linux_dsm_epyc7002/include/linux/skbuff.h

3163 lines
88 KiB
C
Raw Normal View History

/*
* Definitions for the 'struct sk_buff' memory handlers.
*
* Authors:
* Alan Cox, <gw4pts@gw4pts.ampr.org>
* Florian La Roche, <rzsfl@rz.uni-sb.de>
*
* 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.
*/
#ifndef _LINUX_SKBUFF_H
#define _LINUX_SKBUFF_H
#include <linux/kernel.h>
#include <linux/kmemcheck.h>
#include <linux/compiler.h>
#include <linux/time.h>
#include <linux/bug.h>
#include <linux/cache.h>
#include <linux/atomic.h>
#include <asm/types.h>
#include <linux/spinlock.h>
#include <linux/net.h>
#include <linux/textsearch.h>
#include <net/checksum.h>
#include <linux/rcupdate.h>
#include <linux/dmaengine.h>
#include <linux/hrtimer.h>
#include <linux/dma-mapping.h>
#include <linux/netdev_features.h>
#include <linux/sched.h>
#include <net/flow_keys.h>
/* A. Checksumming of received packets by device.
*
* CHECKSUM_NONE:
*
* Device failed to checksum this packet e.g. due to lack of capabilities.
* The packet contains full (though not verified) checksum in packet but
* not in skb->csum. Thus, skb->csum is undefined in this case.
*
* CHECKSUM_UNNECESSARY:
*
* The hardware you're dealing with doesn't calculate the full checksum
* (as in CHECKSUM_COMPLETE), but it does parse headers and verify checksums
* for specific protocols e.g. TCP/UDP/SCTP, then, for such packets it will
* set CHECKSUM_UNNECESSARY if their checksums are okay. skb->csum is still
* undefined in this case though. It is a bad option, but, unfortunately,
* nowadays most vendors do this. Apparently with the secret goal to sell
* you new devices, when you will add new protocol to your host, f.e. IPv6 8)
*
* CHECKSUM_COMPLETE:
*
* This is the most generic way. The device supplied checksum of the _whole_
* packet as seen by netif_rx() and fills out in skb->csum. Meaning, the
* hardware doesn't need to parse L3/L4 headers to implement this.
*
* Note: Even if device supports only some protocols, but is able to produce
* skb->csum, it MUST use CHECKSUM_COMPLETE, not CHECKSUM_UNNECESSARY.
*
* CHECKSUM_PARTIAL:
*
* This is identical to the case for output below. This may occur on a packet
* received directly from another Linux OS, e.g., a virtualized Linux kernel
* on the same host. The packet can be treated in the same way as
* CHECKSUM_UNNECESSARY, except that on output (i.e., forwarding) the
* checksum must be filled in by the OS or the hardware.
*
* B. Checksumming on output.
*
* CHECKSUM_NONE:
*
* The skb was already checksummed by the protocol, or a checksum is not
* required.
*
* CHECKSUM_PARTIAL:
*
* The device is required to checksum the packet as seen by hard_start_xmit()
* from skb->csum_start up to the end, and to record/write the checksum at
* offset skb->csum_start + skb->csum_offset.
*
* The device must show its capabilities in dev->features, set up at device
* setup time, e.g. netdev_features.h:
*
* NETIF_F_HW_CSUM - It's a clever device, it's able to checksum everything.
* NETIF_F_IP_CSUM - Device is dumb, it's able to checksum only TCP/UDP over
* IPv4. Sigh. Vendors like this way for an unknown reason.
* Though, see comment above about CHECKSUM_UNNECESSARY. 8)
* NETIF_F_IPV6_CSUM - About as dumb as the last one but does IPv6 instead.
* NETIF_F_... - Well, you get the picture.
*
* CHECKSUM_UNNECESSARY:
*
* Normally, the device will do per protocol specific checksumming. Protocol
* implementations that do not want the NIC to perform the checksum
* calculation should use this flag in their outgoing skbs.
*
* NETIF_F_FCOE_CRC - This indicates that the device can do FCoE FC CRC
* offload. Correspondingly, the FCoE protocol driver
* stack should use CHECKSUM_UNNECESSARY.
*
* Any questions? No questions, good. --ANK
*/
/* Don't change this without changing skb_csum_unnecessary! */
#define CHECKSUM_NONE 0
#define CHECKSUM_UNNECESSARY 1
#define CHECKSUM_COMPLETE 2
#define CHECKSUM_PARTIAL 3
#define SKB_DATA_ALIGN(X) (((X) + (SMP_CACHE_BYTES - 1)) & \
~(SMP_CACHE_BYTES - 1))
#define SKB_WITH_OVERHEAD(X) \
((X) - SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
#define SKB_MAX_ORDER(X, ORDER) \
SKB_WITH_OVERHEAD((PAGE_SIZE << (ORDER)) - (X))
#define SKB_MAX_HEAD(X) (SKB_MAX_ORDER((X), 0))
#define SKB_MAX_ALLOC (SKB_MAX_ORDER(0, 2))
/* return minimum truesize of one skb containing X bytes of data */
#define SKB_TRUESIZE(X) ((X) + \
SKB_DATA_ALIGN(sizeof(struct sk_buff)) + \
SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
struct net_device;
struct scatterlist;
struct pipe_inode_info;
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
struct nf_conntrack {
atomic_t use;
};
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
struct nf_bridge_info {
atomic_t use;
unsigned int mask;
struct net_device *physindev;
struct net_device *physoutdev;
unsigned long data[32 / sizeof(unsigned long)];
};
#endif
struct sk_buff_head {
/* These two members must be first. */
struct sk_buff *next;
struct sk_buff *prev;
__u32 qlen;
spinlock_t lock;
};
struct sk_buff;
/* To allow 64K frame to be packed as single skb without frag_list we
* require 64K/PAGE_SIZE pages plus 1 additional page to allow for
* buffers which do not start on a page boundary.
*
* Since GRO uses frags we allocate at least 16 regardless of page
* size.
