
|
|
|
struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex);
|
|
|
|
int netdev_get_name(struct net *net, char *name, int ifindex);
|
|
|
|
int dev_restart(struct net_device *dev);
|
|
|
|
int skb_gro_receive(struct sk_buff **head, struct sk_buff *skb);
|
2009-01-29 22:19:50 +08:00
|
|
|
|
|
|
|
static inline unsigned int skb_gro_offset(const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
return NAPI_GRO_CB(skb)->data_offset;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline unsigned int skb_gro_len(const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
return skb->len - NAPI_GRO_CB(skb)->data_offset;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void skb_gro_pull(struct sk_buff *skb, unsigned int len)
|
|
|
|
{
|
|
|
|
NAPI_GRO_CB(skb)->data_offset += len;
|
|
|
|
}
|
|
|
|
|
2009-05-27 02:50:28 +08:00
|
|
|
static inline void *skb_gro_header_fast(struct sk_buff *skb,
|
|
|
|
unsigned int offset)
|
2009-01-29 22:19:50 +08:00
|
|
|
{
|
2009-05-27 02:50:28 +08:00
|
|
|
return NAPI_GRO_CB(skb)->frag0 + offset;
|
|
|
|
}
|
2009-05-27 02:50:21 +08:00
|
|
|
|
2009-05-27 02:50:28 +08:00
|
|
|
static inline int skb_gro_header_hard(struct sk_buff *skb, unsigned int hlen)
|
|
|
|
{
|
|
|
|
return NAPI_GRO_CB(skb)->frag0_len < hlen;
|
|
|
|
}
|
2009-05-27 02:50:21 +08:00
|
|
|
|
2009-05-27 02:50:28 +08:00
|
|
|
static inline void *skb_gro_header_slow(struct sk_buff *skb, unsigned int hlen,
|
|
|
|
unsigned int offset)
|
|
|
|
{
|
2011-07-27 21:16:28 +08:00
|
|
|
if (!pskb_may_pull(skb, hlen))
|
|
|
|
return NULL;
|
|
|
|
|
2009-05-27 02:50:28 +08:00
|
|
|
NAPI_GRO_CB(skb)->frag0 = NULL;
|
|
|
|
NAPI_GRO_CB(skb)->frag0_len = 0;
|
2011-07-27 21:16:28 +08:00
|
|
|
return skb->data + offset;
|
2009-01-29 22:19:50 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-04-27 20:44:45 +08:00
|
|
|
static inline void *skb_gro_network_header(struct sk_buff *skb)
|
|
|
|
{
|
2009-05-27 02:50:23 +08:00
|
|
|
return (NAPI_GRO_CB(skb)->frag0 ?: skb->data) +
|
|
|
|
skb_network_offset(skb);
|
2009-04-27 20:44:45 +08:00
|
|
|
}
|
|
|
|
|
net-gre-gro: Add GRE support to the GRO stack
This patch built on top of Commit 299603e8370a93dd5d8e8d800f0dff1ce2c53d36
("net-gro: Prepare GRO stack for the upcoming tunneling support") to add
the support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO
stack. It also serves as an example for supporting other encapsulation
protocols in the GRO stack in the future.
The patch supports version 0 and all the flags (key, csum, seq#) but
will flush any pkt with the S (seq#) flag. This is because the S flag
is not support by GSO, and a GRO pkt may end up in the forwarding path,
thus requiring GSO support to break it up correctly.
Currently the "packet_offload" structure only contains L3 (ETH_P_IP/
ETH_P_IPV6) GRO offload support so the encapped pkts are limited to
IP pkts (i.e., w/o L2 hdr). But support for other protocol type can
be easily added, so is the support for GRE variations like NVGRE.
The patch also support csum offload. Specifically if the csum flag is on
and the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),
the code will take advantage of the csum computed by the h/w when
validating the GRE csum.
Note that commit 60769a5dcd8755715c7143b4571d5c44f01796f1 "ipv4: gre:
add GRO capability" already introduces GRO capability to IPv4 GRE
tunnels, using the gro_cells infrastructure. But GRO is done after
GRE hdr has been removed (i.e., decapped). The following patch applies
GRO when pkts first come in (before hitting the GRE tunnel code). There
is some performance advantage for applying GRO as early as possible.
Also this approach is transparent to other subsystem like Open vSwitch
where GRE decap is handled outside of the IP stack hence making it
harder for the gro_cells stuff to apply. On the other hand, some NICs
are still not capable of hashing on the inner hdr of a GRE pkt (RSS).
In that case the GRO processing of pkts from the same remote host will
all happen on the same CPU and the performance may be suboptimal.
I'm including some rough preliminary performance numbers below. Note
that the performance will be highly dependent on traffic load, mix as
usual. Moreover it also depends on NIC offload features hence the
following is by no means a comprehesive study. Local testing and tuning
will be needed to decide the best setting.
All tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.
(super_netperf 50 -H 192.168.1.18 -l 30)
An IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1
mode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)
is configured.
The GRO support for pkts AFTER decap are controlled through the device
feature of the GRE device (e.g., ethtool -K gre1 gro on/off).
1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off
thruput: 9.16Gbps
CPU utilization: 19%
1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off
thruput: 5.9Gbps
CPU utilization: 15%
1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 12-13%
1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 10%
The following tests were performed on a different NIC that is capable of
csum offload. I.e., the h/w is capable of computing IP payload csum
(CHECKSUM_COMPLETE).
2.1 ethtool -K gre1 gro on (hence will use gro_cells)
2.1.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 8.53Gbps
CPU utilization: 9%
2.1.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 8.97Gbps
CPU utilization: 7-8%
2.1.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 8.83Gbps
CPU utilization: 5-6%
2.1.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.98Gbps
CPU utilization: 5%
2.2 ethtool -K gre1 gro off
2.2.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 5.93Gbps
CPU utilization: 9%
2.2.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 5.62Gbps
CPU utilization: 8%
2.2.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 7.69Gbps
CPU utilization: 8%
2.2.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.96Gbps
CPU utilization: 5-6%
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-01-08 02:23:19 +08:00
|
|
|
static inline void skb_gro_postpull_rcsum(struct sk_buff *skb,
|
|
|
|
const void *start, unsigned int len)
|
|
|
|
{
|
2014-08-23 04:33:47 +08:00
|
|
|
if (NAPI_GRO_CB(skb)->csum_valid)
|
net-gre-gro: Add GRE support to the GRO stack
This patch built on top of Commit 299603e8370a93dd5d8e8d800f0dff1ce2c53d36
("net-gro: Prepare GRO stack for the upcoming tunneling support") to add
the support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO
stack. It also serves as an example for supporting other encapsulation
protocols in the GRO stack in the future.
The patch supports version 0 and all the flags (key, csum, seq#) but
will flush any pkt with the S (seq#) flag. This is because the S flag
is not support by GSO, and a GRO pkt may end up in the forwarding path,
thus requiring GSO support to break it up correctly.
Currently the "packet_offload" structure only contains L3 (ETH_P_IP/
ETH_P_IPV6) GRO offload support so the encapped pkts are limited to
IP pkts (i.e., w/o L2 hdr). But support for other protocol type can
be easily added, so is the support for GRE variations like NVGRE.
The patch also support csum offload. Specifically if the csum flag is on
and the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),
the code will take advantage of the csum computed by the h/w when
validating the GRE csum.
Note that commit 60769a5dcd8755715c7143b4571d5c44f01796f1 "ipv4: gre:
add GRO capability" already introduces GRO capability to IPv4 GRE
tunnels, using the gro_cells infrastructure. But GRO is done after
GRE hdr has been removed (i.e., decapped). The following patch applies
GRO when pkts first come in (before hitting the GRE tunnel code). There
is some performance advantage for applying GRO as early as possible.
Also this approach is transparent to other subsystem like Open vSwitch
where GRE decap is handled outside of the IP stack hence making it
harder for the gro_cells stuff to apply. On the other hand, some NICs
are still not capable of hashing on the inner hdr of a GRE pkt (RSS).
In that case the GRO processing of pkts from the same remote host will
all happen on the same CPU and the performance may be suboptimal.
I'm including some rough preliminary performance numbers below. Note
that the performance will be highly dependent on traffic load, mix as
usual. Moreover it also depends on NIC offload features hence the
following is by no means a comprehesive study. Local testing and tuning
will be needed to decide the best setting.
All tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.
(super_netperf 50 -H 192.168.1.18 -l 30)
An IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1
mode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)
is configured.
The GRO support for pkts AFTER decap are controlled through the device
feature of the GRE device (e.g., ethtool -K gre1 gro on/off).
1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off
thruput: 9.16Gbps
CPU utilization: 19%
1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off
thruput: 5.9Gbps
CPU utilization: 15%
1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 12-13%
1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 10%
The following tests were performed on a different NIC that is capable of
csum offload. I.e., the h/w is capable of computing IP payload csum
(CHECKSUM_COMPLETE).
2.1 ethtool -K gre1 gro on (hence will use gro_cells)
2.1.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 8.53Gbps
CPU utilization: 9%
2.1.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 8.97Gbps
CPU utilization: 7-8%
2.1.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 8.83Gbps
CPU utilization: 5-6%
2.1.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.98Gbps
CPU utilization: 5%
2.2 ethtool -K gre1 gro off
2.2.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 5.93Gbps
CPU utilization: 9%
2.2.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 5.62Gbps
CPU utilization: 8%
2.2.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 7.69Gbps
CPU utilization: 8%
2.2.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.96Gbps
CPU utilization: 5-6%
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-01-08 02:23:19 +08:00
|
|
|
NAPI_GRO_CB(skb)->csum = csum_sub(NAPI_GRO_CB(skb)->csum,
|
|
|
|
csum_partial(start, len, 0));
|
|
|
|
}
|
|
|
|
|
2014-08-23 04:33:47 +08:00
|
|
|
/* GRO checksum functions. These are logical equivalents of the normal
|
|
|
|
* checksum functions (in skbuff.h) except that they operate on the GRO
|
|
|
|
* offsets and fields in sk_buff.
|
|
|
|
*/
|
|
|
|
|
|
|
|
__sum16 __skb_gro_checksum_complete(struct sk_buff *skb);
|
|
|
|
|
2015-02-11 08:30:31 +08:00
|
|
|
static inline bool skb_at_gro_remcsum_start(struct sk_buff *skb)
|
|
|
|
{
|
2015-08-20 08:07:32 +08:00
|
|
|
return (NAPI_GRO_CB(skb)->gro_remcsum_start == skb_gro_offset(skb));
|
2015-02-11 08:30:31 +08:00
|
|
|
}
|
|
|
|
|
2014-08-23 04:33:47 +08:00
|
|
|
static inline bool __skb_gro_checksum_validate_needed(struct sk_buff *skb,
|
|
|
|
bool zero_okay,
|
|
|
|
__sum16 check)
|
|
|
|
{
|
2015-02-11 08:30:28 +08:00
|
|
|
return ((skb->ip_summed != CHECKSUM_PARTIAL ||
|
|
|
|
skb_checksum_start_offset(skb) <
|
|
|
|
skb_gro_offset(skb)) &&
|
2015-02-11 08:30:31 +08:00
|
|
|
!skb_at_gro_remcsum_start(skb) &&
|
2014-08-28 12:26:56 +08:00
|
|
|
NAPI_GRO_CB(skb)->csum_cnt == 0 &&
|
2014-08-23 04:33:47 +08:00
|
|
|
(!zero_okay || check));
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline __sum16 __skb_gro_checksum_validate_complete(struct sk_buff *skb,
|
|
|
|
__wsum psum)
|
|
|
|
{
|
|
|
|
if (NAPI_GRO_CB(skb)->csum_valid &&
|
|
|
|
!csum_fold(csum_add(psum, NAPI_GRO_CB(skb)->csum)))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
NAPI_GRO_CB(skb)->csum = psum;
|
|
|
|
|
|
|
|
return __skb_gro_checksum_complete(skb);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void skb_gro_incr_csum_unnecessary(struct sk_buff *skb)
|
|
|
|
{
|
2014-08-28 12:26:56 +08:00
|
|
|
if (NAPI_GRO_CB(skb)->csum_cnt > 0) {
|
|
|
|
/* Consume a checksum from CHECKSUM_UNNECESSARY */
|
|
|
|
NAPI_GRO_CB(skb)->csum_cnt--;
|
|
|
|
} else {
|
|
|
|
/* Update skb for CHECKSUM_UNNECESSARY and csum_level when we
|
|
|
|
* verified a new top level checksum or an encapsulated one
|
|
|
|
* during GRO. This saves work if we fallback to normal path.