*/
#if (65536/PAGE_SIZE + 1) < 16
#define MAX_SKB_FRAGS 16UL
#else
#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1)
#endif
typedef struct skb_frag_struct skb_frag_t;
struct skb_frag_struct {
struct {
struct page *p;
} page;
#if (BITS_PER_LONG > 32) || (PAGE_SIZE >= 65536)
__u32 page_offset;
__u32 size;
#else
__u16 page_offset;
__u16 size;
#endif
};
static inline unsigned int skb_frag_size(const skb_frag_t *frag)
{
return frag->size;
}
static inline void skb_frag_size_set(skb_frag_t *frag, unsigned int size)
{
frag->size = size;
}
static inline void skb_frag_size_add(skb_frag_t *frag, int delta)
{
frag->size += delta;
}
static inline void skb_frag_size_sub(skb_frag_t *frag, int delta)
{
frag->size -= delta;
}
#define HAVE_HW_TIME_STAMP
/**
* struct skb_shared_hwtstamps - hardware time stamps
* @hwtstamp: hardware time stamp transformed into duration
* since arbitrary point in time
* @syststamp: hwtstamp transformed to system time base
*
* Software time stamps generated by ktime_get_real() are stored in
* skb->tstamp. The relation between the different kinds of time
* stamps is as follows:
*
* syststamp and tstamp can be compared against each other in
* arbitrary combinations. The accuracy of a
* syststamp/tstamp/"syststamp from other device" comparison is
* limited by the accuracy of the transformation into system time
* base. This depends on the device driver and its underlying
* hardware.
*
* hwtstamps can only be compared against other hwtstamps from
* the same device.
*
* This structure is attached to packets as part of the
* &skb_shared_info. Use skb_hwtstamps() to get a pointer.
*/
struct skb_shared_hwtstamps {
ktime_t hwtstamp;
ktime_t syststamp;
};
/* Definitions for tx_flags in struct skb_shared_info */
enum {
/* generate hardware time stamp */
SKBTX_HW_TSTAMP = 1 << 0,
/* generate software time stamp */
SKBTX_SW_TSTAMP = 1 << 1,
/* device driver is going to provide hardware time stamp */
SKBTX_IN_PROGRESS = 1 << 2,
/* device driver supports TX zero-copy buffers */
SKBTX_DEV_ZEROCOPY = 1 << 3,
/* generate wifi status information (where possible) */
SKBTX_WIFI_STATUS = 1 << 4,
/* This indicates at least one fragment might be overwritten
* (as in vmsplice(), sendfile() ...)
* If we need to compute a TX checksum, we'll need to copy
* all frags to avoid possible bad checksum
*/
SKBTX_SHARED_FRAG = 1 << 5,
};
/*
* The callback notifies userspace to release buffers when skb DMA is done in
* lower device, the skb last reference should be 0 when calling this.
* The zerocopy_success argument is true if zero copy transmit occurred,
* false on data copy or out of memory error caused by data copy attempt.
* The ctx field is used to track device context.
* The desc field is used to track userspace buffer index.
*/
struct ubuf_info {
void (*callback)(struct ubuf_info *, bool zerocopy_success);
void *ctx;
unsigned long desc;
};
/* This data is invariant across clones and lives at
* the end of the header data, ie. at skb->end.
*/
struct skb_shared_info {
unsigned char nr_frags;
__u8 tx_flags;
unsigned short gso_size;
/* Warning: this field is not always filled in (UFO)! */
unsigned short gso_segs;
unsigned short gso_type;
struct sk_buff *frag_list;
struct skb_shared_hwtstamps hwtstamps;
__be32 ip6_frag_id;
/*
* Warning : all fields before dataref are cleared in __alloc_skb()
*/
atomic_t dataref;
/* Intermediate layers must ensure that destructor_arg
* remains valid until skb destructor */
void * destructor_arg;
/* must be last field, see pskb_expand_head() */
skb_frag_t frags[MAX_SKB_FRAGS];
};
/* We divide dataref into two halves. The higher 16 bits hold references
* to the payload part of skb->data. The lower 16 bits hold references to
[SKBUFF]: Keep track of writable header len of headerless clones Currently NAT (and others) that want to modify cloned skbs copy them, even if in the vast majority of cases its not necessary because the skb is a clone made by TCP and the portion NAT wants to modify is actually writable because TCP release the header reference before cloning. The problem is that there is no clean way for NAT to find out how long the writable header area is, so this patch introduces skb->hdr_len to hold this length. When a headerless skb is cloned skb->hdr_len is set to the current headroom, for regular clones it is copied from the original. A new function skb_clone_writable(skb, len) returns whether the skb is writable up to len bytes from skb->data. To avoid enlarging the skb the mac_len field is reduced to 16 bit and the new hdr_len field is put in the remaining 16 bit. I've done a few rough benchmarks of NAT (not with this exact patch, but a very similar one). As expected it saves huge amounts of system time in case of sendfile, bringing it down to basically the same amount as without NAT, with sendmsg it only helps on loopback, probably because of the large MTU. Transmit a 1GB file using sendfile/sendmsg over eth0/lo with and without NAT: - sendfile eth0, no NAT: sys 0m0.388s - sendfile eth0, NAT: sys 0m1.835s - sendfile eth0: NAT + path: sys 0m0.370s (~ -80%) - sendfile lo, no NAT: sys 0m0.258s - sendfile lo, NAT: sys 0m2.609s - sendfile lo, NAT + patch: sys 0m0.260s (~ -90%) - sendmsg eth0, no NAT: sys 0m2.508s - sendmsg eth0, NAT: sys 0m2.539s - sendmsg eth0, NAT + patch: sys 0m2.445s (no change) - sendmsg lo, no NAT: sys 0m2.151s - sendmsg lo, NAT: sys 0m3.557s - sendmsg lo, NAT + patch: sys 0m2.159s (~ -40%) I expect other users can see a similar performance improvement, packet mangling iptables targets, ipip and ip_gre come to mind .. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-06-25 18:35:20 +07:00
* the entire skb->data. A clone of a headerless skb holds the length of
* the header in skb->hdr_len.
*
* All users must obey the rule that the skb->data reference count must be
* greater than or equal to the payload reference count.
*
* Holding a reference to the payload part means that the user does not
* care about modifications to the header part of skb->data.
*/
#define SKB_DATAREF_SHIFT 16
#define SKB_DATAREF_MASK ((1 << SKB_DATAREF_SHIFT) - 1)
enum {
SKB_FCLONE_UNAVAILABLE,
SKB_FCLONE_ORIG,
SKB_FCLONE_CLONE,
};
enum {
SKB_GSO_TCPV4 = 1 << 0,
[IPV6]: Added GSO support for TCPv6 This patch adds GSO support for IPv6 and TCPv6. This is based on a patch by Ananda Raju <Ananda.Raju@neterion.com>. His original description is: This patch enables TSO over IPv6. Currently Linux network stacks restricts TSO over IPv6 by clearing of the NETIF_F_TSO bit from "dev->features". This patch will remove this restriction. This patch will introduce a new flag NETIF_F_TSO6 which will be used to check whether device supports TSO over IPv6. If device support TSO over IPv6 then we don't clear of NETIF_F_TSO and which will make the TCP layer to create TSO packets. Any device supporting TSO over IPv6 will set NETIF_F_TSO6 flag in "dev->features" along with NETIF_F_TSO. In case when user disables TSO using ethtool, NETIF_F_TSO will get cleared from "dev->features". So even if we have NETIF_F_TSO6 we don't get TSO packets created by TCP layer. SKB_GSO_TCPV4 renamed to SKB_GSO_TCP to make it generic GSO packet. SKB_GSO_UDPV4 renamed to SKB_GSO_UDP as UFO is not a IPv4 feature. UFO is supported over IPv6 also The following table shows there is significant improvement in throughput with normal frames and CPU usage for both normal and jumbo. -------------------------------------------------- | | 1500 | 9600 | | ------------------|-------------------| | | thru CPU | thru CPU | -------------------------------------------------- | TSO OFF | 2.00 5.5% id | 5.66 20.0% id | -------------------------------------------------- | TSO ON | 2.63 78.0 id | 5.67 39.0% id | -------------------------------------------------- Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-07-01 03:37:03 +07:00
SKB_GSO_UDP = 1 << 1,
/* This indicates the skb is from an untrusted source. */
SKB_GSO_DODGY = 1 << 2,
/* This indicates the tcp segment has CWR set. */
[IPV6]: Added GSO support for TCPv6 This patch adds GSO support for IPv6 and TCPv6. This is based on a patch by Ananda Raju <Ananda.Raju@neterion.com>. His original description is: This patch enables TSO over IPv6. Currently Linux network stacks restricts TSO over IPv6 by clearing of the NETIF_F_TSO bit from "dev->features". This patch will remove this restriction. This patch will introduce a new flag NETIF_F_TSO6 which will be used to check whether device supports TSO over IPv6. If device support TSO over IPv6 then we don't clear of NETIF_F_TSO and which will make the TCP layer to create TSO packets. Any device supporting TSO over IPv6 will set NETIF_F_TSO6 flag in "dev->features" along with NETIF_F_TSO. In case when user disables TSO using ethtool, NETIF_F_TSO will get cleared from "dev->features". So even if we have NETIF_F_TSO6 we don't get TSO packets created by TCP layer. SKB_GSO_TCPV4 renamed to SKB_GSO_TCP to make it generic GSO packet. SKB_GSO_UDPV4 renamed to SKB_GSO_UDP as UFO is not a IPv4 feature. UFO is supported over IPv6 also The following table shows there is significant improvement in throughput with normal frames and CPU usage for both normal and jumbo. -------------------------------------------------- | | 1500 | 9600 | | ------------------|-------------------| | | thru CPU | thru CPU | -------------------------------------------------- | TSO OFF | 2.00 5.5% id | 5.66 20.0% id | -------------------------------------------------- | TSO ON | 2.63 78.0 id | 5.67 39.0% id | -------------------------------------------------- Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-07-01 03:37:03 +07:00
SKB_GSO_TCP_ECN = 1 << 3,
SKB_GSO_TCPV6 = 1 << 4,
SKB_GSO_FCOE = 1 << 5,
SKB_GSO_GRE = 1 << 6,
SKB_GSO_GRE_CSUM = 1 << 7,
MPLS: Add limited GSO support In the case where a non-MPLS packet is received and an MPLS stack is added it may well be the case that the original skb is GSO but the NIC used for transmit does not support GSO of MPLS packets. The aim of this code is to provide GSO in software for MPLS packets whose skbs are GSO. SKB Usage: When an implementation adds an MPLS stack to a non-MPLS packet it should do the following to skb metadata: * Set skb->inner_protocol to the old non-MPLS ethertype of the packet. skb->inner_protocol is added by this patch. * Set skb->protocol to the new MPLS ethertype of the packet. * Set skb->network_header to correspond to the end of the L3 header, including the MPLS label stack. I have posted a patch, "[PATCH v3.29] datapath: Add basic MPLS support to kernel" which adds MPLS support to the kernel datapath of Open vSwtich. That patch sets the above requirements in datapath/actions.c:push_mpls() and was used to exercise this code. The datapath patch is against the Open vSwtich tree but it is intended that it be added to the Open vSwtich code present in the mainline Linux kernel at some point. Features: I believe that the approach that I have taken is at least partially consistent with the handling of other protocols. Jesse, I understand that you have some ideas here. I am more than happy to change my implementation. This patch adds dev->mpls_features which may be used by devices to advertise features supported for MPLS packets. A new NETIF_F_MPLS_GSO feature is added for devices which support hardware MPLS GSO offload. Currently no devices support this and MPLS GSO always falls back to software. Alternate Implementation: One possible alternate implementation is to teach netif_skb_features() and skb_network_protocol() about MPLS, in a similar way to their understanding of VLANs. I believe this would avoid the need for net/mpls/mpls_gso.c and in particular the calls to __skb_push() and __skb_push() in mpls_gso_segment(). I have decided on the implementation in this patch as it should not introduce any overhead in the case where mpls_gso is not compiled into the kernel or inserted as a module. MPLS GSO suggested by Jesse Gross. Based in part on "v4 GRE: Add TCP segmentation offload for GRE" by Pravin B Shelar. Cc: Jesse Gross <jesse@nicira.com> Cc: Pravin B Shelar <pshelar@nicira.com> Signed-off-by: Simon Horman <horms@verge.net.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-05-24 04:02:52 +07:00
SKB_GSO_IPIP = 1 << 8,
SKB_GSO_SIT = 1 << 9,
SKB_GSO_UDP_TUNNEL = 1 << 10,
SKB_GSO_UDP_TUNNEL_CSUM = 1 << 11,
SKB_GSO_MPLS = 1 << 12,
};
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
#if BITS_PER_LONG > 32
#define NET_SKBUFF_DATA_USES_OFFSET 1
#endif
#ifdef NET_SKBUFF_DATA_USES_OFFSET
typedef unsigned int sk_buff_data_t;
#else
typedef unsigned char *sk_buff_data_t;
#endif
/**
* struct skb_mstamp - multi resolution time stamps
* @stamp_us: timestamp in us resolution
* @stamp_jiffies: timestamp in jiffies
*/
struct skb_mstamp {
union {
u64 v64;
struct {
u32 stamp_us;
u32 stamp_jiffies;
};
};
};
/**
* skb_mstamp_get - get current timestamp
* @cl: place to store timestamps
*/
static inline void skb_mstamp_get(struct skb_mstamp *cl)
{
u64 val = local_clock();
do_div(val, NSEC_PER_USEC);
cl->stamp_us = (u32)val;
cl->stamp_jiffies = (u32)jiffies;
}
/**
* skb_mstamp_delta - compute the difference in usec between two skb_mstamp
* @t1: pointer to newest sample
* @t0: pointer to oldest sample
*/
static inline u32 skb_mstamp_us_delta(const struct skb_mstamp *t1,
const struct skb_mstamp *t0)
{
s32 delta_us = t1->stamp_us - t0->stamp_us;
u32 delta_jiffies = t1->stamp_jiffies - t0->stamp_jiffies;
/* If delta_us is negative, this might be because interval is too big,
* or local_clock() drift is too big : fallback using jiffies.