|
|
|
|
*/
|
|
|
|
__skb_incr_checksum_unnecessary(skb);
|
2014-08-23 04:33:47 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#define __skb_gro_checksum_validate(skb, proto, zero_okay, check, \
|
|
|
|
compute_pseudo) \
|
|
|
|
({ \
|
|
|
|
__sum16 __ret = 0; \
|
|
|
|
if (__skb_gro_checksum_validate_needed(skb, zero_okay, check)) \
|
|
|
|
__ret = __skb_gro_checksum_validate_complete(skb, \
|
|
|
|
compute_pseudo(skb, proto)); \
|
2014-09-01 06:12:41 +08:00
|
|
|
if (__ret) \
|
|
|
|
__skb_mark_checksum_bad(skb); \
|
|
|
|
else \
|
2014-08-23 04:33:47 +08:00
|
|
|
skb_gro_incr_csum_unnecessary(skb); \
|
|
|
|
__ret; \
|
|
|
|
})
|
|
|
|
|
|
|
|
#define skb_gro_checksum_validate(skb, proto, compute_pseudo) \
|
|
|
|
__skb_gro_checksum_validate(skb, proto, false, 0, compute_pseudo)
|
|
|
|
|
|
|
|
#define skb_gro_checksum_validate_zero_check(skb, proto, check, \
|
|
|
|
compute_pseudo) \
|
|
|
|
__skb_gro_checksum_validate(skb, proto, true, check, compute_pseudo)
|
|
|
|
|
|
|
|
#define skb_gro_checksum_simple_validate(skb) \
|
|
|
|
__skb_gro_checksum_validate(skb, 0, false, 0, null_compute_pseudo)
|
|
|
|
|
2014-09-01 06:12:42 +08:00
|
|
|
static inline bool __skb_gro_checksum_convert_check(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
return (NAPI_GRO_CB(skb)->csum_cnt == 0 &&
|
|
|
|
!NAPI_GRO_CB(skb)->csum_valid);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void __skb_gro_checksum_convert(struct sk_buff *skb,
|
|
|
|
__sum16 check, __wsum pseudo)
|
|
|
|
{
|
|
|
|
NAPI_GRO_CB(skb)->csum = ~pseudo;
|
|
|
|
NAPI_GRO_CB(skb)->csum_valid = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
#define skb_gro_checksum_try_convert(skb, proto, check, compute_pseudo) \
|
|
|
|
do { \
|
|
|
|
if (__skb_gro_checksum_convert_check(skb)) \
|
|
|
|
__skb_gro_checksum_convert(skb, check, \
|
|
|
|
compute_pseudo(skb, proto)); \
|
|
|
|
} while (0)
|
|
|
|
|
2015-02-11 08:30:27 +08:00
|
|
|
struct gro_remcsum {
|
|
|
|
int offset;
|
|
|
|
__wsum delta;
|
|
|
|
};
|
|
|
|
|
|
|
|
static inline void skb_gro_remcsum_init(struct gro_remcsum *grc)
|
|
|
|
{
|
2015-02-18 18:38:06 +08:00
|
|
|
grc->offset = 0;
|
2015-02-11 08:30:27 +08:00
|
|
|
grc->delta = 0;
|
|
|
|
}
|
|
|
|
|
2015-08-20 08:07:32 +08:00
|
|
|
static inline void *skb_gro_remcsum_process(struct sk_buff *skb, void *ptr,
|
|
|
|
unsigned int off, size_t hdrlen,
|
|
|
|
int start, int offset,
|
|
|
|
struct gro_remcsum *grc,
|
|
|
|
bool nopartial)
|
2015-02-03 08:07:34 +08:00
|
|
|
{
|
|
|
|
__wsum delta;
|
2015-08-20 08:07:32 +08:00
|
|
|
size_t plen = hdrlen + max_t(size_t, offset + sizeof(u16), start);
|
2015-02-03 08:07:34 +08:00
|
|
|
|
|
|
|
BUG_ON(!NAPI_GRO_CB(skb)->csum_valid);
|
|
|
|
|
2015-02-11 08:30:31 +08:00
|
|
|
if (!nopartial) {
|
2015-08-20 08:07:32 +08:00
|
|
|
NAPI_GRO_CB(skb)->gro_remcsum_start = off + hdrlen + start;
|
|
|
|
return ptr;
|
|
|
|
}
|
|
|
|
|
|
|
|
ptr = skb_gro_header_fast(skb, off);
|
|
|
|
if (skb_gro_header_hard(skb, off + plen)) {
|
|
|
|
ptr = skb_gro_header_slow(skb, off + plen, off);
|
|
|
|
if (!ptr)
|
|
|
|
return NULL;
|
2015-02-11 08:30:31 +08:00
|
|
|
}
|
|
|
|
|
2015-08-20 08:07:32 +08:00
|
|
|
delta = remcsum_adjust(ptr + hdrlen, NAPI_GRO_CB(skb)->csum,
|
|
|
|
start, offset);
|
2015-02-03 08:07:34 +08:00
|
|
|
|
|
|
|
/* Adjust skb->csum since we changed the packet */
|
|
|
|
NAPI_GRO_CB(skb)->csum = csum_add(NAPI_GRO_CB(skb)->csum, delta);
|
2015-02-11 08:30:27 +08:00
|
|
|
|
2015-08-20 08:07:32 +08:00
|
|
|
grc->offset = off + hdrlen + offset;
|
2015-02-11 08:30:27 +08:00
|
|
|
grc->delta = delta;
|
2015-08-20 08:07:32 +08:00
|
|
|
|
|
|
|
return ptr;
|
2015-02-03 08:07:34 +08:00
|
|
|
}
|
|
|
|
|
2015-02-11 08:30:27 +08:00
|
|
|
static inline void skb_gro_remcsum_cleanup(struct sk_buff *skb,
|
|
|
|
struct gro_remcsum *grc)
|
|
|
|
{
|
2015-08-20 08:07:32 +08:00
|
|
|
void *ptr;
|
|
|
|
size_t plen = grc->offset + sizeof(u16);
|
|
|
|
|
2015-02-11 08:30:27 +08:00
|
|
|
if (!grc->delta)
|
|
|
|
return;
|
|
|
|
|
2015-08-20 08:07:32 +08:00
|
|
|
ptr = skb_gro_header_fast(skb, grc->offset);
|
|
|
|
if (skb_gro_header_hard(skb, grc->offset + sizeof(u16))) {
|
|
|
|
ptr = skb_gro_header_slow(skb, plen, grc->offset);
|
|
|
|
if (!ptr)
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
remcsum_unadjust((__sum16 *)ptr, grc->delta);
|
2015-02-11 08:30:27 +08:00
|
|
|
}
|
2015-02-03 08:07:34 +08:00
|
|
|
|
2007-10-09 16:36:32 +08:00
|
|
|
static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev,
|
|
|
|
unsigned short type,
|
2007-10-09 16:40:57 +08:00
|
|
|
const void *daddr, const void *saddr,
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int len)
|
2007-10-09 16:36:32 +08:00
|
|
|
{
|
2007-10-22 22:16:14 +08:00
|
|
|
if (!dev->header_ops || !dev->header_ops->create)
|
2007-10-09 16:36:32 +08:00
|
|
|
return 0;
|
2007-10-09 16:40:57 +08:00
|
|
|
|
|
|
|
return dev->header_ops->create(skb, dev, type, daddr, saddr, len);
|
2007-10-09 16:36:32 +08:00
|
|
|
}
|
|
|
|
|
2007-09-27 13:13:38 +08:00
|
|
|
static inline int dev_parse_header(const struct sk_buff *skb,
|
|
|
|
unsigned char *haddr)
|
|
|
|
{
|
|
|
|
const struct net_device *dev = skb->dev;
|
|
|
|
|
2007-10-18 20:09:28 +08:00
|
|
|
if (!dev->header_ops || !dev->header_ops->parse)
|
2007-09-27 13:13:38 +08:00
|
|
|
return 0;
|
2007-10-09 16:40:57 +08:00
|
|
|
return dev->header_ops->parse(skb, haddr);
|
2007-09-27 13:13:38 +08:00
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
typedef int gifconf_func_t(struct net_device * dev, char __user * bufptr, int len);
|
2013-09-27 05:48:15 +08:00
|
|
|
int register_gifconf(unsigned int family, gifconf_func_t *gifconf);
|
2005-04-17 06:20:36 +08:00
|
|
|
static inline int unregister_gifconf(unsigned int family)
|
|
|
|
{
|
|
|
|
return register_gifconf(family, NULL);
|
|
|
|
}
|
|
|
|
|
2013-05-20 12:02:32 +08:00
|
|
|
#ifdef CONFIG_NET_FLOW_LIMIT
|
2013-06-14 03:29:38 +08:00
|
|
|
#define FLOW_LIMIT_HISTORY (1 << 7) /* must be ^2 and !overflow buckets */
|
2013-05-20 12:02:32 +08:00
|
|
|
struct sd_flow_limit {
|
|
|
|
u64 count;
|
|
|
|
unsigned int num_buckets;
|
|
|
|
unsigned int history_head;
|
|
|
|
u16 history[FLOW_LIMIT_HISTORY];
|
|
|
|
u8 buckets[];
|
|
|
|
};
|
|
|
|
|
|
|
|
extern int netdev_flow_limit_table_len;
|
|
|
|
#endif /* CONFIG_NET_FLOW_LIMIT */
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
2010-04-19 13:07:33 +08:00
|
|
|
* Incoming packets are placed on per-cpu queues
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2009-11-05 01:50:58 +08:00
|
|
|
struct softnet_data {
|
2005-04-17 06:20:36 +08:00
|
|
|
struct list_head poll_list;
|
2010-04-28 06:07:33 +08:00
|
|
|
struct sk_buff_head process_queue;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2010-05-02 13:42:16 +08:00
|
|
|
/* stats */
|
2010-05-03 13:27:59 +08:00
|
|
|
unsigned int processed;
|
|
|
|
unsigned int time_squeeze;
|
|
|
|
unsigned int cpu_collision;
|
|
|
|
unsigned int received_rps;
|
2010-04-15 15:16:59 +08:00
|
|
|
#ifdef CONFIG_RPS
|
2010-04-19 13:07:33 +08:00
|
|
|
struct softnet_data *rps_ipi_list;
|
2014-11-02 22:00:12 +08:00
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_NET_FLOW_LIMIT
|
|
|
|
struct sd_flow_limit __rcu *flow_limit;
|
|
|
|
#endif
|
|
|
|
struct Qdisc *output_queue;
|
|
|
|
struct Qdisc **output_queue_tailp;
|
|
|
|
struct sk_buff *completion_queue;
|
2010-04-19 13:07:33 +08:00
|
|
|
|
2014-11-02 22:00:12 +08:00
|
|
|
#ifdef CONFIG_RPS
|
2010-04-19 13:07:33 +08:00
|
|
|
/* Elements below can be accessed between CPUs for RPS */
|
2010-03-16 16:03:29 +08:00
|
|
|
struct call_single_data csd ____cacheline_aligned_in_smp;
|
2010-04-19 13:07:33 +08:00
|
|
|
struct softnet_data *rps_ipi_next;
|
|
|
|
unsigned int cpu;
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
unsigned int input_queue_head;
|
2010-05-21 02:37:59 +08:00
|
|
|
unsigned int input_queue_tail;
|
2010-03-19 08:45:44 +08:00
|
|
|
#endif
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int dropped;
|
2010-03-16 16:03:29 +08:00
|
|
|
struct sk_buff_head input_pkt_queue;
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
struct napi_struct backlog;
|
2013-05-20 12:02:32 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
};
|
|
|
|
|
2010-05-21 02:37:59 +08:00
|
|
|
static inline void input_queue_head_incr(struct softnet_data *sd)
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
{
|
|
|
|
#ifdef CONFIG_RPS
|
2010-05-21 02:37:59 +08:00
|
|
|
sd->input_queue_head++;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void input_queue_tail_incr_save(struct softnet_data *sd,
|
|
|
|
unsigned int *qtail)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_RPS
|
|
|
|
*qtail = ++sd->input_queue_tail;
|
rfs: Receive Flow Steering
This patch implements receive flow steering (RFS). RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running. RFS is an
extension of Receive Packet Steering (RPS).
The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure. The rxhash is passed in skb's received on
the connection from netif_receive_skb. For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.
The convolution of the simple approach is that it would potentially
allow OOO packets. If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.
To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.
rps_sock_table is a global hash table. Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.
rps_dev_flow_table is specific to each device queue. Each entry
contains a CPU and a tail queue counter. The CPU is the "current"
CPU for a matching flow. The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.
Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length. When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.
And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted. When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:
- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table. This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.
Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality. 2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.
This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.
There are two configuration parameters for RFS. The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue. Both are rounded to power of two.
The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).
The benefits of RFS are dependent on cache hierarchy, application
load, and other factors. On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation. However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp. The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.
e1000e on 8 core Intel
No RFS or RPS 104K tps at 30% CPU
No RFS (best RPS config): 290K tps at 63% CPU
RFS 303K tps at 61% CPU
RPC test tps CPU% 50/90/99% usec latency Latency StdDev
No RFS/RPS 103K 48% 757/900/3185 4472.35
RPS only: 174K 73% 415/993/2468 491.66
RFS 223K 73% 379/651/1382 315.61
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 07:01:27 +08:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2010-03-16 16:03:29 +08:00
|
|
|
DECLARE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void __netif_schedule(struct Qdisc *q);
|
2014-09-13 11:04:52 +08:00
|
|
|
void netif_schedule_queue(struct netdev_queue *txq);
|
2008-07-09 14:11:25 +08:00
|
|
|
|
2008-07-17 16:56:23 +08:00
|
|
|
static inline void netif_tx_schedule_all(struct net_device *dev)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
for (i = 0; i < dev->num_tx_queues; i++)
|
|
|
|
netif_schedule_queue(netdev_get_tx_queue(dev, i));
|
|
|
|
}
|
|
|
|
|
2008-07-23 05:09:06 +08:00
|
|
|
static inline void netif_tx_start_queue(struct netdev_queue *dev_queue)
|
|
|
|
{
|
2011-11-29 00:32:44 +08:00
|
|
|
clear_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
|
2008-07-23 05:09:06 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_start_queue - allow transmit
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Allow upper layers to call the device hard_start_xmit routine.
|
|
|
|
*/
|
2005-04-17 06:20:36 +08:00
|
|
|
static inline void netif_start_queue(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
netif_tx_start_queue(netdev_get_tx_queue(dev, 0));
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2008-07-17 16:56:23 +08:00
|
|
|
static inline void netif_tx_start_all_queues(struct net_device *dev)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
|
|
|
netif_tx_start_queue(txq);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-09-13 11:04:52 +08:00
|
|
|
void netif_tx_wake_queue(struct netdev_queue *dev_queue);
|
2008-07-09 14:14:46 +08:00
|
|
|
|
2008-07-23 05:09:06 +08:00
|
|
|
/**
|
|
|
|
* netif_wake_queue - restart transmit
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Allow upper layers to call the device hard_start_xmit routine.
|
|
|
|
* Used for flow control when transmit resources are available.
|
|
|
|
*/
|
2008-07-09 14:14:46 +08:00
|
|
|
static inline void netif_wake_queue(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
netif_tx_wake_queue(netdev_get_tx_queue(dev, 0));
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2008-07-17 16:56:23 +08:00
|
|
|
static inline void netif_tx_wake_all_queues(struct net_device *dev)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
|
|
|
netif_tx_wake_queue(txq);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-07-23 05:09:06 +08:00
|
|
|
static inline void netif_tx_stop_queue(struct netdev_queue *dev_queue)
|
|
|
|
{
|
2011-11-29 00:32:44 +08:00
|
|
|
set_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
|
2008-07-23 05:09:06 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_stop_queue - stop transmitted packets
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Stop upper layers calling the device hard_start_xmit routine.
|
|
|
|
* Used for flow control when transmit resources are unavailable.
|
|
|
|
*/
|
2005-04-17 06:20:36 +08:00
|
|
|
static inline void netif_stop_queue(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
netif_tx_stop_queue(netdev_get_tx_queue(dev, 0));
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2015-05-12 03:17:53 +08:00
|
|
|
void netif_tx_stop_all_queues(struct net_device *dev);
|
2008-07-17 16:56:23 +08:00
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_tx_queue_stopped(const struct netdev_queue *dev_queue)
|
2008-07-23 05:09:06 +08:00
|
|
|
{
|
2011-11-29 00:32:44 +08:00
|
|
|
return test_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
|
2008-07-23 05:09:06 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_queue_stopped - test if transmit queue is flowblocked
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Test if transmit queue on device is currently unable to send.