*/
if (delta_us <= 0 ||
delta_jiffies >= (INT_MAX / (USEC_PER_SEC / HZ)))
delta_us = jiffies_to_usecs(delta_jiffies);
return delta_us;
}
/**
* struct sk_buff - socket buffer
* @next: Next buffer in list
* @prev: Previous buffer in list
* @tstamp: Time we arrived/left
* @sk: Socket we are owned by
* @dev: Device we arrived on/are leaving by
* @cb: Control buffer. Free for use by every layer. Put private vars here
* @_skb_refdst: destination entry (with norefcount bit)
* @sp: the security path, used for xfrm
* @len: Length of actual data
* @data_len: Data length
* @mac_len: Length of link layer header
[SKBUFF]: Keep track of writable header len of headerless clones Currently NAT (and others) that want to modify cloned skbs copy them, even if in the vast majority of cases its not necessary because the skb is a clone made by TCP and the portion NAT wants to modify is actually writable because TCP release the header reference before cloning. The problem is that there is no clean way for NAT to find out how long the writable header area is, so this patch introduces skb->hdr_len to hold this length. When a headerless skb is cloned skb->hdr_len is set to the current headroom, for regular clones it is copied from the original. A new function skb_clone_writable(skb, len) returns whether the skb is writable up to len bytes from skb->data. To avoid enlarging the skb the mac_len field is reduced to 16 bit and the new hdr_len field is put in the remaining 16 bit. I've done a few rough benchmarks of NAT (not with this exact patch, but a very similar one). As expected it saves huge amounts of system time in case of sendfile, bringing it down to basically the same amount as without NAT, with sendmsg it only helps on loopback, probably because of the large MTU. Transmit a 1GB file using sendfile/sendmsg over eth0/lo with and without NAT: - sendfile eth0, no NAT: sys 0m0.388s - sendfile eth0, NAT: sys 0m1.835s - sendfile eth0: NAT + path: sys 0m0.370s (~ -80%) - sendfile lo, no NAT: sys 0m0.258s - sendfile lo, NAT: sys 0m2.609s - sendfile lo, NAT + patch: sys 0m0.260s (~ -90%) - sendmsg eth0, no NAT: sys 0m2.508s - sendmsg eth0, NAT: sys 0m2.539s - sendmsg eth0, NAT + patch: sys 0m2.445s (no change) - sendmsg lo, no NAT: sys 0m2.151s - sendmsg lo, NAT: sys 0m3.557s - sendmsg lo, NAT + patch: sys 0m2.159s (~ -40%) I expect other users can see a similar performance improvement, packet mangling iptables targets, ipip and ip_gre come to mind .. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-06-25 18:35:20 +07:00
* @hdr_len: writable header length of cloned skb
* @csum: Checksum (must include start/offset pair)
* @csum_start: Offset from skb->head where checksumming should start
* @csum_offset: Offset from csum_start where checksum should be stored
* @priority: Packet queueing priority
* @ignore_df: allow local fragmentation
* @cloned: Head may be cloned (check refcnt to be sure)
* @ip_summed: Driver fed us an IP checksum
* @nohdr: Payload reference only, must not modify header
* @nfctinfo: Relationship of this skb to the connection
* @pkt_type: Packet class
* @fclone: skbuff clone status
* @ipvs_property: skbuff is owned by ipvs
* @peeked: this packet has been seen already, so stats have been
* done for it, don't do them again
* @nf_trace: netfilter packet trace flag
* @protocol: Packet protocol from driver
* @destructor: Destruct function
* @nfct: Associated connection, if any
* @nf_bridge: Saved data about a bridged frame - see br_netfilter.c
* @skb_iif: ifindex of device we arrived on
* @tc_index: Traffic control index
* @tc_verd: traffic control verdict
* @hash: the packet hash
* @queue_mapping: Queue mapping for multiqueue devices
* @ndisc_nodetype: router type (from link layer)
* @ooo_okay: allow the mapping of a socket to a queue to be changed
* @l4_hash: indicate hash is a canonical 4-tuple hash over transport
* ports.
* @wifi_acked_valid: wifi_acked was set
* @wifi_acked: whether frame was acked on wifi or not
* @no_fcs: Request NIC to treat last 4 bytes as Ethernet FCS
* @dma_cookie: a cookie to one of several possible DMA operations
* done by skb DMA functions
* @napi_id: id of the NAPI struct this skb came from
* @secmark: security marking
* @mark: Generic packet mark
* @dropcount: total number of sk_receive_queue overflows
* @vlan_proto: vlan encapsulation protocol
* @vlan_tci: vlan tag control information
MPLS: Add limited GSO support In the case where a non-MPLS packet is received and an MPLS stack is added it may well be the case that the original skb is GSO but the NIC used for transmit does not support GSO of MPLS packets. The aim of this code is to provide GSO in software for MPLS packets whose skbs are GSO. SKB Usage: When an implementation adds an MPLS stack to a non-MPLS packet it should do the following to skb metadata: * Set skb->inner_protocol to the old non-MPLS ethertype of the packet. skb->inner_protocol is added by this patch. * Set skb->protocol to the new MPLS ethertype of the packet. * Set skb->network_header to correspond to the end of the L3 header, including the MPLS label stack. I have posted a patch, "[PATCH v3.29] datapath: Add basic MPLS support to kernel" which adds MPLS support to the kernel datapath of Open vSwtich. That patch sets the above requirements in datapath/actions.c:push_mpls() and was used to exercise this code. The datapath patch is against the Open vSwtich tree but it is intended that it be added to the Open vSwtich code present in the mainline Linux kernel at some point. Features: I believe that the approach that I have taken is at least partially consistent with the handling of other protocols. Jesse, I understand that you have some ideas here. I am more than happy to change my implementation. This patch adds dev->mpls_features which may be used by devices to advertise features supported for MPLS packets. A new NETIF_F_MPLS_GSO feature is added for devices which support hardware MPLS GSO offload. Currently no devices support this and MPLS GSO always falls back to software. Alternate Implementation: One possible alternate implementation is to teach netif_skb_features() and skb_network_protocol() about MPLS, in a similar way to their understanding of VLANs. I believe this would avoid the need for net/mpls/mpls_gso.c and in particular the calls to __skb_push() and __skb_push() in mpls_gso_segment(). I have decided on the implementation in this patch as it should not introduce any overhead in the case where mpls_gso is not compiled into the kernel or inserted as a module. MPLS GSO suggested by Jesse Gross. Based in part on "v4 GRE: Add TCP segmentation offload for GRE" by Pravin B Shelar. Cc: Jesse Gross <jesse@nicira.com> Cc: Pravin B Shelar <pshelar@nicira.com> Signed-off-by: Simon Horman <horms@verge.net.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-05-24 04:02:52 +07:00
* @inner_protocol: Protocol (encapsulation)
* @inner_transport_header: Inner transport layer header (encapsulation)
* @inner_network_header: Network layer header (encapsulation)
* @inner_mac_header: Link layer header (encapsulation)
* @transport_header: Transport layer header
* @network_header: Network layer header
* @mac_header: Link layer header
* @tail: Tail pointer
* @end: End pointer
* @head: Head of buffer
* @data: Data head pointer
* @truesize: Buffer size
* @users: User count - see {datagram,tcp}.c
*/
struct sk_buff {
/* These two members must be first. */
struct sk_buff *next;
struct sk_buff *prev;
union {
ktime_t tstamp;
struct skb_mstamp skb_mstamp;
};
struct sock *sk;
struct net_device *dev;
/*
* This is the control buffer. It is free to use for every
* layer. Please put your private variables there. If you
* want to keep them across layers you have to do a skb_clone()
* first. This is owned by whoever has the skb queued ATM.
*/
char cb[48] __aligned(8);
unsigned long _skb_refdst;
#ifdef CONFIG_XFRM
struct sec_path *sp;
#endif
unsigned int len,
[SKBUFF]: Keep track of writable header len of headerless clones Currently NAT (and others) that want to modify cloned skbs copy them, even if in the vast majority of cases its not necessary because the skb is a clone made by TCP and the portion NAT wants to modify is actually writable because TCP release the header reference before cloning. The problem is that there is no clean way for NAT to find out how long the writable header area is, so this patch introduces skb->hdr_len to hold this length. When a headerless skb is cloned skb->hdr_len is set to the current headroom, for regular clones it is copied from the original. A new function skb_clone_writable(skb, len) returns whether the skb is writable up to len bytes from skb->data. To avoid enlarging the skb the mac_len field is reduced to 16 bit and the new hdr_len field is put in the remaining 16 bit. I've done a few rough benchmarks of NAT (not with this exact patch, but a very similar one). As expected it saves huge amounts of system time in case of sendfile, bringing it down to basically the same amount as without NAT, with sendmsg it only helps on loopback, probably because of the large MTU. Transmit a 1GB file using sendfile/sendmsg over eth0/lo with and without NAT: - sendfile eth0, no NAT: sys 0m0.388s - sendfile eth0, NAT: sys 0m1.835s - sendfile eth0: NAT + path: sys 0m0.370s (~ -80%) - sendfile lo, no NAT: sys 0m0.258s - sendfile lo, NAT: sys 0m2.609s - sendfile lo, NAT + patch: sys 0m0.260s (~ -90%) - sendmsg eth0, no NAT: sys 0m2.508s - sendmsg eth0, NAT: sys 0m2.539s - sendmsg eth0, NAT + patch: sys 0m2.445s (no change) - sendmsg lo, no NAT: sys 0m2.151s - sendmsg lo, NAT: sys 0m3.557s - sendmsg lo, NAT + patch: sys 0m2.159s (~ -40%) I expect other users can see a similar performance improvement, packet mangling iptables targets, ipip and ip_gre come to mind .. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-06-25 18:35:20 +07:00
data_len;
__u16 mac_len,
hdr_len;
union {
__wsum csum;
struct {
__u16 csum_start;
__u16 csum_offset;
};
};
__u32 priority;
kmemcheck_bitfield_begin(flags1);
__u8 ignore_df:1,
cloned:1,
ip_summed:2,
nohdr:1,
nfctinfo:3;
__u8 pkt_type:3,
fclone:2,
ipvs_property:1,
peeked:1,
nf_trace:1;
kmemcheck_bitfield_end(flags1);
__be16 protocol;
void (*destructor)(struct sk_buff *skb);