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_queue_stopped(const struct net_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
return netif_tx_queue_stopped(netdev_get_tx_queue(dev, 0));
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_xmit_stopped(const struct netdev_queue *dev_queue)
|
2008-08-01 07:58:50 +08:00
|
|
|
{
|
2011-11-29 00:32:44 +08:00
|
|
|
return dev_queue->state & QUEUE_STATE_ANY_XOFF;
|
|
|
|
}
|
|
|
|
|
2014-04-03 02:52:57 +08:00
|
|
|
static inline bool
|
|
|
|
netif_xmit_frozen_or_stopped(const struct netdev_queue *dev_queue)
|
2011-11-29 00:32:44 +08:00
|
|
|
{
|
|
|
|
return dev_queue->state & QUEUE_STATE_ANY_XOFF_OR_FROZEN;
|
|
|
|
}
|
|
|
|
|
2014-04-03 02:52:57 +08:00
|
|
|
static inline bool
|
|
|
|
netif_xmit_frozen_or_drv_stopped(const struct netdev_queue *dev_queue)
|
|
|
|
{
|
|
|
|
return dev_queue->state & QUEUE_STATE_DRV_XOFF_OR_FROZEN;
|
|
|
|
}
|
|
|
|
|
2014-10-08 23:19:27 +08:00
|
|
|
/**
|
|
|
|
* netdev_txq_bql_enqueue_prefetchw - prefetch bql data for write
|
|
|
|
* @dev_queue: pointer to transmit queue
|
|
|
|
*
|
|
|
|
* BQL enabled drivers might use this helper in their ndo_start_xmit(),
|
|
|
|
* to give appropriate hint to the cpu.
|
|
|
|
*/
|
|
|
|
static inline void netdev_txq_bql_enqueue_prefetchw(struct netdev_queue *dev_queue)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_BQL
|
|
|
|
prefetchw(&dev_queue->dql.num_queued);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* netdev_txq_bql_complete_prefetchw - prefetch bql data for write
|
|
|
|
* @dev_queue: pointer to transmit queue
|
|
|
|
*
|
|
|
|
* BQL enabled drivers might use this helper in their TX completion path,
|
|
|
|
* to give appropriate hint to the cpu.
|
|
|
|
*/
|
|
|
|
static inline void netdev_txq_bql_complete_prefetchw(struct netdev_queue *dev_queue)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_BQL
|
|
|
|
prefetchw(&dev_queue->dql.limit);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2011-11-29 00:32:52 +08:00
|
|
|
static inline void netdev_tx_sent_queue(struct netdev_queue *dev_queue,
|
|
|
|
unsigned int bytes)
|
|
|
|
{
|
2011-11-29 00:33:09 +08:00
|
|
|
#ifdef CONFIG_BQL
|
|
|
|
dql_queued(&dev_queue->dql, bytes);
|
2012-02-07 10:29:06 +08:00
|
|
|
|
|
|
|
if (likely(dql_avail(&dev_queue->dql) >= 0))
|
|
|
|
return;
|
|
|
|
|
|
|
|
set_bit(__QUEUE_STATE_STACK_XOFF, &dev_queue->state);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The XOFF flag must be set before checking the dql_avail below,
|
|
|
|
* because in netdev_tx_completed_queue we update the dql_completed
|
|
|
|
* before checking the XOFF flag.
|
|
|
|
*/
|
|
|
|
smp_mb();
|
|
|
|
|
|
|
|
/* check again in case another CPU has just made room avail */
|
|
|
|
if (unlikely(dql_avail(&dev_queue->dql) >= 0))
|
|
|
|
clear_bit(__QUEUE_STATE_STACK_XOFF, &dev_queue->state);
|
2011-11-29 00:33:09 +08:00
|
|
|
#endif
|
2011-11-29 00:32:52 +08:00
|
|
|
}
|
|
|
|
|
2013-09-06 23:58:00 +08:00
|
|
|
/**
|
|
|
|
* netdev_sent_queue - report the number of bytes queued to hardware
|
|
|
|
* @dev: network device
|
|
|
|
* @bytes: number of bytes queued to the hardware device queue
|
|
|
|
*
|
|
|
|
* Report the number of bytes queued for sending/completion to the network
|
|
|
|
* device hardware queue. @bytes should be a good approximation and should
|
|
|
|
* exactly match netdev_completed_queue() @bytes
|
|
|
|
*/
|
2011-11-29 00:32:52 +08:00
|
|
|
static inline void netdev_sent_queue(struct net_device *dev, unsigned int bytes)
|
|
|
|
{
|
|
|
|
netdev_tx_sent_queue(netdev_get_tx_queue(dev, 0), bytes);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netdev_tx_completed_queue(struct netdev_queue *dev_queue,
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int pkts, unsigned int bytes)
|
2011-11-29 00:32:52 +08:00
|
|
|
{
|
2011-11-29 00:33:09 +08:00
|
|
|
#ifdef CONFIG_BQL
|
2012-02-07 10:29:06 +08:00
|
|
|
if (unlikely(!bytes))
|
|
|
|
return;
|
|
|
|
|
|
|
|
dql_completed(&dev_queue->dql, bytes);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Without the memory barrier there is a small possiblity that
|
|
|
|
* netdev_tx_sent_queue will miss the update and cause the queue to
|
|
|
|
* be stopped forever
|
|
|
|
*/
|
|
|
|
smp_mb();
|
|
|
|
|
|
|
|
if (dql_avail(&dev_queue->dql) < 0)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (test_and_clear_bit(__QUEUE_STATE_STACK_XOFF, &dev_queue->state))
|
|
|
|
netif_schedule_queue(dev_queue);
|
2011-11-29 00:33:09 +08:00
|
|
|
#endif
|
2011-11-29 00:32:52 +08:00
|
|
|
}
|
|
|
|
|
2013-09-06 23:58:00 +08:00
|
|
|
/**
|
|
|
|
* netdev_completed_queue - report bytes and packets completed by device
|
|
|
|
* @dev: network device
|
|
|
|
* @pkts: actual number of packets sent over the medium
|
|
|
|
* @bytes: actual number of bytes sent over the medium
|
|
|
|
*
|
|
|
|
* Report the number of bytes and packets transmitted by the network device
|
|
|
|
* hardware queue over the physical medium, @bytes must exactly match the
|
|
|
|
* @bytes amount passed to netdev_sent_queue()
|
|
|
|
*/
|
2011-11-29 00:32:52 +08:00
|
|
|
static inline void netdev_completed_queue(struct net_device *dev,
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int pkts, unsigned int bytes)
|
2011-11-29 00:32:52 +08:00
|
|
|
{
|
|
|
|
netdev_tx_completed_queue(netdev_get_tx_queue(dev, 0), pkts, bytes);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netdev_tx_reset_queue(struct netdev_queue *q)
|
|
|
|
{
|
2011-11-29 00:33:09 +08:00
|
|
|
#ifdef CONFIG_BQL
|
2012-02-07 10:29:01 +08:00
|
|
|
clear_bit(__QUEUE_STATE_STACK_XOFF, &q->state);
|
2011-11-29 00:33:09 +08:00
|
|
|
dql_reset(&q->dql);
|
|
|
|
#endif
|
2011-11-29 00:32:52 +08:00
|
|
|
}
|
|
|
|
|
2013-09-06 23:58:00 +08:00
|
|
|
/**
|
|
|
|
* netdev_reset_queue - reset the packets and bytes count of a network device
|
|
|
|
* @dev_queue: network device
|
|
|
|
*
|
|
|
|
* Reset the bytes and packet count of a network device and clear the
|
|
|
|
* software flow control OFF bit for this network device
|
|
|
|
*/
|
2011-11-29 00:32:52 +08:00
|
|
|
static inline void netdev_reset_queue(struct net_device *dev_queue)
|
|
|
|
{
|
|
|
|
netdev_tx_reset_queue(netdev_get_tx_queue(dev_queue, 0));
|
2008-08-01 07:58:50 +08:00
|
|
|
}
|
|
|
|
|
2014-02-16 22:55:21 +08:00
|
|
|
/**
|
|
|
|
* netdev_cap_txqueue - check if selected tx queue exceeds device queues
|
|
|
|
* @dev: network device
|
|
|
|
* @queue_index: given tx queue index
|
|
|
|
*
|
|
|
|
* Returns 0 if given tx queue index >= number of device tx queues,
|
|
|
|
* otherwise returns the originally passed tx queue index.
|
|
|
|
*/
|
|
|
|
static inline u16 netdev_cap_txqueue(struct net_device *dev, u16 queue_index)
|
|
|
|
{
|
|
|
|
if (unlikely(queue_index >= dev->real_num_tx_queues)) {
|
|
|
|
net_warn_ratelimited("%s selects TX queue %d, but real number of TX queues is %d\n",
|
|
|
|
dev->name, queue_index,
|
|
|
|
dev->real_num_tx_queues);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return queue_index;
|
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_running - test if up
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Test if the device has been brought up.
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_running(const struct net_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
return test_bit(__LINK_STATE_START, &dev->state);
|
|
|
|
}
|
|
|
|
|
2007-07-07 04:36:20 +08:00
|
|
|
/*
|
|
|
|
* Routines to manage the subqueues on a device. We only need start
|
|
|
|
* stop, and a check if it's stopped. All other device management is
|
|
|
|
* done at the overall netdevice level.
|
|
|
|
* Also test the device if we're multiqueue.
|
|
|
|
*/
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
|
|
|
|
/**
|
|
|
|
* netif_start_subqueue - allow sending packets on subqueue
|
|
|
|
* @dev: network device
|
|
|
|
* @queue_index: sub queue index
|
|
|
|
*
|
|
|
|
* Start individual transmit queue of a device with multiple transmit queues.
|
|
|
|
*/
|
2007-07-07 04:36:20 +08:00
|
|
|
static inline void netif_start_subqueue(struct net_device *dev, u16 queue_index)
|
|
|
|
{
|
2008-07-17 16:56:23 +08:00
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
|
2009-08-30 04:21:21 +08:00
|
|
|
|
|
|
|
netif_tx_start_queue(txq);
|
2007-07-07 04:36:20 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_stop_subqueue - stop sending packets on subqueue
|
|
|
|
* @dev: network device
|
|
|
|
* @queue_index: sub queue index
|
|
|
|
*
|
|
|
|
* Stop individual transmit queue of a device with multiple transmit queues.
|
|
|
|
*/
|
2007-07-07 04:36:20 +08:00
|
|
|
static inline void netif_stop_subqueue(struct net_device *dev, u16 queue_index)
|
|
|
|
{
|
2008-07-17 16:56:23 +08:00
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
|
2009-08-30 04:21:21 +08:00
|
|
|
netif_tx_stop_queue(txq);
|
2007-07-07 04:36:20 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_subqueue_stopped - test status of subqueue
|
|
|
|
* @dev: network device
|
|
|
|
* @queue_index: sub queue index
|
|
|
|
*
|
|
|
|
* Check individual transmit queue of a device with multiple transmit queues.
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool __netif_subqueue_stopped(const struct net_device *dev,
|
|
|
|
u16 queue_index)
|
2007-07-07 04:36:20 +08:00
|
|
|
{
|
2008-07-17 16:56:23 +08:00
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
|
2009-08-30 04:21:21 +08:00
|
|
|
|
|
|
|
return netif_tx_queue_stopped(txq);
|
2007-07-07 04:36:20 +08:00
|
|
|
}
|
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_subqueue_stopped(const struct net_device *dev,
|
|
|
|
struct sk_buff *skb)
|
2007-10-22 08:01:56 +08:00
|
|
|
{
|
|
|
|
return __netif_subqueue_stopped(dev, skb_get_queue_mapping(skb));
|
|
|
|
}
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
|
2014-09-13 11:04:52 +08:00
|
|
|
void netif_wake_subqueue(struct net_device *dev, u16 queue_index);
|
2007-07-07 04:36:20 +08:00
|
|
|
|
2013-01-10 16:57:02 +08:00
|
|
|
#ifdef CONFIG_XPS
|
2013-10-09 11:07:53 +08:00
|
|
|
int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
|
2013-09-27 05:48:15 +08:00
|
|
|
u16 index);
|
2013-01-10 16:57:02 +08:00
|
|
|
#else
|
|
|
|
static inline int netif_set_xps_queue(struct net_device *dev,
|
2013-10-02 14:14:06 +08:00
|
|
|
const struct cpumask *mask,
|
2013-01-10 16:57:02 +08:00
|
|
|
u16 index)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2015-05-12 20:56:12 +08:00
|
|
|
u16 __skb_tx_hash(const struct net_device *dev, struct sk_buff *skb,
|
|
|
|
unsigned int num_tx_queues);
|
|
|
|
|
2010-12-13 14:27:10 +08:00
|
|
|
/*
|
|
|
|
* Returns a Tx hash for the given packet when dev->real_num_tx_queues is used
|
|
|
|
* as a distribution range limit for the returned value.
|
|
|
|
*/
|
|
|
|
static inline u16 skb_tx_hash(const struct net_device *dev,
|
2014-07-02 12:32:27 +08:00
|
|
|
struct sk_buff *skb)
|
2010-12-13 14:27:10 +08:00
|
|
|
{
|
|
|
|
return __skb_tx_hash(dev, skb, dev->real_num_tx_queues);
|
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_is_multiqueue - test if device has multiple transmit queues
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Check if device has multiple transmit queues
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_is_multiqueue(const struct net_device *dev)
|
2007-07-07 04:36:20 +08:00
|
|
|
{
|
2010-09-23 04:43:57 +08:00
|
|
|
return dev->num_tx_queues > 1;
|
2007-07-07 04:36:20 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int netif_set_real_num_tx_queues(struct net_device *dev, unsigned int txq);
|
2010-07-01 21:21:57 +08:00
|
|
|
|
2014-01-17 14:23:28 +08:00
|
|
|
#ifdef CONFIG_SYSFS
|
2013-09-27 05:48:15 +08:00
|
|
|
int netif_set_real_num_rx_queues(struct net_device *dev, unsigned int rxq);
|
2010-09-27 16:24:33 +08:00
|
|
|
#else
|
|
|
|
static inline int netif_set_real_num_rx_queues(struct net_device *dev,
|
|
|
|
unsigned int rxq)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2014-01-17 14:23:28 +08:00
|
|
|
#ifdef CONFIG_SYSFS
|
|
|
|
static inline unsigned int get_netdev_rx_queue_index(
|
|
|
|
struct netdev_rx_queue *queue)
|
|
|
|
{
|
|
|
|
struct net_device *dev = queue->dev;
|
|
|
|
int index = queue - dev->_rx;
|
|
|
|
|
|
|
|
BUG_ON(index >= dev->num_rx_queues);
|
|
|
|
return index;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2012-07-01 11:18:50 +08:00
|
|
|
#define DEFAULT_MAX_NUM_RSS_QUEUES (8)
|
2013-09-27 05:48:15 +08:00
|
|
|
int netif_get_num_default_rss_queues(void);
|
2012-07-01 11:18:50 +08:00
|
|
|
|
2013-12-05 20:45:08 +08:00
|
|
|
enum skb_free_reason {
|
|
|
|
SKB_REASON_CONSUMED,
|
|
|
|
SKB_REASON_DROPPED,
|
|
|
|
};
|
|
|
|
|
|
|
|
void __dev_kfree_skb_irq(struct sk_buff *skb, enum skb_free_reason reason);
|
|
|
|
void __dev_kfree_skb_any(struct sk_buff *skb, enum skb_free_reason reason);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-12-05 20:45:08 +08:00
|
|
|
/*
|
|
|
|
* It is not allowed to call kfree_skb() or consume_skb() from hardware
|
|
|
|
* interrupt context or with hardware interrupts being disabled.
|
|
|
|
* (in_irq() || irqs_disabled())
|
|
|
|
*
|
|
|
|
* We provide four helpers that can be used in following contexts :
|
|
|
|
*
|
|
|
|
* dev_kfree_skb_irq(skb) when caller drops a packet from irq context,
|
|
|
|
* replacing kfree_skb(skb)
|
|
|
|
*
|
|
|
|
* dev_consume_skb_irq(skb) when caller consumes a packet from irq context.
|
|
|
|
* Typically used in place of consume_skb(skb) in TX completion path
|
|
|
|
*
|
|
|
|
* dev_kfree_skb_any(skb) when caller doesn't know its current irq context,
|
|
|
|
* replacing kfree_skb(skb)
|
|
|
|
*
|
|
|
|
* dev_consume_skb_any(skb) when caller doesn't know its current irq context,
|
|
|
|
* and consumed a packet. Used in place of consume_skb(skb)
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2013-12-05 20:45:08 +08:00
|
|
|
static inline void dev_kfree_skb_irq(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
__dev_kfree_skb_irq(skb, SKB_REASON_DROPPED);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void dev_consume_skb_irq(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
__dev_kfree_skb_irq(skb, SKB_REASON_CONSUMED);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void dev_kfree_skb_any(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
__dev_kfree_skb_any(skb, SKB_REASON_DROPPED);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void dev_consume_skb_any(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
__dev_kfree_skb_any(skb, SKB_REASON_CONSUMED);
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int netif_rx(struct sk_buff *skb);
|
|
|
|
int netif_rx_ni(struct sk_buff *skb);
|
2015-04-06 10:19:04 +08:00
|
|
|
int netif_receive_skb_sk(struct sock *sk, struct sk_buff *skb);
|
|
|
|
static inline int netif_receive_skb(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
return netif_receive_skb_sk(skb->sk, skb);
|
|
|
|
}
|
2013-09-27 05:48:15 +08:00
|
|
|
gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb);
|
|
|
|
void napi_gro_flush(struct napi_struct *napi, bool flush_old);
|
|
|
|
struct sk_buff *napi_get_frags(struct napi_struct *napi);
|
|
|
|
gro_result_t napi_gro_frags(struct napi_struct *napi);
|
net-gre-gro: Add GRE support to the GRO stack
This patch built on top of Commit 299603e8370a93dd5d8e8d800f0dff1ce2c53d36
("net-gro: Prepare GRO stack for the upcoming tunneling support") to add
the support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO
stack. It also serves as an example for supporting other encapsulation
protocols in the GRO stack in the future.