[NETFILTER]: Add nf_conntrack subsystem. The existing connection tracking subsystem in netfilter can only handle ipv4. There were basically two choices present to add connection tracking support for ipv6. We could either duplicate all of the ipv4 connection tracking code into an ipv6 counterpart, or (the choice taken by these patches) we could design a generic layer that could handle both ipv4 and ipv6 and thus requiring only one sub-protocol (TCP, UDP, etc.) connection tracking helper module to be written. In fact nf_conntrack is capable of working with any layer 3 protocol. The existing ipv4 specific conntrack code could also not deal with the pecularities of doing connection tracking on ipv6, which is also cured here. For example, these issues include: 1) ICMPv6 handling, which is used for neighbour discovery in ipv6 thus some messages such as these should not participate in connection tracking since effectively they are like ARP messages 2) fragmentation must be handled differently in ipv6, because the simplistic "defrag, connection track and NAT, refrag" (which the existing ipv4 connection tracking does) approach simply isn't feasible in ipv6 3) ipv6 extension header parsing must occur at the correct spots before and after connection tracking decisions, and there were no provisions for this in the existing connection tracking design 4) ipv6 has no need for stateful NAT The ipv4 specific conntrack layer is kept around, until all of the ipv4 specific conntrack helpers are ported over to nf_conntrack and it is feature complete. Once that occurs, the old conntrack stuff will get placed into the feature-removal-schedule and we will fully kill it off 6 months later. Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp> Signed-off-by: Harald Welte <laforge@netfilter.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-11-10 07:38:16 +07:00
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
struct nf_conntrack *nfct;
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
struct nf_bridge_info *nf_bridge;
#endif
int skb_iif;
__u32 hash;
__be16 vlan_proto;
__u16 vlan_tci;
#ifdef CONFIG_NET_SCHED
__u16 tc_index; /* traffic control index */
#ifdef CONFIG_NET_CLS_ACT
__u16 tc_verd; /* traffic control verdict */
#endif
#endif
net: filter: Just In Time compiler for x86-64 In order to speedup packet filtering, here is an implementation of a JIT compiler for x86_64 It is disabled by default, and must be enabled by the admin. echo 1 >/proc/sys/net/core/bpf_jit_enable It uses module_alloc() and module_free() to get memory in the 2GB text kernel range since we call helpers functions from the generated code. EAX : BPF A accumulator EBX : BPF X accumulator RDI : pointer to skb (first argument given to JIT function) RBP : frame pointer (even if CONFIG_FRAME_POINTER=n) r9d : skb->len - skb->data_len (headlen) r8 : skb->data To get a trace of generated code, use : echo 2 >/proc/sys/net/core/bpf_jit_enable Example of generated code : # tcpdump -p -n -s 0 -i eth1 host 192.168.20.0/24 flen=18 proglen=147 pass=3 image=ffffffffa00b5000 JIT code: ffffffffa00b5000: 55 48 89 e5 48 83 ec 60 48 89 5d f8 44 8b 4f 60 JIT code: ffffffffa00b5010: 44 2b 4f 64 4c 8b 87 b8 00 00 00 be 0c 00 00 00 JIT code: ffffffffa00b5020: e8 24 7b f7 e0 3d 00 08 00 00 75 28 be 1a 00 00 JIT code: ffffffffa00b5030: 00 e8 fe 7a f7 e0 24 00 3d 00 14 a8 c0 74 49 be JIT code: ffffffffa00b5040: 1e 00 00 00 e8 eb 7a f7 e0 24 00 3d 00 14 a8 c0 JIT code: ffffffffa00b5050: 74 36 eb 3b 3d 06 08 00 00 74 07 3d 35 80 00 00 JIT code: ffffffffa00b5060: 75 2d be 1c 00 00 00 e8 c8 7a f7 e0 24 00 3d 00 JIT code: ffffffffa00b5070: 14 a8 c0 74 13 be 26 00 00 00 e8 b5 7a f7 e0 24 JIT code: ffffffffa00b5080: 00 3d 00 14 a8 c0 75 07 b8 ff ff 00 00 eb 02 31 JIT code: ffffffffa00b5090: c0 c9 c3 BPF program is 144 bytes long, so native program is almost same size ;) (000) ldh [12] (001) jeq #0x800 jt 2 jf 8 (002) ld [26] (003) and #0xffffff00 (004) jeq #0xc0a81400 jt 16 jf 5 (005) ld [30] (006) and #0xffffff00 (007) jeq #0xc0a81400 jt 16 jf 17 (008) jeq #0x806 jt 10 jf 9 (009) jeq #0x8035 jt 10 jf 17 (010) ld [28] (011) and #0xffffff00 (012) jeq #0xc0a81400 jt 16 jf 13 (013) ld [38] (014) and #0xffffff00 (015) jeq #0xc0a81400 jt 16 jf 17 (016) ret #65535 (017) ret #0 Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@infradead.org> Cc: Ben Hutchings <bhutchings@solarflare.com> Cc: Hagen Paul Pfeifer <hagen@jauu.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-20 16:27:32 +07:00
__u16 queue_mapping;
kmemcheck_bitfield_begin(flags2);
#ifdef CONFIG_IPV6_NDISC_NODETYPE
__u8 ndisc_nodetype:2;
#endif
__u8 pfmemalloc:1;
__u8 ooo_okay:1;
__u8 l4_hash:1;
__u8 wifi_acked_valid:1;
__u8 wifi_acked:1;
__u8 no_fcs:1;
net: allow skb->head to be a page fragment skb->head is currently allocated from kmalloc(). This is convenient but has the drawback the data cannot be converted to a page fragment if needed. We have three spots were it hurts : 1) GRO aggregation When a linear skb must be appended to another skb, GRO uses the frag_list fallback, very inefficient since we keep all struct sk_buff around. So drivers enabling GRO but delivering linear skbs to network stack aren't enabling full GRO power. 2) splice(socket -> pipe). We must copy the linear part to a page fragment. This kind of defeats splice() purpose (zero copy claim) 3) TCP coalescing. Recently introduced, this permits to group several contiguous segments into a single skb. This shortens queue lengths and save kernel memory, and greatly reduce probabilities of TCP collapses. This coalescing doesnt work on linear skbs (or we would need to copy data, this would be too slow) Given all these issues, the following patch introduces the possibility of having skb->head be a fragment in itself. We use a new skb flag, skb->head_frag to carry this information. build_skb() is changed to accept a frag_size argument. Drivers willing to provide a page fragment instead of kmalloc() data will set a non zero value, set to the fragment size. Then, on situations we need to convert the skb head to a frag in itself, we can check if skb->head_frag is set and avoid the copies or various fallbacks we have. This means drivers currently using frags could be updated to avoid the current skb->head allocation and reduce their memory footprint (aka skb truesize). (thats 512 or 1024 bytes saved per skb). This also makes bpf/netfilter faster since the 'first frag' will be part of skb linear part, no need to copy data. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Cc: Herbert Xu <herbert@gondor.apana.org.au> Cc: Maciej Żenczykowski <maze@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Tom Herbert <therbert@google.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Ben Hutchings <bhutchings@solarflare.com> Cc: Matt Carlson <mcarlson@broadcom.com> Cc: Michael Chan <mchan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-27 07:33:38 +07:00
__u8 head_frag:1;
/* Encapsulation protocol and NIC drivers should use
* this flag to indicate to each other if the skb contains
* encapsulated packet or not and maybe use the inner packet
* headers if needed
*/
__u8 encapsulation:1;
__u8 encap_hdr_csum:1;
__u8 csum_valid:1;
__u8 csum_complete_sw:1;
/* 3/5 bit hole (depending on ndisc_nodetype presence) */
kmemcheck_bitfield_end(flags2);
#if defined CONFIG_NET_DMA || defined CONFIG_NET_RX_BUSY_POLL
union {
unsigned int napi_id;
dma_cookie_t dma_cookie;
};
#endif
#ifdef CONFIG_NETWORK_SECMARK
__u32 secmark;
#endif
net: Generalize socket rx gap / receive queue overflow cmsg Create a new socket level option to report number of queue overflows Recently I augmented the AF_PACKET protocol to report the number of frames lost on the socket receive queue between any two enqueued frames. This value was exported via a SOL_PACKET level cmsg. AFter I completed that work it was requested that this feature be generalized so that any datagram oriented socket could make use of this option. As such I've created this patch, It creates a new SOL_SOCKET level option called SO_RXQ_OVFL, which when enabled exports a SOL_SOCKET level cmsg that reports the nubmer of times the sk_receive_queue overflowed between any two given frames. It also augments the AF_PACKET protocol to take advantage of this new feature (as it previously did not touch sk->sk_drops, which this patch uses to record the overflow count). Tested successfully by me. Notes: 1) Unlike my previous patch, this patch simply records the sk_drops value, which is not a number of drops between packets, but rather a total number of drops. Deltas must be computed in user space. 2) While this patch currently works with datagram oriented protocols, it will also be accepted by non-datagram oriented protocols. I'm not sure if thats agreeable to everyone, but my argument in favor of doing so is that, for those protocols which aren't applicable to this option, sk_drops will always be zero, and reporting no drops on a receive queue that isn't used for those non-participating protocols seems reasonable to me. This also saves us having to code in a per-protocol opt in mechanism. 3) This applies cleanly to net-next assuming that commit 977750076d98c7ff6cbda51858bb5a5894a9d9ab (my af packet cmsg patch) is reverted Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 03:26:31 +07:00
union {
__u32 mark;
__u32 dropcount;
__u32 reserved_tailroom;
net: Generalize socket rx gap / receive queue overflow cmsg Create a new socket level option to report number of queue overflows Recently I augmented the AF_PACKET protocol to report the number of frames lost on the socket receive queue between any two enqueued frames. This value was exported via a SOL_PACKET level cmsg. AFter I completed that work it was requested that this feature be generalized so that any datagram oriented socket could make use of this option. As such I've created this patch, It creates a new SOL_SOCKET level option called SO_RXQ_OVFL, which when enabled exports a SOL_SOCKET level cmsg that reports the nubmer of times the sk_receive_queue overflowed between any two given frames. It also augments the AF_PACKET protocol to take advantage of this new feature (as it previously did not touch sk->sk_drops, which this patch uses to record the overflow count). Tested successfully by me. Notes: 1) Unlike my previous patch, this patch simply records the sk_drops value, which is not a number of drops between packets, but rather a total number of drops. Deltas must be computed in user space. 2) While this patch currently works with datagram oriented protocols, it will also be accepted by non-datagram oriented protocols. I'm not sure if thats agreeable to everyone, but my argument in favor of doing so is that, for those protocols which aren't applicable to this option, sk_drops will always be zero, and reporting no drops on a receive queue that isn't used for those non-participating protocols seems reasonable to me. This also saves us having to code in a per-protocol opt in mechanism. 3) This applies cleanly to net-next assuming that commit 977750076d98c7ff6cbda51858bb5a5894a9d9ab (my af packet cmsg patch) is reverted Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 03:26:31 +07:00
};
MPLS: Add limited GSO support In the case where a non-MPLS packet is received and an MPLS stack is added it may well be the case that the original skb is GSO but the NIC used for transmit does not support GSO of MPLS packets. The aim of this code is to provide GSO in software for MPLS packets whose skbs are GSO. SKB Usage: When an implementation adds an MPLS stack to a non-MPLS packet it should do the following to skb metadata: * Set skb->inner_protocol to the old non-MPLS ethertype of the packet. skb->inner_protocol is added by this patch. * Set skb->protocol to the new MPLS ethertype of the packet. * Set skb->network_header to correspond to the end of the L3 header, including the MPLS label stack. I have posted a patch, "[PATCH v3.29] datapath: Add basic MPLS support to kernel" which adds MPLS support to the kernel datapath of Open vSwtich. That patch sets the above requirements in datapath/actions.c:push_mpls() and was used to exercise this code. The datapath patch is against the Open vSwtich tree but it is intended that it be added to the Open vSwtich code present in the mainline Linux kernel at some point. Features: I believe that the approach that I have taken is at least partially consistent with the handling of other protocols. Jesse, I understand that you have some ideas here. I am more than happy to change my implementation. This patch adds dev->mpls_features which may be used by devices to advertise features supported for MPLS packets. A new NETIF_F_MPLS_GSO feature is added for devices which support hardware MPLS GSO offload. Currently no devices support this and MPLS GSO always falls back to software. Alternate Implementation: One possible alternate implementation is to teach netif_skb_features() and skb_network_protocol() about MPLS, in a similar way to their understanding of VLANs. I believe this would avoid the need for net/mpls/mpls_gso.c and in particular the calls to __skb_push() and __skb_push() in mpls_gso_segment(). I have decided on the implementation in this patch as it should not introduce any overhead in the case where mpls_gso is not compiled into the kernel or inserted as a module. MPLS GSO suggested by Jesse Gross. Based in part on "v4 GRE: Add TCP segmentation offload for GRE" by Pravin B Shelar. Cc: Jesse Gross <jesse@nicira.com> Cc: Pravin B Shelar <pshelar@nicira.com> Signed-off-by: Simon Horman <horms@verge.net.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-05-24 04:02:52 +07:00
__be16 inner_protocol;
__u16 inner_transport_header;
__u16 inner_network_header;
__u16 inner_mac_header;
__u16 transport_header;
__u16 network_header;
__u16 mac_header;
/* These elements must be at the end, see alloc_skb() for details. */
sk_buff_data_t tail;
sk_buff_data_t end;
unsigned char *head,
*data;
unsigned int truesize;
atomic_t users;
};
#ifdef __KERNEL__
/*
* Handling routines are only of interest to the kernel
*/
#include <linux/slab.h>
#define SKB_ALLOC_FCLONE 0x01
#define SKB_ALLOC_RX 0x02
/* Returns true if the skb was allocated from PFMEMALLOC reserves */
static inline bool skb_pfmemalloc(const struct sk_buff *skb)
{
return unlikely(skb->pfmemalloc);
}
/*
* skb might have a dst pointer attached, refcounted or not.
* _skb_refdst low order bit is set if refcount was _not_ taken
*/
#define SKB_DST_NOREF 1UL
#define SKB_DST_PTRMASK ~(SKB_DST_NOREF)
/**
* skb_dst - returns skb dst_entry
* @skb: buffer
*
* Returns skb dst_entry, regardless of reference taken or not.