The patch supports version 0 and all the flags (key, csum, seq#) but
will flush any pkt with the S (seq#) flag. This is because the S flag
is not support by GSO, and a GRO pkt may end up in the forwarding path,
thus requiring GSO support to break it up correctly.
Currently the "packet_offload" structure only contains L3 (ETH_P_IP/
ETH_P_IPV6) GRO offload support so the encapped pkts are limited to
IP pkts (i.e., w/o L2 hdr). But support for other protocol type can
be easily added, so is the support for GRE variations like NVGRE.
The patch also support csum offload. Specifically if the csum flag is on
and the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),
the code will take advantage of the csum computed by the h/w when
validating the GRE csum.
Note that commit 60769a5dcd8755715c7143b4571d5c44f01796f1 "ipv4: gre:
add GRO capability" already introduces GRO capability to IPv4 GRE
tunnels, using the gro_cells infrastructure. But GRO is done after
GRE hdr has been removed (i.e., decapped). The following patch applies
GRO when pkts first come in (before hitting the GRE tunnel code). There
is some performance advantage for applying GRO as early as possible.
Also this approach is transparent to other subsystem like Open vSwitch
where GRE decap is handled outside of the IP stack hence making it
harder for the gro_cells stuff to apply. On the other hand, some NICs
are still not capable of hashing on the inner hdr of a GRE pkt (RSS).
In that case the GRO processing of pkts from the same remote host will
all happen on the same CPU and the performance may be suboptimal.
I'm including some rough preliminary performance numbers below. Note
that the performance will be highly dependent on traffic load, mix as
usual. Moreover it also depends on NIC offload features hence the
following is by no means a comprehesive study. Local testing and tuning
will be needed to decide the best setting.
All tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.
(super_netperf 50 -H 192.168.1.18 -l 30)
An IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1
mode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)
is configured.
The GRO support for pkts AFTER decap are controlled through the device
feature of the GRE device (e.g., ethtool -K gre1 gro on/off).
1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off
thruput: 9.16Gbps
CPU utilization: 19%
1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off
thruput: 5.9Gbps
CPU utilization: 15%
1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 12-13%
1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on
thruput: 9.26Gbps
CPU utilization: 10%
The following tests were performed on a different NIC that is capable of
csum offload. I.e., the h/w is capable of computing IP payload csum
(CHECKSUM_COMPLETE).
2.1 ethtool -K gre1 gro on (hence will use gro_cells)
2.1.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 8.53Gbps
CPU utilization: 9%
2.1.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 8.97Gbps
CPU utilization: 7-8%
2.1.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 8.83Gbps
CPU utilization: 5-6%
2.1.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.98Gbps
CPU utilization: 5%
2.2 ethtool -K gre1 gro off
2.2.1 ethtool -K eth0 gro off; csum offload disabled
thruput: 5.93Gbps
CPU utilization: 9%
2.2.2 ethtool -K eth0 gro off; csum offload enabled
thruput: 5.62Gbps
CPU utilization: 8%
2.2.3 ethtool -K eth0 gro on; csum offload disabled
thruput: 7.69Gbps
CPU utilization: 8%
2.2.4 ethtool -K eth0 gro on; csum offload enabled
thruput: 8.96Gbps
CPU utilization: 5-6%
Signed-off-by: H.K. Jerry Chu <hkchu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-01-08 02:23:19 +08:00
|
|
|
struct packet_offload *gro_find_receive_by_type(__be16 type);
|
|
|
|
struct packet_offload *gro_find_complete_by_type(__be16 type);
|
2009-04-16 17:02:07 +08:00
|
|
|
|
|
|
|
static inline void napi_free_frags(struct napi_struct *napi)
|
|
|
|
{
|
|
|
|
kfree_skb(napi->skb);
|
|
|
|
napi->skb = NULL;
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int netdev_rx_handler_register(struct net_device *dev,
|
|
|
|
rx_handler_func_t *rx_handler,
|
|
|
|
void *rx_handler_data);
|
|
|
|
void netdev_rx_handler_unregister(struct net_device *dev);
|
|
|
|
|
|
|
|
bool dev_valid_name(const char *name);
|
|
|
|
int dev_ioctl(struct net *net, unsigned int cmd, void __user *);
|
|
|
|
int dev_ethtool(struct net *net, struct ifreq *);
|
|
|
|
unsigned int dev_get_flags(const struct net_device *);
|
|
|
|
int __dev_change_flags(struct net_device *, unsigned int flags);
|
|
|
|
int dev_change_flags(struct net_device *, unsigned int);
|
2013-10-01 03:36:45 +08:00
|
|
|
void __dev_notify_flags(struct net_device *, unsigned int old_flags,
|
|
|
|
unsigned int gchanges);
|
2013-09-27 05:48:15 +08:00
|
|
|
int dev_change_name(struct net_device *, const char *);
|
|
|
|
int dev_set_alias(struct net_device *, const char *, size_t);
|
|
|
|
int dev_change_net_namespace(struct net_device *, struct net *, const char *);
|
|
|
|
int dev_set_mtu(struct net_device *, int);
|
|
|
|
void dev_set_group(struct net_device *, int);
|
|
|
|
int dev_set_mac_address(struct net_device *, struct sockaddr *);
|
|
|
|
int dev_change_carrier(struct net_device *, bool new_carrier);
|
|
|
|
int dev_get_phys_port_id(struct net_device *dev,
|
2014-11-28 21:34:16 +08:00
|
|
|
struct netdev_phys_item_id *ppid);
|
2015-03-18 10:23:15 +08:00
|
|
|
int dev_get_phys_port_name(struct net_device *dev,
|
|
|
|
char *name, size_t len);
|
2015-07-15 04:43:19 +08:00
|
|
|
int dev_change_proto_down(struct net_device *dev, bool proto_down);
|
2014-10-04 06:31:07 +08:00
|
|
|
struct sk_buff *validate_xmit_skb_list(struct sk_buff *skb, struct net_device *dev);
|
2014-08-31 10:22:20 +08:00
|
|
|
struct sk_buff *dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
|
|
|
|
struct netdev_queue *txq, int *ret);
|
2014-04-17 13:45:03 +08:00
|
|
|
int __dev_forward_skb(struct net_device *dev, struct sk_buff *skb);
|
2013-09-27 05:48:15 +08:00
|
|
|
int dev_forward_skb(struct net_device *dev, struct sk_buff *skb);
|
2014-03-28 05:32:29 +08:00
|
|
|
bool is_skb_forwardable(struct net_device *dev, struct sk_buff *skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-08-16 13:18:02 +08:00
|
|
|
extern int netdev_budget;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Called by rtnetlink.c:rtnl_unlock() */
|
2013-09-27 05:48:15 +08:00
|
|
|
void netdev_run_todo(void);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* dev_put - release reference to device
|
|
|
|
* @dev: network device
|
|
|
|
*
|
2007-10-11 12:18:17 +08:00
|
|
|
* Release reference to device to allow it to be freed.
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
*/
|
2005-04-17 06:20:36 +08:00
|
|
|
static inline void dev_put(struct net_device *dev)
|
|
|
|
{
|
2011-12-23 01:58:51 +08:00
|
|
|
this_cpu_dec(*dev->pcpu_refcnt);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* dev_hold - get reference to device
|
|
|
|
* @dev: network device
|
|
|
|
*
|
2007-10-11 12:18:17 +08:00
|
|
|
* Hold reference to device to keep it from being freed.
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
*/
|
2006-03-21 14:32:28 +08:00
|
|
|
static inline void dev_hold(struct net_device *dev)
|
|
|
|
{
|
2011-12-23 01:58:51 +08:00
|
|
|
this_cpu_inc(*dev->pcpu_refcnt);
|
2006-03-21 14:32:28 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Carrier loss detection, dial on demand. The functions netif_carrier_on
|
|
|
|
* and _off may be called from IRQ context, but it is caller
|
|
|
|
* who is responsible for serialization of these calls.
|
2006-03-21 09:09:11 +08:00
|
|
|
*
|
|
|
|
* The name carrier is inappropriate, these functions should really be
|
|
|
|
* called netif_lowerlayer_*() because they represent the state of any
|
|
|
|
* kind of lower layer not just hardware media.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void linkwatch_init_dev(struct net_device *dev);
|
|
|
|
void linkwatch_fire_event(struct net_device *dev);
|
|
|
|
void linkwatch_forget_dev(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_carrier_ok - test if carrier present
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Check if carrier is present on device
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_carrier_ok(const struct net_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
return !test_bit(__LINK_STATE_NOCARRIER, &dev->state);
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
unsigned long dev_trans_start(struct net_device *dev);
|
2009-05-18 11:55:16 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void __netdev_watchdog_up(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void netif_carrier_on(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void netif_carrier_off(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_dormant_on - mark device as dormant.
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Mark device as dormant (as per RFC2863).
|
|
|
|
*
|
|
|
|
* The dormant state indicates that the relevant interface is not
|
|
|
|
* actually in a condition to pass packets (i.e., it is not 'up') but is
|
|
|
|
* in a "pending" state, waiting for some external event. For "on-
|
|
|
|
* demand" interfaces, this new state identifies the situation where the
|
|
|
|
* interface is waiting for events to place it in the up state.
|
|
|
|
*
|
|
|
|
*/
|
2006-03-21 09:09:11 +08:00
|
|
|
static inline void netif_dormant_on(struct net_device *dev)
|
|
|
|
{
|
|
|
|
if (!test_and_set_bit(__LINK_STATE_DORMANT, &dev->state))
|
|
|
|
linkwatch_fire_event(dev);
|
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_dormant_off - set device as not dormant.
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Device is not in dormant state.
|
|
|
|
*/
|
2006-03-21 09:09:11 +08:00
|
|
|
static inline void netif_dormant_off(struct net_device *dev)
|
|
|
|
{
|
|
|
|
if (test_and_clear_bit(__LINK_STATE_DORMANT, &dev->state))
|
|
|
|
linkwatch_fire_event(dev);
|
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_dormant - test if carrier present
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Check if carrier is present on device
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_dormant(const struct net_device *dev)
|
2006-03-21 09:09:11 +08:00
|
|
|
{
|
|
|
|
return test_bit(__LINK_STATE_DORMANT, &dev->state);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_oper_up - test if device is operational
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Check if carrier is operational
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_oper_up(const struct net_device *dev)
|
2009-11-05 01:50:58 +08:00
|
|
|
{
|
2006-03-21 09:09:11 +08:00
|
|
|
return (dev->operstate == IF_OPER_UP ||
|
|
|
|
dev->operstate == IF_OPER_UNKNOWN /* backward compat */);
|
|
|
|
}
|
|
|
|
|
[NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.
In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.
The signature of the ->poll() call back goes from:
int foo_poll(struct net_device *dev, int *budget)
to
int foo_poll(struct napi_struct *napi, int budget)
The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract). The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.
The napi_struct is to be embedded in the device driver private data
structures.
Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler. Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.
With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.
Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.
[ Ported to current tree and all drivers converted. Integrated
Stephen's follow-on kerneldoc additions, and restored poll_list
handling to the old style to fix mutual exclusion issues. -DaveM ]
Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 07:41:36 +08:00
|
|
|
/**
|
|
|
|
* netif_device_present - is device available or removed
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Check if device has not been removed from system.
|
|
|
|
*/
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_device_present(struct net_device *dev)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
return test_bit(__LINK_STATE_PRESENT, &dev->state);
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void netif_device_detach(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void netif_device_attach(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Network interface message level settings
|
|
|
|
*/
|
|
|
|
|
|
|
|
enum {
|
|
|
|
NETIF_MSG_DRV = 0x0001,
|
|
|
|
NETIF_MSG_PROBE = 0x0002,
|
|
|
|
NETIF_MSG_LINK = 0x0004,
|
|
|
|
NETIF_MSG_TIMER = 0x0008,
|
|
|
|
NETIF_MSG_IFDOWN = 0x0010,
|
|
|
|
NETIF_MSG_IFUP = 0x0020,
|
|
|
|
NETIF_MSG_RX_ERR = 0x0040,
|
|
|
|
NETIF_MSG_TX_ERR = 0x0080,
|
|
|
|
NETIF_MSG_TX_QUEUED = 0x0100,
|
|
|
|
NETIF_MSG_INTR = 0x0200,
|
|
|
|
NETIF_MSG_TX_DONE = 0x0400,
|
|
|
|
NETIF_MSG_RX_STATUS = 0x0800,
|
|
|
|
NETIF_MSG_PKTDATA = 0x1000,
|
|
|
|
NETIF_MSG_HW = 0x2000,
|
|
|
|
NETIF_MSG_WOL = 0x4000,
|
|
|
|
};
|
|
|
|
|
|
|
|
#define netif_msg_drv(p) ((p)->msg_enable & NETIF_MSG_DRV)
|
|
|
|
#define netif_msg_probe(p) ((p)->msg_enable & NETIF_MSG_PROBE)
|
|
|
|
#define netif_msg_link(p) ((p)->msg_enable & NETIF_MSG_LINK)
|
|
|
|
#define netif_msg_timer(p) ((p)->msg_enable & NETIF_MSG_TIMER)
|
|
|
|
#define netif_msg_ifdown(p) ((p)->msg_enable & NETIF_MSG_IFDOWN)
|
|
|
|
#define netif_msg_ifup(p) ((p)->msg_enable & NETIF_MSG_IFUP)
|
|
|
|
#define netif_msg_rx_err(p) ((p)->msg_enable & NETIF_MSG_RX_ERR)
|
|
|
|
#define netif_msg_tx_err(p) ((p)->msg_enable & NETIF_MSG_TX_ERR)
|
|
|
|
#define netif_msg_tx_queued(p) ((p)->msg_enable & NETIF_MSG_TX_QUEUED)
|
|
|
|
#define netif_msg_intr(p) ((p)->msg_enable & NETIF_MSG_INTR)
|
|
|
|
#define netif_msg_tx_done(p) ((p)->msg_enable & NETIF_MSG_TX_DONE)
|
|
|
|
#define netif_msg_rx_status(p) ((p)->msg_enable & NETIF_MSG_RX_STATUS)
|
|
|
|
#define netif_msg_pktdata(p) ((p)->msg_enable & NETIF_MSG_PKTDATA)
|
|
|
|
#define netif_msg_hw(p) ((p)->msg_enable & NETIF_MSG_HW)
|
|
|
|
#define netif_msg_wol(p) ((p)->msg_enable & NETIF_MSG_WOL)
|
|
|
|
|
|
|
|
static inline u32 netif_msg_init(int debug_value, int default_msg_enable_bits)
|
|
|
|
{
|
|
|
|
/* use default */
|
|
|
|
if (debug_value < 0 || debug_value >= (sizeof(u32) * 8))
|
|
|
|
return default_msg_enable_bits;
|
|
|
|
if (debug_value == 0) /* no output */
|
|
|
|
return 0;
|
|
|
|
/* set low N bits */
|
|
|
|
return (1 << debug_value) - 1;
|
|
|
|
}
|
|
|
|
|
2008-07-09 14:13:53 +08:00
|
|
|
static inline void __netif_tx_lock(struct netdev_queue *txq, int cpu)
|
2006-06-10 03:20:56 +08:00
|
|
|
{
|
2008-07-09 14:13:53 +08:00
|
|
|
spin_lock(&txq->_xmit_lock);
|
|
|
|
txq->xmit_lock_owner = cpu;
|
2007-09-17 05:40:49 +08:00
|
|
|
}
|
|
|
|
|
2008-07-17 16:56:23 +08:00
|
|
|
static inline void __netif_tx_lock_bh(struct netdev_queue *txq)
|
|
|
|
{
|
|
|
|
spin_lock_bh(&txq->_xmit_lock);
|
|
|
|
txq->xmit_lock_owner = smp_processor_id();
|
|
|
|
}
|
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool __netif_tx_trylock(struct netdev_queue *txq)
|
2008-08-01 07:58:50 +08:00
|
|
|
{
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
bool ok = spin_trylock(&txq->_xmit_lock);
|
2008-08-01 07:58:50 +08:00
|
|
|
if (likely(ok))
|
|
|
|
txq->xmit_lock_owner = smp_processor_id();
|
|
|
|
return ok;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void __netif_tx_unlock(struct netdev_queue *txq)
|
|
|
|
{
|
|
|
|
txq->xmit_lock_owner = -1;
|
|
|
|
spin_unlock(&txq->_xmit_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void __netif_tx_unlock_bh(struct netdev_queue *txq)
|
|
|
|
{
|
|
|
|
txq->xmit_lock_owner = -1;
|
|
|
|
spin_unlock_bh(&txq->_xmit_lock);
|
|
|
|
}
|
|
|
|
|
2009-05-26 13:58:01 +08:00
|
|
|
static inline void txq_trans_update(struct netdev_queue *txq)
|
|
|
|
{
|
|
|
|
if (txq->xmit_lock_owner != -1)
|
|
|
|
txq->trans_start = jiffies;
|
|
|
|
}
|
|
|
|
|
2008-07-23 05:09:06 +08:00
|
|
|
/**
|
|
|
|
* netif_tx_lock - grab network device transmit lock
|
|
|
|
* @dev: network device
|
|
|
|
*
|
|
|
|
* Get network device transmit lock
|
|
|
|
*/
|
2007-09-17 05:40:49 +08:00
|
|
|
static inline void netif_tx_lock(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
unsigned int i;
|
2008-08-01 07:58:50 +08:00
|
|
|
int cpu;
|
2008-07-09 14:13:53 +08:00
|
|
|
|
2008-08-01 07:58:50 +08:00
|
|
|
spin_lock(&dev->tx_global_lock);
|
|
|
|
cpu = smp_processor_id();
|
2008-07-17 15:34:19 +08:00
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
2008-08-01 07:58:50 +08:00
|
|
|
|
|
|
|
/* We are the only thread of execution doing a
|
|
|
|
* freeze, but we have to grab the _xmit_lock in
|
|
|
|
* order to synchronize with threads which are in
|
|
|
|
* the ->hard_start_xmit() handler and already
|
|
|
|
* checked the frozen bit.
|
|
|
|
*/
|
2008-07-17 15:34:19 +08:00
|
|
|
__netif_tx_lock(txq, cpu);
|
2008-08-01 07:58:50 +08:00
|
|
|
set_bit(__QUEUE_STATE_FROZEN, &txq->state);
|
|
|
|
__netif_tx_unlock(txq);
|
2008-07-17 15:34:19 +08:00
|
|
|
}
|
2006-06-10 03:20:56 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netif_tx_lock_bh(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
local_bh_disable();
|
|
|
|
netif_tx_lock(dev);
|
2006-06-10 03:20:56 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netif_tx_unlock(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
2008-07-09 14:13:53 +08:00
|
|
|
|
2008-08-01 07:58:50 +08:00
|
|
|
/* No need to grab the _xmit_lock here. If the
|
|
|
|
* queue is not stopped for another reason, we
|
|
|
|
* force a schedule.
|
|
|
|
*/
|
|
|
|
clear_bit(__QUEUE_STATE_FROZEN, &txq->state);
|
2009-08-30 04:21:21 +08:00
|
|
|
netif_schedule_queue(txq);
|
2008-08-01 07:58:50 +08:00
|
|
|
}
|
|
|
|
spin_unlock(&dev->tx_global_lock);
|
2006-06-10 03:20:56 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netif_tx_unlock_bh(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 15:34:19 +08:00
|
|
|
netif_tx_unlock(dev);
|
|
|
|
local_bh_enable();
|
2006-06-10 03:20:56 +08:00
|
|
|
}
|
|
|
|
|
2008-07-09 14:13:53 +08:00
|
|
|
#define HARD_TX_LOCK(dev, txq, cpu) { \
|
2007-09-17 05:40:49 +08:00
|
|
|
if ((dev->features & NETIF_F_LLTX) == 0) { \
|
2008-07-09 14:13:53 +08:00
|
|
|
__netif_tx_lock(txq, cpu); \
|
2007-09-17 05:40:49 +08:00
|
|
|
} \
|
|
|
|
}
|
|
|
|
|
2014-03-28 06:42:20 +08:00
|
|
|
#define HARD_TX_TRYLOCK(dev, txq) \
|
|
|
|
(((dev->features & NETIF_F_LLTX) == 0) ? \
|
|
|
|
__netif_tx_trylock(txq) : \
|
|
|
|
true )
|
|
|
|
|
2008-07-09 14:13:53 +08:00
|
|
|
#define HARD_TX_UNLOCK(dev, txq) { \
|
2007-09-17 05:40:49 +08:00
|
|
|
if ((dev->features & NETIF_F_LLTX) == 0) { \
|
2008-07-09 14:13:53 +08:00
|
|
|
__netif_tx_unlock(txq); \
|
2007-09-17 05:40:49 +08:00
|
|
|
} \
|
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
static inline void netif_tx_disable(struct net_device *dev)
|
|
|
|
{
|
2008-07-17 16:56:23 +08:00
|
|
|
unsigned int i;
|
2008-08-01 07:58:50 +08:00
|
|
|
int cpu;
|
2008-07-17 16:56:23 +08:00
|
|
|
|
2008-08-01 07:58:50 +08:00
|
|
|
local_bh_disable();
|
|
|
|
cpu = smp_processor_id();
|
2008-07-17 16:56:23 +08:00
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
2008-08-01 07:58:50 +08:00
|
|
|
|
|
|
|
__netif_tx_lock(txq, cpu);
|
2008-07-17 16:56:23 +08:00
|
|
|
netif_tx_stop_queue(txq);
|
2008-08-01 07:58:50 +08:00
|
|
|
__netif_tx_unlock(txq);
|
2008-07-17 16:56:23 +08:00
|
|
|
}
|
2008-08-01 07:58:50 +08:00
|
|
|
local_bh_enable();
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
2008-07-15 15:13:44 +08:00
|
|
|
static inline void netif_addr_lock(struct net_device *dev)
|
|
|
|
{
|
|
|
|
spin_lock(&dev->addr_list_lock);
|
|
|
|
}
|
|
|
|
|
2012-01-09 14:36:54 +08:00
|
|
|
static inline void netif_addr_lock_nested(struct net_device *dev)
|
|
|
|
{
|
2014-05-17 05:04:54 +08:00
|
|
|
int subclass = SINGLE_DEPTH_NESTING;
|
|
|
|
|
|
|
|
if (dev->netdev_ops->ndo_get_lock_subclass)
|
|
|
|
subclass = dev->netdev_ops->ndo_get_lock_subclass(dev);
|
|
|
|
|
|
|
|
spin_lock_nested(&dev->addr_list_lock, subclass);
|
2012-01-09 14:36:54 +08:00
|
|
|
}
|
|
|
|
|
2008-07-15 15:13:44 +08:00
|
|
|
static inline void netif_addr_lock_bh(struct net_device *dev)
|
|
|
|
{
|
|
|
|
spin_lock_bh(&dev->addr_list_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netif_addr_unlock(struct net_device *dev)
|
|
|
|
{
|
|
|
|
spin_unlock(&dev->addr_list_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netif_addr_unlock_bh(struct net_device *dev)
|
|
|
|
{
|
|
|
|
spin_unlock_bh(&dev->addr_list_lock);
|
|
|
|
}
|
|
|
|
|
2009-05-05 10:48:28 +08:00
|
|
|
/*
|
2009-06-17 09:12:19 +08:00
|
|
|
* dev_addrs walker. Should be used only for read access. Call with
|
2009-05-05 10:48:28 +08:00
|
|
|
* rcu_read_lock held.
|
|
|
|
*/
|
|
|
|
#define for_each_dev_addr(dev, ha) \
|
2009-06-17 09:12:19 +08:00
|
|
|
list_for_each_entry_rcu(ha, &dev->dev_addrs.list, list)
|
2009-05-05 10:48:28 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* These functions live elsewhere (drivers/net/net_init.c, but related) */
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void ether_setup(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Support for loadable net-drivers */
|
2013-09-27 05:48:15 +08:00
|
|
|
struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
|
net: set name_assign_type in alloc_netdev()
Extend alloc_netdev{,_mq{,s}}() to take name_assign_type as argument, and convert
all users to pass NET_NAME_UNKNOWN.
Coccinelle patch:
@@
expression sizeof_priv, name, setup, txqs, rxqs, count;
@@
(
-alloc_netdev_mqs(sizeof_priv, name, setup, txqs, rxqs)
+alloc_netdev_mqs(sizeof_priv, name, NET_NAME_UNKNOWN, setup, txqs, rxqs)
|
-alloc_netdev_mq(sizeof_priv, name, setup, count)
+alloc_netdev_mq(sizeof_priv, name, NET_NAME_UNKNOWN, setup, count)
|
-alloc_netdev(sizeof_priv, name, setup)
+alloc_netdev(sizeof_priv, name, NET_NAME_UNKNOWN, setup)
)
v9: move comments here from the wrong commit
Signed-off-by: Tom Gundersen <teg@jklm.no>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-14 22:37:24 +08:00
|
|
|
unsigned char name_assign_type,
|
2013-09-27 05:48:15 +08:00
|
|
|
void (*setup)(struct net_device *),
|
|
|
|
unsigned int txqs, unsigned int rxqs);
|
net: set name_assign_type in alloc_netdev()
Extend alloc_netdev{,_mq{,s}}() to take name_assign_type as argument, and convert
all users to pass NET_NAME_UNKNOWN.
Coccinelle patch:
@@
expression sizeof_priv, name, setup, txqs, rxqs, count;
@@
(
-alloc_netdev_mqs(sizeof_priv, name, setup, txqs, rxqs)
+alloc_netdev_mqs(sizeof_priv, name, NET_NAME_UNKNOWN, setup, txqs, rxqs)
|
-alloc_netdev_mq(sizeof_priv, name, setup, count)
+alloc_netdev_mq(sizeof_priv, name, NET_NAME_UNKNOWN, setup, count)
|
-alloc_netdev(sizeof_priv, name, setup)
+alloc_netdev(sizeof_priv, name, NET_NAME_UNKNOWN, setup)
)
v9: move comments here from the wrong commit
Signed-off-by: Tom Gundersen <teg@jklm.no>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-14 22:37:24 +08:00
|
|
|
#define alloc_netdev(sizeof_priv, name, name_assign_type, setup) \
|
|
|
|
alloc_netdev_mqs(sizeof_priv, name, name_assign_type, setup, 1, 1)
|
2011-01-10 03:36:31 +08:00
|
|
|
|
net: set name_assign_type in alloc_netdev()
Extend alloc_netdev{,_mq{,s}}() to take name_assign_type as argument, and convert
all users to pass NET_NAME_UNKNOWN.
Coccinelle patch:
@@
expression sizeof_priv, name, setup, txqs, rxqs, count;
@@
(
-alloc_netdev_mqs(sizeof_priv, name, setup, txqs, rxqs)
+alloc_netdev_mqs(sizeof_priv, name, NET_NAME_UNKNOWN, setup, txqs, rxqs)
|
-alloc_netdev_mq(sizeof_priv, name, setup, count)
+alloc_netdev_mq(sizeof_priv, name, NET_NAME_UNKNOWN, setup, count)
|
-alloc_netdev(sizeof_priv, name, setup)
+alloc_netdev(sizeof_priv, name, NET_NAME_UNKNOWN, setup)
)
v9: move comments here from the wrong commit
Signed-off-by: Tom Gundersen <teg@jklm.no>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-14 22:37:24 +08:00
|
|
|
#define alloc_netdev_mq(sizeof_priv, name, name_assign_type, setup, count) \
|
|
|
|
alloc_netdev_mqs(sizeof_priv, name, name_assign_type, setup, count, \
|
|
|
|
count)
|
2011-01-10 03:36:31 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
int register_netdev(struct net_device *dev);
|
|
|
|
void unregister_netdev(struct net_device *dev);
|
2009-05-05 10:48:28 +08:00
|
|
|
|
2010-04-02 05:22:57 +08:00
|
|
|
/* General hardware address lists handling functions */
|
2013-09-27 05:48:15 +08:00
|
|
|
int __hw_addr_sync(struct netdev_hw_addr_list *to_list,
|
|
|
|
struct netdev_hw_addr_list *from_list, int addr_len);
|
|
|
|
void __hw_addr_unsync(struct netdev_hw_addr_list *to_list,
|
|
|
|
struct netdev_hw_addr_list *from_list, int addr_len);
|
2014-05-29 09:44:46 +08:00
|
|
|
int __hw_addr_sync_dev(struct netdev_hw_addr_list *list,
|
|
|
|
struct net_device *dev,
|
|
|
|
int (*sync)(struct net_device *, const unsigned char *),
|
|
|
|
int (*unsync)(struct net_device *,
|
|
|
|
const unsigned char *));
|
|
|
|
void __hw_addr_unsync_dev(struct netdev_hw_addr_list *list,
|
|
|
|
struct net_device *dev,
|
|
|
|
int (*unsync)(struct net_device *,
|
|
|
|
const unsigned char *));
|
2013-09-27 05:48:15 +08:00
|
|
|
void __hw_addr_init(struct netdev_hw_addr_list *list);
|
2010-04-02 05:22:57 +08:00
|
|
|
|
2009-05-05 10:48:28 +08:00
|
|
|
/* Functions used for device addresses handling */
|
2013-09-27 05:48:15 +08:00
|
|
|
int dev_addr_add(struct net_device *dev, const unsigned char *addr,
|
|
|
|
unsigned char addr_type);
|
|
|
|
int dev_addr_del(struct net_device *dev, const unsigned char *addr,
|
|
|
|
unsigned char addr_type);
|
|
|
|
void dev_addr_flush(struct net_device *dev);
|
|
|
|
int dev_addr_init(struct net_device *dev);
|
2010-04-02 05:22:09 +08:00
|
|
|
|
|
|
|
/* Functions used for unicast addresses handling */
|
2013-09-27 05:48:15 +08:00
|
|
|
int dev_uc_add(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_uc_add_excl(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_uc_del(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_uc_sync(struct net_device *to, struct net_device *from);
|
|
|
|
int dev_uc_sync_multiple(struct net_device *to, struct net_device *from);
|
|
|
|
void dev_uc_unsync(struct net_device *to, struct net_device *from);
|
|
|
|
void dev_uc_flush(struct net_device *dev);
|
|
|
|
void dev_uc_init(struct net_device *dev);
|
2009-05-05 10:48:28 +08:00
|
|
|
|
2014-05-29 09:44:46 +08:00
|
|
|
/**
|
|
|
|
* __dev_uc_sync - Synchonize device's unicast list
|
|
|
|
* @dev: device to sync
|
|
|
|
* @sync: function to call if address should be added
|
|
|
|
* @unsync: function to call if address should be removed
|
|
|
|
*
|
|
|
|
* Add newly added addresses to the interface, and release
|
|
|
|
* addresses that have been deleted.
|
|
|
|
**/
|
|
|
|
static inline int __dev_uc_sync(struct net_device *dev,
|
|
|
|
int (*sync)(struct net_device *,
|
|
|
|
const unsigned char *),
|
|
|
|
int (*unsync)(struct net_device *,
|
|
|
|
const unsigned char *))
|
|
|
|
{
|
|
|
|
return __hw_addr_sync_dev(&dev->uc, dev, sync, unsync);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2014-09-04 22:44:36 +08:00
|
|
|
* __dev_uc_unsync - Remove synchronized addresses from device
|
2014-05-29 09:44:46 +08:00
|
|
|
* @dev: device to sync
|
|
|
|
* @unsync: function to call if address should be removed
|
|
|
|
*
|
|
|
|
* Remove all addresses that were added to the device by dev_uc_sync().
|
|
|
|
**/
|
|
|
|
static inline void __dev_uc_unsync(struct net_device *dev,
|
|
|
|
int (*unsync)(struct net_device *,
|
|
|
|
const unsigned char *))
|
|
|
|
{
|
|
|
|
__hw_addr_unsync_dev(&dev->uc, dev, unsync);
|
|
|
|
}
|
|
|
|
|
2010-04-02 05:22:57 +08:00
|
|
|
/* Functions used for multicast addresses handling */
|
2013-09-27 05:48:15 +08:00
|
|
|
int dev_mc_add(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_mc_add_global(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_mc_add_excl(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_mc_del(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_mc_del_global(struct net_device *dev, const unsigned char *addr);
|
|
|
|
int dev_mc_sync(struct net_device *to, struct net_device *from);
|
|
|
|
int dev_mc_sync_multiple(struct net_device *to, struct net_device *from);
|
|
|
|
void dev_mc_unsync(struct net_device *to, struct net_device *from);
|
|
|
|
void dev_mc_flush(struct net_device *dev);
|
|
|
|
void dev_mc_init(struct net_device *dev);
|
2009-05-05 10:48:28 +08:00
|
|
|
|
2014-05-29 09:44:46 +08:00
|
|
|
/**
|
|
|
|
* __dev_mc_sync - Synchonize device's multicast list
|
|
|
|
* @dev: device to sync
|
|
|
|
* @sync: function to call if address should be added
|
|
|
|
* @unsync: function to call if address should be removed
|
|
|
|
*
|
|
|
|
* Add newly added addresses to the interface, and release
|
|
|
|
* addresses that have been deleted.
|
|
|
|
**/
|
|
|
|
static inline int __dev_mc_sync(struct net_device *dev,
|
|
|
|
int (*sync)(struct net_device *,
|
|
|
|
const unsigned char *),
|
|
|
|
int (*unsync)(struct net_device *,
|
|
|
|
const unsigned char *))
|
|
|
|
{
|
|
|
|
return __hw_addr_sync_dev(&dev->mc, dev, sync, unsync);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2014-09-04 22:44:36 +08:00
|
|
|
* __dev_mc_unsync - Remove synchronized addresses from device
|
2014-05-29 09:44:46 +08:00
|
|
|
* @dev: device to sync
|
|
|
|
* @unsync: function to call if address should be removed
|
|
|
|
*
|
|
|
|
* Remove all addresses that were added to the device by dev_mc_sync().
|
|
|
|
**/
|
|
|
|
static inline void __dev_mc_unsync(struct net_device *dev,
|
|
|
|
int (*unsync)(struct net_device *,
|
|
|
|
const unsigned char *))
|
|
|
|
{
|
|
|
|
__hw_addr_unsync_dev(&dev->mc, dev, unsync);
|
|
|
|
}
|
|
|
|
|
2007-06-27 16:28:10 +08:00
|
|
|
/* Functions used for secondary unicast and multicast support */
|
2013-09-27 05:48:15 +08:00
|
|
|
void dev_set_rx_mode(struct net_device *dev);
|
|
|
|
void __dev_set_rx_mode(struct net_device *dev);
|
|
|
|
int dev_set_promiscuity(struct net_device *dev, int inc);
|
|
|
|
int dev_set_allmulti(struct net_device *dev, int inc);
|
|
|
|
void netdev_state_change(struct net_device *dev);
|
|
|
|
void netdev_notify_peers(struct net_device *dev);
|
|
|
|
void netdev_features_change(struct net_device *dev);
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Load a device via the kmod */
|
2013-09-27 05:48:15 +08:00
|
|
|
void dev_load(struct net *net, const char *name);
|
|
|
|
struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev,
|
|
|
|
struct rtnl_link_stats64 *storage);
|
|
|
|
void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64,
|
|
|
|
const struct net_device_stats *netdev_stats);
|
2008-11-20 13:40:23 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
extern int netdev_max_backlog;
|
net: Consistent skb timestamping
With RPS inclusion, skb timestamping is not consistent in RX path.
If netif_receive_skb() is used, its deferred after RPS dispatch.
If netif_rx() is used, its done before RPS dispatch.
This can give strange tcpdump timestamps results.
I think timestamping should be done as soon as possible in the receive
path, to get meaningful values (ie timestamps taken at the time packet
was delivered by NIC driver to our stack), even if NAPI already can
defer timestamping a bit (RPS can help to reduce the gap)
Tom Herbert prefer to sample timestamps after RPS dispatch. In case
sampling is expensive (HPET/acpi_pm on x86), this makes sense.
Let admins switch from one mode to another, using a new
sysctl, /proc/sys/net/core/netdev_tstamp_prequeue
Its default value (1), means timestamps are taken as soon as possible,
before backlog queueing, giving accurate timestamps.
Setting a 0 value permits to sample timestamps when processing backlog,
after RPS dispatch, to lower the load of the pre-RPS cpu.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-16 14:57:10 +08:00
|
|
|
extern int netdev_tstamp_prequeue;
|
2005-04-17 06:20:36 +08:00
|
|
|
extern int weight_p;
|
2011-04-20 17:27:32 +08:00
|
|
|
extern int bpf_jit_enable;
|
2013-01-04 06:48:49 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
bool netdev_has_upper_dev(struct net_device *dev, struct net_device *upper_dev);
|
2014-05-17 05:20:38 +08:00
|
|
|
struct net_device *netdev_upper_get_next_dev_rcu(struct net_device *dev,
|
|
|
|
struct list_head **iter);
|
2013-09-27 05:48:15 +08:00
|
|
|
struct net_device *netdev_all_upper_get_next_dev_rcu(struct net_device *dev,
|
|
|
|
struct list_head **iter);
|
2013-08-29 05:25:08 +08:00
|
|
|
|
2014-05-17 05:20:38 +08:00
|
|
|
/* iterate through upper list, must be called under RCU read lock */
|
|
|
|
#define netdev_for_each_upper_dev_rcu(dev, updev, iter) \
|
|
|
|
for (iter = &(dev)->adj_list.upper, \
|
|
|
|
updev = netdev_upper_get_next_dev_rcu(dev, &(iter)); \
|
|
|
|
updev; \
|
|
|
|
updev = netdev_upper_get_next_dev_rcu(dev, &(iter)))
|
|
|
|
|
2013-08-29 05:25:08 +08:00
|
|
|
/* iterate through upper list, must be called under RCU read lock */
|
net: add adj_list to save only neighbours
Currently, we distinguish neighbours (first-level linked devices) from
non-neighbours by the neighbour bool in the netdev_adjacent. This could be
quite time-consuming in case we would like to traverse *only* through
neighbours - cause we'd have to traverse through all devices and check for
this flag, and in a (quite common) scenario where we have lots of vlans on
top of bridge, which is on top of a bond - the bonding would have to go
through all those vlans to get its upper neighbour linked devices.
This situation is really unpleasant, cause there are already a lot of cases
when a device with slaves needs to go through them in hot path.
To fix this, introduce a new upper/lower device lists structure -
adj_list, which contains only the neighbours. It works always in
pair with the all_adj_list structure (renamed from upper/lower_dev_list),
i.e. both of them contain the same links, only that all_adj_list contains
also non-neighbour device links. It's really a small change visible,
currently, only for __netdev_adjacent_dev_insert/remove(), and doesn't
change the main linked logic at all.
Also, add some comments a fix a name collision in
netdev_for_each_upper_dev_rcu() and rework the naming by the following
rules:
netdev_(all_)(upper|lower)_*
If "all_" is present, then we work with the whole list of upper/lower
devices, otherwise - only with direct neighbours. Uninline functions - to
get better stack traces.
CC: "David S. Miller" <davem@davemloft.net>
CC: Eric Dumazet <edumazet@google.com>
CC: Jiri Pirko <jiri@resnulli.us>
CC: Alexander Duyck <alexander.h.duyck@intel.com>
CC: Cong Wang <amwang@redhat.com>
Signed-off-by: Veaceslav Falico <vfalico@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-09-25 15:20:07 +08:00
|
|
|
#define netdev_for_each_all_upper_dev_rcu(dev, updev, iter) \
|
|
|
|
for (iter = &(dev)->all_adj_list.upper, \
|
|
|
|
updev = netdev_all_upper_get_next_dev_rcu(dev, &(iter)); \
|
|
|
|
updev; \
|
|
|
|
updev = netdev_all_upper_get_next_dev_rcu(dev, &(iter)))
|
2013-08-29 05:25:08 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void *netdev_lower_get_next_private(struct net_device *dev,
|
|
|
|
struct list_head **iter);
|
|
|
|
void *netdev_lower_get_next_private_rcu(struct net_device *dev,
|
|
|
|
struct list_head **iter);
|
2013-09-25 15:20:12 +08:00
|
|
|
|
|
|
|
#define netdev_for_each_lower_private(dev, priv, iter) \
|
|
|
|
for (iter = (dev)->adj_list.lower.next, \
|
|
|
|
priv = netdev_lower_get_next_private(dev, &(iter)); \
|
|
|
|
priv; \
|
|
|
|
priv = netdev_lower_get_next_private(dev, &(iter)))
|
|
|
|
|
|
|
|
#define netdev_for_each_lower_private_rcu(dev, priv, iter) \
|
|
|
|
for (iter = &(dev)->adj_list.lower, \
|
|
|
|
priv = netdev_lower_get_next_private_rcu(dev, &(iter)); \
|
|
|
|
priv; \
|
|
|
|
priv = netdev_lower_get_next_private_rcu(dev, &(iter)))
|
|
|
|
|
2014-05-17 05:04:53 +08:00
|
|
|
void *netdev_lower_get_next(struct net_device *dev,
|
|
|
|
struct list_head **iter);
|
|
|
|
#define netdev_for_each_lower_dev(dev, ldev, iter) \
|
|
|
|
for (iter = &(dev)->adj_list.lower, \
|
|
|
|
ldev = netdev_lower_get_next(dev, &(iter)); \
|
|
|
|
ldev; \
|
|
|
|
ldev = netdev_lower_get_next(dev, &(iter)))
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void *netdev_adjacent_get_private(struct list_head *adj_list);
|
2013-12-13 10:19:55 +08:00
|
|
|
void *netdev_lower_get_first_private_rcu(struct net_device *dev);
|
2013-09-27 05:48:15 +08:00
|
|
|
struct net_device *netdev_master_upper_dev_get(struct net_device *dev);
|
|
|
|
struct net_device *netdev_master_upper_dev_get_rcu(struct net_device *dev);
|
|
|
|
int netdev_upper_dev_link(struct net_device *dev, struct net_device *upper_dev);
|
|
|
|
int netdev_master_upper_dev_link(struct net_device *dev,
|
2013-01-04 06:48:49 +08:00
|
|
|
struct net_device *upper_dev);
|
2013-09-27 05:48:15 +08:00
|
|
|
int netdev_master_upper_dev_link_private(struct net_device *dev,
|
|
|
|
struct net_device *upper_dev,
|
|
|
|
void *private);
|
|
|
|
void netdev_upper_dev_unlink(struct net_device *dev,
|
|
|
|
struct net_device *upper_dev);
|
2014-01-15 04:58:51 +08:00
|
|
|
void netdev_adjacent_rename_links(struct net_device *dev, char *oldname);
|
2013-09-27 05:48:15 +08:00
|
|
|
void *netdev_lower_dev_get_private(struct net_device *dev,
|
|
|
|
struct net_device *lower_dev);
|
2014-11-16 22:23:05 +08:00
|
|
|
|
|
|
|
/* RSS keys are 40 or 52 bytes long */
|
|
|
|
#define NETDEV_RSS_KEY_LEN 52
|
|
|
|
extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN];
|
|
|
|
void netdev_rss_key_fill(void *buffer, size_t len);
|
|
|
|
|
2014-05-17 05:04:53 +08:00
|
|
|
int dev_get_nest_level(struct net_device *dev,
|
|
|
|
bool (*type_check)(struct net_device *dev));
|
2013-09-27 05:48:15 +08:00
|
|
|
int skb_checksum_help(struct sk_buff *skb);
|
|
|
|
struct sk_buff *__skb_gso_segment(struct sk_buff *skb,
|
|
|
|
netdev_features_t features, bool tx_path);
|
|
|
|
struct sk_buff *skb_mac_gso_segment(struct sk_buff *skb,
|
|
|
|
netdev_features_t features);
|
2013-02-06 00:36:38 +08:00
|
|
|
|
2015-02-03 22:48:29 +08:00
|
|
|
struct netdev_bonding_info {
|
|
|
|
ifslave slave;
|
|
|
|
ifbond master;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct netdev_notifier_bonding_info {
|
|
|
|
struct netdev_notifier_info info; /* must be first */
|
|
|
|
struct netdev_bonding_info bonding_info;
|
|
|
|
};
|
|
|
|
|
|
|
|
void netdev_bonding_info_change(struct net_device *dev,
|
|
|
|
struct netdev_bonding_info *bonding_info);
|
|
|
|
|
2013-02-06 00:36:38 +08:00
|
|
|
static inline
|
|
|
|
struct sk_buff *skb_gso_segment(struct sk_buff *skb, netdev_features_t features)
|
|
|
|
{
|
|
|
|
return __skb_gso_segment(skb, features, true);
|
|
|
|
}
|
2014-03-28 05:26:18 +08:00
|
|
|
__be16 skb_network_protocol(struct sk_buff *skb, int *depth);
|
2013-03-07 17:28:01 +08:00
|
|
|
|
|
|
|
static inline bool can_checksum_protocol(netdev_features_t features,
|
|
|
|
__be16 protocol)
|
|
|
|
{
|
|
|
|
return ((features & NETIF_F_GEN_CSUM) ||
|
|
|
|
((features & NETIF_F_V4_CSUM) &&
|
|
|
|
protocol == htons(ETH_P_IP)) ||
|
|
|
|
((features & NETIF_F_V6_CSUM) &&
|
|
|
|
protocol == htons(ETH_P_IPV6)) ||
|
|
|
|
((features & NETIF_F_FCOE_CRC) &&
|
|
|
|
protocol == htons(ETH_P_FCOE)));
|
|
|
|
}
|
2013-02-06 00:36:38 +08:00
|
|
|
|
2005-11-11 05:01:24 +08:00
|
|
|
#ifdef CONFIG_BUG
|
2013-09-27 05:48:15 +08:00
|
|
|
void netdev_rx_csum_fault(struct net_device *dev);
|
2005-11-11 05:01:24 +08:00
|
|
|
#else
|
|
|
|
static inline void netdev_rx_csum_fault(struct net_device *dev)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif
|
2005-04-17 06:20:36 +08:00
|
|
|
/* rx skb timestamps */
|
2013-09-27 05:48:15 +08:00
|
|
|
void net_enable_timestamp(void);
|
|
|
|
void net_disable_timestamp(void);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-08-16 13:18:02 +08:00
|
|
|
#ifdef CONFIG_PROC_FS
|
2013-09-27 05:48:15 +08:00
|
|
|
int __init dev_proc_init(void);
|
2013-02-19 03:20:33 +08:00
|
|
|
#else
|
|
|
|
#define dev_proc_init() 0
|
2005-08-16 13:18:02 +08:00
|
|
|
#endif
|
|
|
|
|
2014-08-23 07:21:53 +08:00
|
|
|
static inline netdev_tx_t __netdev_start_xmit(const struct net_device_ops *ops,
|
2014-08-30 12:55:22 +08:00
|
|
|
struct sk_buff *skb, struct net_device *dev,
|
|
|
|
bool more)
|
2014-08-23 07:21:53 +08:00
|
|
|
{
|
2014-08-30 12:55:22 +08:00
|
|
|
skb->xmit_more = more ? 1 : 0;
|
2014-08-26 06:51:53 +08:00
|
|
|
return ops->ndo_start_xmit(skb, dev);
|
2014-08-23 07:21:53 +08:00
|
|
|
}
|
|
|
|
|
2014-08-30 12:07:24 +08:00
|
|
|
static inline netdev_tx_t netdev_start_xmit(struct sk_buff *skb, struct net_device *dev,
|
2014-08-30 12:55:22 +08:00
|
|
|
struct netdev_queue *txq, bool more)
|
2014-08-23 07:21:53 +08:00
|
|
|
{
|
|
|
|
const struct net_device_ops *ops = dev->netdev_ops;
|
2014-08-30 12:07:24 +08:00
|
|
|
int rc;
|
2014-08-23 07:21:53 +08:00
|
|
|
|
2014-08-30 12:55:22 +08:00
|
|
|
rc = __netdev_start_xmit(ops, skb, dev, more);
|
2014-08-30 12:07:24 +08:00
|
|
|
if (rc == NETDEV_TX_OK)
|
|
|
|
txq_trans_update(txq);
|
|
|
|
|
|
|
|
return rc;
|
2014-08-23 07:21:53 +08:00
|
|
|
}
|
|
|
|
|
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next
Pull networking updates from David Miller:
1) The addition of nftables. No longer will we need protocol aware
firewall filtering modules, it can all live in userspace.
At the core of nftables is a, for lack of a better term, virtual
machine that executes byte codes to inspect packet or metadata
(arriving interface index, etc.) and make verdict decisions.
Besides support for loading packet contents and comparing them, the
interpreter supports lookups in various datastructures as
fundamental operations. For example sets are supports, and
therefore one could create a set of whitelist IP address entries
which have ACCEPT verdicts attached to them, and use the appropriate
byte codes to do such lookups.
Since the interpreted code is composed in userspace, userspace can
do things like optimize things before giving it to the kernel.
Another major improvement is the capability of atomically updating
portions of the ruleset. In the existing netfilter implementation,
one has to update the entire rule set in order to make a change and
this is very expensive.
Userspace tools exist to create nftables rules using existing
netfilter rule sets, but both kernel implementations will need to
co-exist for quite some time as we transition from the old to the
new stuff.
Kudos to Patrick McHardy, Pablo Neira Ayuso, and others who have
worked so hard on this.
2) Daniel Borkmann and Hannes Frederic Sowa made several improvements
to our pseudo-random number generator, mostly used for things like
UDP port randomization and netfitler, amongst other things.
In particular the taus88 generater is updated to taus113, and test
cases are added.
3) Support 64-bit rates in HTB and TBF schedulers, from Eric Dumazet
and Yang Yingliang.
4) Add support for new 577xx tigon3 chips to tg3 driver, from Nithin
Sujir.
5) Fix two fatal flaws in TCP dynamic right sizing, from Eric Dumazet,
Neal Cardwell, and Yuchung Cheng.
6) Allow IP_TOS and IP_TTL to be specified in sendmsg() ancillary
control message data, much like other socket option attributes.
From Francesco Fusco.
7) Allow applications to specify a cap on the rate computed
automatically by the kernel for pacing flows, via a new
SO_MAX_PACING_RATE socket option. From Eric Dumazet.
8) Make the initial autotuned send buffer sizing in TCP more closely
reflect actual needs, from Eric Dumazet.
9) Currently early socket demux only happens for TCP sockets, but we
can do it for connected UDP sockets too. Implementation from Shawn
Bohrer.
10) Refactor inet socket demux with the goal of improving hash demux
performance for listening sockets. With the main goals being able
to use RCU lookups on even request sockets, and eliminating the
listening lock contention. From Eric Dumazet.
11) The bonding layer has many demuxes in it's fast path, and an RCU
conversion was started back in 3.11, several changes here extend the
RCU usage to even more locations. From Ding Tianhong and Wang
Yufen, based upon suggestions by Nikolay Aleksandrov and Veaceslav
Falico.
12) Allow stackability of segmentation offloads to, in particular, allow
segmentation offloading over tunnels. From Eric Dumazet.
13) Significantly improve the handling of secret keys we input into the
various hash functions in the inet hashtables, TCP fast open, as
well as syncookies. From Hannes Frederic Sowa. The key fundamental
operation is "net_get_random_once()" which uses static keys.
Hannes even extended this to ipv4/ipv6 fragmentation handling and
our generic flow dissector.
14) The generic driver layer takes care now to set the driver data to
NULL on device removal, so it's no longer necessary for drivers to
explicitly set it to NULL any more. Many drivers have been cleaned
up in this way, from Jingoo Han.
15) Add a BPF based packet scheduler classifier, from Daniel Borkmann.
16) Improve CRC32 interfaces and generic SKB checksum iterators so that
SCTP's checksumming can more cleanly be handled. Also from Daniel
Borkmann.
17) Add a new PMTU discovery mode, IP_PMTUDISC_INTERFACE, which forces
using the interface MTU value. This helps avoid PMTU attacks,
particularly on DNS servers. From Hannes Frederic Sowa.
18) Use generic XPS for transmit queue steering rather than internal
(re-)implementation in virtio-net. From Jason Wang.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1622 commits)
random32: add test cases for taus113 implementation
random32: upgrade taus88 generator to taus113 from errata paper
random32: move rnd_state to linux/random.h
random32: add prandom_reseed_late() and call when nonblocking pool becomes initialized
random32: add periodic reseeding
random32: fix off-by-one in seeding requirement
PHY: Add RTL8201CP phy_driver to realtek
xtsonic: add missing platform_set_drvdata() in xtsonic_probe()
macmace: add missing platform_set_drvdata() in mace_probe()
ethernet/arc/arc_emac: add missing platform_set_drvdata() in arc_emac_probe()
ipv6: protect for_each_sk_fl_rcu in mem_check with rcu_read_lock_bh
vlan: Implement vlan_dev_get_egress_qos_mask as an inline.
ixgbe: add warning when max_vfs is out of range.
igb: Update link modes display in ethtool
netfilter: push reasm skb through instead of original frag skbs
ip6_output: fragment outgoing reassembled skb properly
MAINTAINERS: mv643xx_eth: take over maintainership from Lennart
net_sched: tbf: support of 64bit rates
ixgbe: deleting dfwd stations out of order can cause null ptr deref
ixgbe: fix build err, num_rx_queues is only available with CONFIG_RPS
...
2013-11-13 16:40:34 +08:00
|
|
|
int netdev_class_create_file_ns(struct class_attribute *class_attr,
|
|
|
|
const void *ns);
|
|
|
|
void netdev_class_remove_file_ns(struct class_attribute *class_attr,
|
|
|
|
const void *ns);
|
2013-09-12 10:29:04 +08:00
|
|
|
|
|
|
|
static inline int netdev_class_create_file(struct class_attribute *class_attr)
|
|
|
|
{
|
|
|
|
return netdev_class_create_file_ns(class_attr, NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void netdev_class_remove_file(struct class_attribute *class_attr)
|
|
|
|
{
|
|
|
|
netdev_class_remove_file_ns(class_attr, NULL);
|
|
|
|
}
|
2008-06-14 09:12:04 +08:00
|
|
|
|
2010-08-05 23:45:15 +08:00
|
|
|
extern struct kobj_ns_type_operations net_ns_type_operations;
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
const char *netdev_drivername(const struct net_device *dev);
|
2008-07-22 04:31:48 +08:00
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
void linkwatch_run_queue(void);
|
2005-08-16 13:18:02 +08:00
|
|
|
|
2014-05-20 14:29:25 +08:00
|
|
|
static inline netdev_features_t netdev_intersect_features(netdev_features_t f1,
|
|
|
|
netdev_features_t f2)
|
|
|
|
{
|
|
|
|
if (f1 & NETIF_F_GEN_CSUM)
|
|
|
|
f1 |= (NETIF_F_ALL_CSUM & ~NETIF_F_GEN_CSUM);
|
|
|
|
if (f2 & NETIF_F_GEN_CSUM)
|
|
|
|
f2 |= (NETIF_F_ALL_CSUM & ~NETIF_F_GEN_CSUM);
|
|
|
|
f1 &= f2;
|
|
|
|
if (f1 & NETIF_F_GEN_CSUM)
|
|
|
|
f1 &= ~(NETIF_F_ALL_CSUM & ~NETIF_F_GEN_CSUM);
|
|
|
|
|
|
|
|
return f1;
|
|
|
|
}
|
|
|
|
|
2011-11-15 23:29:55 +08:00
|
|
|
static inline netdev_features_t netdev_get_wanted_features(
|
|
|
|
struct net_device *dev)
|
2011-02-16 00:59:17 +08:00
|
|
|
{
|
|
|
|
return (dev->features & ~dev->hw_features) | dev->wanted_features;
|
|
|
|
}
|
2011-11-15 23:29:55 +08:00
|
|
|
netdev_features_t netdev_increment_features(netdev_features_t all,
|
|
|
|
netdev_features_t one, netdev_features_t mask);
|
2013-05-16 15:34:53 +08:00
|
|
|
|
|
|
|
/* Allow TSO being used on stacked device :
|
|
|
|
* Performing the GSO segmentation before last device
|
|
|
|
* is a performance improvement.
|
|
|
|
*/
|
|
|
|
static inline netdev_features_t netdev_add_tso_features(netdev_features_t features,
|
|
|
|
netdev_features_t mask)
|
|
|
|
{
|
|
|
|
return netdev_increment_features(features, NETIF_F_ALL_TSO, mask);
|
|
|
|
}
|
|
|
|
|
2011-04-03 13:48:47 +08:00
|
|
|
int __netdev_update_features(struct net_device *dev);
|
2011-02-16 00:59:17 +08:00
|
|
|
void netdev_update_features(struct net_device *dev);
|
2011-05-07 11:22:17 +08:00
|
|
|
void netdev_change_features(struct net_device *dev);
|
2007-08-11 06:47:58 +08:00
|
|
|
|
2009-12-04 07:59:22 +08:00
|
|
|
void netif_stacked_transfer_operstate(const struct net_device *rootdev,
|
|
|
|
struct net_device *dev);
|
|
|
|
|
2015-03-27 13:31:13 +08:00
|
|
|
netdev_features_t passthru_features_check(struct sk_buff *skb,
|
|
|
|
struct net_device *dev,
|
|
|
|
netdev_features_t features);
|
2014-05-05 21:00:44 +08:00
|
|
|
netdev_features_t netif_skb_features(struct sk_buff *skb);
|
2010-10-29 20:14:55 +08:00
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool net_gso_ok(netdev_features_t features, int gso_type)
|
2006-06-28 04:22:38 +08:00
|
|
|
{
|
2011-11-15 23:29:55 +08:00
|
|
|
netdev_features_t feature = gso_type << NETIF_F_GSO_SHIFT;
|
2011-11-16 22:05:33 +08:00
|
|
|
|
|
|
|
/* check flags correspondence */
|
|
|
|
BUILD_BUG_ON(SKB_GSO_TCPV4 != (NETIF_F_TSO >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_UDP != (NETIF_F_UFO >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_DODGY != (NETIF_F_GSO_ROBUST >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_TCP_ECN != (NETIF_F_TSO_ECN >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_TCPV6 != (NETIF_F_TSO6 >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_FCOE != (NETIF_F_FSO >> NETIF_F_GSO_SHIFT));
|
2014-06-15 14:23:52 +08:00
|
|
|
BUILD_BUG_ON(SKB_GSO_GRE != (NETIF_F_GSO_GRE >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_GRE_CSUM != (NETIF_F_GSO_GRE_CSUM >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_IPIP != (NETIF_F_GSO_IPIP >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_SIT != (NETIF_F_GSO_SIT >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_UDP_TUNNEL != (NETIF_F_GSO_UDP_TUNNEL >> NETIF_F_GSO_SHIFT));
|
|
|
|
BUILD_BUG_ON(SKB_GSO_UDP_TUNNEL_CSUM != (NETIF_F_GSO_UDP_TUNNEL_CSUM >> NETIF_F_GSO_SHIFT));
|
2014-11-05 01:06:54 +08:00
|
|
|
BUILD_BUG_ON(SKB_GSO_TUNNEL_REMCSUM != (NETIF_F_GSO_TUNNEL_REMCSUM >> NETIF_F_GSO_SHIFT));
|
2011-11-16 22:05:33 +08:00
|
|
|
|
2006-06-30 03:25:53 +08:00
|
|
|
return (features & feature) == feature;
|
2006-06-28 04:22:38 +08:00
|
|
|
}
|
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool skb_gso_ok(struct sk_buff *skb, netdev_features_t features)
|
2006-07-01 04:36:35 +08:00
|
|
|
{
|
2009-06-04 12:20:51 +08:00
|
|
|
return net_gso_ok(features, skb_shinfo(skb)->gso_type) &&
|
2010-08-23 15:13:46 +08:00
|
|
|
(!skb_has_frag_list(skb) || (features & NETIF_F_FRAGLIST));
|
2006-07-01 04:36:35 +08:00
|
|
|
}
|
|
|
|
|
2015-04-17 21:45:04 +08:00
|
|
|
static inline bool netif_needs_gso(struct sk_buff *skb,
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
netdev_features_t features)
|
2006-06-22 17:40:14 +08:00
|
|
|
{
|
2011-01-09 14:23:32 +08:00
|
|
|
return skb_is_gso(skb) && (!skb_gso_ok(skb, features) ||
|
2012-03-17 07:08:11 +08:00
|
|
|
unlikely((skb->ip_summed != CHECKSUM_PARTIAL) &&
|
|
|
|
(skb->ip_summed != CHECKSUM_UNNECESSARY)));
|
2006-06-22 17:40:14 +08:00
|
|
|
}
|
|
|
|
|
[NET]: Add per-connection option to set max TSO frame size
Update: My mailer ate one of Jarek's feedback mails... Fixed the
parameter in netif_set_gso_max_size() to be u32, not u16. Fixed the
whitespace issue due to a patch import botch. Changed the types from
u32 to unsigned int to be more consistent with other variables in the
area. Also brought the patch up to the latest net-2.6.26 tree.
Update: Made gso_max_size container 32 bits, not 16. Moved the
location of gso_max_size within netdev to be less hotpath. Made more
consistent names between the sock and netdev layers, and added a
define for the max GSO size.
Update: Respun for net-2.6.26 tree.
Update: changed max_gso_frame_size and sk_gso_max_size from signed to
unsigned - thanks Stephen!
This patch adds the ability for device drivers to control the size of
the TSO frames being sent to them, per TCP connection. By setting the
netdevice's gso_max_size value, the socket layer will set the GSO
frame size based on that value. This will propogate into the TCP
layer, and send TSO's of that size to the hardware.
This can be desirable to help tune the bursty nature of TSO on a
per-adapter basis, where one may have 1 GbE and 10 GbE devices
coexisting in a system, one running multiqueue and the other not, etc.
This can also be desirable for devices that cannot support full 64 KB
TSO's, but still want to benefit from some level of segmentation
offloading.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-21 18:43:19 +08:00
|
|
|
static inline void netif_set_gso_max_size(struct net_device *dev,
|
|
|
|
unsigned int size)
|
|
|
|
{
|
|
|
|
dev->gso_max_size = size;
|
|
|
|
}
|
|
|
|
|
2013-12-27 05:10:22 +08:00
|
|
|
static inline void skb_gso_error_unwind(struct sk_buff *skb, __be16 protocol,
|
|
|
|
int pulled_hlen, u16 mac_offset,
|
|
|
|
int mac_len)
|
|
|
|
{
|
|
|
|
skb->protocol = protocol;
|
|
|
|
skb->encapsulation = 1;
|
|
|
|
skb_push(skb, pulled_hlen);
|
|
|
|
skb_reset_transport_header(skb);
|
|
|
|
skb->mac_header = mac_offset;
|
|
|
|
skb->network_header = skb->mac_header + mac_len;
|
|
|
|
skb->mac_len = mac_len;
|
|
|
|
}
|
|
|
|
|
2013-11-07 01:54:46 +08:00
|
|
|
static inline bool netif_is_macvlan(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_MACVLAN;
|
|
|
|
}
|
|
|
|
|
2014-12-07 07:53:04 +08:00
|
|
|
static inline bool netif_is_macvlan_port(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_MACVLAN_PORT;
|
|
|
|
}
|
|
|
|
|
2014-12-07 07:53:33 +08:00
|
|
|
static inline bool netif_is_ipvlan(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_IPVLAN_SLAVE;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool netif_is_ipvlan_port(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_IPVLAN_MASTER;
|
|
|
|
}
|
|
|
|
|
2013-03-12 10:49:01 +08:00
|
|
|
static inline bool netif_is_bond_master(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->flags & IFF_MASTER && dev->priv_flags & IFF_BONDING;
|
|
|
|
}
|
|
|
|
|
net: Use bool in netdevice.h helpers.
Specifically use it in napi_disable_pending(), napi_schedule_prep(),
napi_reschedule(), netif_tx_queue_stopped(), netif_queue_stopped(),
netif_xmit_stopped(), netif_xmit_frozen_or_stopped(), netif_running(),
__netif_subqueue_stopped(), netif_subqueue_stopped(),
netif_is_multiquue(), netif_carrier_ok(), netif_dormant(),
netif_oper_up(), netif_device_present(), __netif_tx_trylock(),
net_gso_ok(), skb_gso_ok(), netif_needs_gso(), and
netif_is_bond_slave().
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-03-08 10:02:35 +08:00
|
|
|
static inline bool netif_is_bond_slave(struct net_device *dev)
|
2011-02-12 14:48:36 +08:00
|
|
|
{
|
|
|
|
return dev->flags & IFF_SLAVE && dev->priv_flags & IFF_BONDING;
|
|
|
|
}
|
|
|
|
|
2012-02-11 23:39:30 +08:00
|
|
|
static inline bool netif_supports_nofcs(struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_SUPP_NOFCS;
|
|
|
|
}
|
|
|
|
|
2015-08-14 04:59:00 +08:00
|
|
|
static inline bool netif_is_vrf(const struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_VRF_MASTER;
|
|
|
|
}
|
|
|
|
|
2015-08-27 15:31:19 +08:00
|
|
|
static inline bool netif_is_bridge_master(const struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_EBRIDGE;
|
|
|
|
}
|
|
|
|
|
2015-08-27 15:31:20 +08:00
|
|
|
static inline bool netif_is_ovs_master(const struct net_device *dev)
|
|
|
|
{
|
|
|
|
return dev->priv_flags & IFF_OPENVSWITCH;
|
|
|
|
}
|
|
|
|
|
2015-08-14 04:59:00 +08:00
|
|
|
static inline bool netif_index_is_vrf(struct net *net, int ifindex)
|
|
|
|
{
|
|
|
|
bool rc = false;
|
|
|
|
|
2015-08-16 21:49:20 +08:00
|
|
|
#if IS_ENABLED(CONFIG_NET_VRF)
|
|
|
|
struct net_device *dev;
|
|
|
|
|
|
|
|
if (ifindex == 0)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
|
|
|
|
dev = dev_get_by_index_rcu(net, ifindex);
|
2015-08-14 04:59:00 +08:00
|
|
|
if (dev)
|
|
|
|
rc = netif_is_vrf(dev);
|
|
|
|
|
2015-08-16 21:49:20 +08:00
|
|
|
rcu_read_unlock();
|
|
|
|
#endif
|
2015-08-14 04:59:00 +08:00
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
2014-10-06 09:38:35 +08:00
|
|
|
/* This device needs to keep skb dst for qdisc enqueue or ndo_start_xmit() */
|
|
|
|
static inline void netif_keep_dst(struct net_device *dev)
|
|
|
|
{
|
|
|
|
dev->priv_flags &= ~(IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM);
|
|
|
|
}
|
|
|
|
|
2008-11-08 14:54:20 +08:00
|
|
|
extern struct pernet_operations __net_initdata loopback_net_ops;
|
2009-04-20 12:49:28 +08:00
|
|
|
|
2010-02-09 19:49:47 +08:00
|
|
|
/* Logging, debugging and troubleshooting/diagnostic helpers. */
|
|
|
|
|
|
|
|
/* netdev_printk helpers, similar to dev_printk */
|
|
|
|
|
|
|
|
static inline const char *netdev_name(const struct net_device *dev)
|
|
|
|
{
|
2014-07-18 01:46:09 +08:00
|
|
|
if (!dev->name[0] || strchr(dev->name, '%'))
|
|
|
|
return "(unnamed net_device)";
|
2010-02-09 19:49:47 +08:00
|
|
|
return dev->name;
|
|
|
|
}
|
|
|
|
|
2014-07-18 01:46:10 +08:00
|
|
|
static inline const char *netdev_reg_state(const struct net_device *dev)
|
|
|
|
{
|
|
|
|
switch (dev->reg_state) {
|
|
|
|
case NETREG_UNINITIALIZED: return " (uninitialized)";
|
|
|
|
case NETREG_REGISTERED: return "";
|
|
|
|
case NETREG_UNREGISTERING: return " (unregistering)";
|
|
|
|
case NETREG_UNREGISTERED: return " (unregistered)";
|
|
|
|
case NETREG_RELEASED: return " (released)";
|
|
|
|
case NETREG_DUMMY: return " (dummy)";
|
|
|
|
}
|
|
|
|
|
|
|
|
WARN_ONCE(1, "%s: unknown reg_state %d\n", dev->name, dev->reg_state);
|
|
|
|
return " (unknown)";
|
|
|
|
}
|
|
|
|
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(3, 4)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_printk(const char *level, const struct net_device *dev,
|
|
|
|
const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_emerg(const struct net_device *dev, const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_alert(const struct net_device *dev, const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_crit(const struct net_device *dev, const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_err(const struct net_device *dev, const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_warn(const struct net_device *dev, const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_notice(const struct net_device *dev, const char *format, ...);
|
2013-09-27 05:48:15 +08:00
|
|
|
__printf(2, 3)
|
2014-09-23 02:10:50 +08:00
|
|
|
void netdev_info(const struct net_device *dev, const char *format, ...);
|
2010-02-09 19:49:47 +08:00
|
|
|
|
2011-03-02 05:33:13 +08:00
|
|
|
#define MODULE_ALIAS_NETDEV(device) \
|
|
|
|
MODULE_ALIAS("netdev-" device)
|
|
|
|
|
2011-12-20 06:11:18 +08:00
|
|
|
#if defined(CONFIG_DYNAMIC_DEBUG)
|
2010-02-09 19:49:47 +08:00
|
|
|
#define netdev_dbg(__dev, format, args...) \
|
|
|
|
do { \
|
2011-08-12 02:36:48 +08:00
|
|
|
dynamic_netdev_dbg(__dev, format, ##args); \
|
2010-02-09 19:49:47 +08:00
|
|
|
} while (0)
|
2011-12-20 06:11:18 +08:00
|
|
|
#elif defined(DEBUG)
|
|
|
|
#define netdev_dbg(__dev, format, args...) \
|
|
|
|
netdev_printk(KERN_DEBUG, __dev, format, ##args)
|
2010-02-09 19:49:47 +08:00
|
|
|
#else
|
|
|
|
#define netdev_dbg(__dev, format, args...) \
|
|
|
|
({ \
|
|
|
|
if (0) \
|
|
|
|
netdev_printk(KERN_DEBUG, __dev, format, ##args); \
|
|
|
|
})
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#if defined(VERBOSE_DEBUG)
|
|
|
|
#define netdev_vdbg netdev_dbg
|
|
|
|
#else
|
|
|
|
|
|
|
|
#define netdev_vdbg(dev, format, args...) \
|
|
|
|
({ \
|
|
|
|
if (0) \
|
|
|
|
netdev_printk(KERN_DEBUG, dev, format, ##args); \
|
|
|
|
0; \
|
|
|
|
})
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/*
|
|
|
|
* netdev_WARN() acts like dev_printk(), but with the key difference
|
|
|
|
* of using a WARN/WARN_ON to get the message out, including the
|
|
|
|
* file/line information and a backtrace.
|
|
|
|
*/
|
|
|
|
#define netdev_WARN(dev, format, args...) \
|
2014-07-18 01:46:10 +08:00
|
|
|
WARN(1, "netdevice: %s%s\n" format, netdev_name(dev), \
|
|
|
|
netdev_reg_state(dev), ##args)
|
2010-02-09 19:49:47 +08:00
|
|
|
|
2010-02-09 19:49:49 +08:00
|
|
|
/* netif printk helpers, similar to netdev_printk */
|
|
|
|
|
|
|
|
#define netif_printk(priv, type, level, dev, fmt, args...) \
|
|
|
|
do { \
|
|
|
|
if (netif_msg_##type(priv)) \
|
|
|
|
netdev_printk(level, (dev), fmt, ##args); \
|
|
|
|
} while (0)
|
|
|
|
|
2010-06-27 09:02:36 +08:00
|
|
|
#define netif_level(level, priv, type, dev, fmt, args...) \
|
|
|
|
do { \
|
|
|
|
if (netif_msg_##type(priv)) \
|
|
|
|
netdev_##level(dev, fmt, ##args); \
|
|
|
|
} while (0)
|
|
|
|
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_emerg(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(emerg, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_alert(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(alert, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_crit(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(crit, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_err(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(err, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_warn(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(warn, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_notice(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(notice, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_info(priv, type, dev, fmt, args...) \
|
2010-06-27 09:02:36 +08:00
|
|
|
netif_level(info, priv, type, dev, fmt, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
|
2012-05-30 15:43:34 +08:00
|
|
|
#if defined(CONFIG_DYNAMIC_DEBUG)
|
2010-02-09 19:49:49 +08:00
|
|
|
#define netif_dbg(priv, type, netdev, format, args...) \
|
|
|
|
do { \
|
|
|
|
if (netif_msg_##type(priv)) \
|
2011-08-12 02:36:53 +08:00
|
|
|
dynamic_netdev_dbg(netdev, format, ##args); \
|
2010-02-09 19:49:49 +08:00
|
|
|
} while (0)
|
2012-05-30 15:43:34 +08:00
|
|
|
#elif defined(DEBUG)
|
|
|
|
#define netif_dbg(priv, type, dev, format, args...) \
|
|
|
|
netif_printk(priv, type, KERN_DEBUG, dev, format, ##args)
|
2010-02-09 19:49:49 +08:00
|
|
|
#else
|
|
|
|
#define netif_dbg(priv, type, dev, format, args...) \
|
|
|
|
({ \
|
|
|
|
if (0) \
|
|
|
|
netif_printk(priv, type, KERN_DEBUG, dev, format, ##args); \
|
|
|
|
0; \
|
|
|
|
})
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#if defined(VERBOSE_DEBUG)
|
2010-07-02 15:08:44 +08:00
|
|
|
#define netif_vdbg netif_dbg
|
2010-02-09 19:49:49 +08:00
|
|
|
#else
|
|
|
|
#define netif_vdbg(priv, type, dev, format, args...) \
|
|
|
|
({ \
|
|
|
|
if (0) \
|
2010-05-18 14:56:32 +08:00
|
|
|
netif_printk(priv, type, KERN_DEBUG, dev, format, ##args); \
|
2010-02-09 19:49:49 +08:00
|
|
|
0; \
|
|
|
|
})
|
|
|
|
#endif
|
2010-02-09 19:49:47 +08:00
|
|
|
|
2013-02-19 03:20:33 +08:00
|
|
|
/*
|
|
|
|
* The list of packet types we will receive (as opposed to discard)
|
|
|
|
* and the routines to invoke.
|
|
|
|
*
|
|
|
|
* Why 16. Because with 16 the only overlap we get on a hash of the
|
|
|
|
* low nibble of the protocol value is RARP/SNAP/X.25.
|
|
|
|
*
|
|
|
|
* NOTE: That is no longer true with the addition of VLAN tags. Not
|
|
|
|
* sure which should go first, but I bet it won't make much
|
|
|
|
* difference if we are running VLANs. The good news is that
|
|
|
|
* this protocol won't be in the list unless compiled in, so
|
|
|
|
* the average user (w/out VLANs) will not be adversely affected.
|
|
|
|
* --BLG
|
|
|
|
*
|
|
|
|
* 0800 IP
|
|
|
|
* 8100 802.1Q VLAN
|
|
|
|
* 0001 802.3
|
|
|
|
* 0002 AX.25
|
|
|
|
* 0004 802.2
|
|
|
|
* 8035 RARP
|
|
|
|
* 0005 SNAP
|
|
|
|
* 0805 X.25
|
|
|
|
* 0806 ARP
|
|
|
|
* 8137 IPX
|
|
|
|
* 0009 Localtalk
|
|
|
|
* 86DD IPv6
|
|
|
|
*/
|
|
|
|
#define PTYPE_HASH_SIZE (16)
|
|
|
|
#define PTYPE_HASH_MASK (PTYPE_HASH_SIZE - 1)
|
|
|
|
|
2009-05-28 06:48:07 +08:00
|
|
|
#endif /* _LINUX_NETDEVICE_H */
|