*/
static inline struct dst_entry *skb_dst(const struct sk_buff *skb)
{
/* If refdst was not refcounted, check we still are in a
* rcu_read_lock section
*/
WARN_ON((skb->_skb_refdst & SKB_DST_NOREF) &&
!rcu_read_lock_held() &&
!rcu_read_lock_bh_held());
return (struct dst_entry *)(skb->_skb_refdst & SKB_DST_PTRMASK);
}
/**
* skb_dst_set - sets skb dst
* @skb: buffer
* @dst: dst entry
*
* Sets skb dst, assuming a reference was taken on dst and should
* be released by skb_dst_drop()
*/
static inline void skb_dst_set(struct sk_buff *skb, struct dst_entry *dst)
{
skb->_skb_refdst = (unsigned long)dst;
}
void __skb_dst_set_noref(struct sk_buff *skb, struct dst_entry *dst,
bool force);
/**
* skb_dst_set_noref - sets skb dst, hopefully, without taking reference
* @skb: buffer
* @dst: dst entry
*
* Sets skb dst, assuming a reference was not taken on dst.
* If dst entry is cached, we do not take reference and dst_release
* will be avoided by refdst_drop. If dst entry is not cached, we take
* reference, so that last dst_release can destroy the dst immediately.
*/
static inline void skb_dst_set_noref(struct sk_buff *skb, struct dst_entry *dst)
{
__skb_dst_set_noref(skb, dst, false);
}
/**
* skb_dst_set_noref_force - sets skb dst, without taking reference
* @skb: buffer
* @dst: dst entry
*
* Sets skb dst, assuming a reference was not taken on dst.
* No reference is taken and no dst_release will be called. While for
* cached dsts deferred reclaim is a basic feature, for entries that are
* not cached it is caller's job to guarantee that last dst_release for
* provided dst happens when nobody uses it, eg. after a RCU grace period.
*/
static inline void skb_dst_set_noref_force(struct sk_buff *skb,
struct dst_entry *dst)
{
__skb_dst_set_noref(skb, dst, true);
}
/**
* skb_dst_is_noref - Test if skb dst isn't refcounted
* @skb: buffer
*/
static inline bool skb_dst_is_noref(const struct sk_buff *skb)
{
return (skb->_skb_refdst & SKB_DST_NOREF) && skb_dst(skb);
}
static inline struct rtable *skb_rtable(const struct sk_buff *skb)
{
return (struct rtable *)skb_dst(skb);
}
void kfree_skb(struct sk_buff *skb);
void kfree_skb_list(struct sk_buff *segs);
void skb_tx_error(struct sk_buff *skb);
void consume_skb(struct sk_buff *skb);
void __kfree_skb(struct sk_buff *skb);
extern struct kmem_cache *skbuff_head_cache;
void kfree_skb_partial(struct sk_buff *skb, bool head_stolen);
bool skb_try_coalesce(struct sk_buff *to, struct sk_buff *from,
bool *fragstolen, int *delta_truesize);
struct sk_buff *__alloc_skb(unsigned int size, gfp_t priority, int flags,
int node);
struct sk_buff *build_skb(void *data, unsigned int frag_size);
static inline struct sk_buff *alloc_skb(unsigned int size,
gfp_t priority)
{
return __alloc_skb(size, priority, 0, NUMA_NO_NODE);
}
static inline struct sk_buff *alloc_skb_fclone(unsigned int size,
gfp_t priority)
{
return __alloc_skb(size, priority, SKB_ALLOC_FCLONE, NUMA_NO_NODE);
}
struct sk_buff *__alloc_skb_head(gfp_t priority, int node);
static inline struct sk_buff *alloc_skb_head(gfp_t priority)
{
return __alloc_skb_head(priority, -1);
}
struct sk_buff *skb_morph(struct sk_buff *dst, struct sk_buff *src);
int skb_copy_ubufs(struct sk_buff *skb, gfp_t gfp_mask);
struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t priority);
struct sk_buff *skb_copy(const struct sk_buff *skb, gfp_t priority);
struct sk_buff *__pskb_copy_fclone(struct sk_buff *skb, int headroom,
gfp_t gfp_mask, bool fclone);
static inline struct sk_buff *__pskb_copy(struct sk_buff *skb, int headroom,
gfp_t gfp_mask)
{
return __pskb_copy_fclone(skb, headroom, gfp_mask, false);
}
int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail, gfp_t gfp_mask);
struct sk_buff *skb_realloc_headroom(struct sk_buff *skb,
unsigned int headroom);
struct sk_buff *skb_copy_expand(const struct sk_buff *skb, int newheadroom,
int newtailroom, gfp_t priority);
int skb_to_sgvec_nomark(struct sk_buff *skb, struct scatterlist *sg,
int offset, int len);
int skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset,
int len);
int skb_cow_data(struct sk_buff *skb, int tailbits, struct sk_buff **trailer);
int skb_pad(struct sk_buff *skb, int pad);
#define dev_kfree_skb(a) consume_skb(a)
int skb_append_datato_frags(struct sock *sk, struct sk_buff *skb,
int getfrag(void *from, char *to, int offset,
int len, int odd, struct sk_buff *skb),
void *from, int length);
[IPv4/IPv6]: UFO Scatter-gather approach Attached is kernel patch for UDP Fragmentation Offload (UFO) feature. 1. This patch incorporate the review comments by Jeff Garzik. 2. Renamed USO as UFO (UDP Fragmentation Offload) 3. udp sendfile support with UFO This patches uses scatter-gather feature of skb to generate large UDP datagram. Below is a "how-to" on changes required in network device driver to use the UFO interface. UDP Fragmentation Offload (UFO) Interface: ------------------------------------------- UFO is a feature wherein the Linux kernel network stack will offload the IP fragmentation functionality of large UDP datagram to hardware. This will reduce the overhead of stack in fragmenting the large UDP datagram to MTU sized packets 1) Drivers indicate their capability of UFO using dev->features |= NETIF_F_UFO | NETIF_F_HW_CSUM | NETIF_F_SG NETIF_F_HW_CSUM is required for UFO over ipv6. 2) UFO packet will be submitted for transmission using driver xmit routine. UFO packet will have a non-zero value for "skb_shinfo(skb)->ufo_size" skb_shinfo(skb)->ufo_size will indicate the length of data part in each IP fragment going out of the adapter after IP fragmentation by hardware. skb->data will contain MAC/IP/UDP header and skb_shinfo(skb)->frags[] contains the data payload. The skb->ip_summed will be set to CHECKSUM_HW indicating that hardware has to do checksum calculation. Hardware should compute the UDP checksum of complete datagram and also ip header checksum of each fragmented IP packet. For IPV6 the UFO provides the fragment identification-id in skb_shinfo(skb)->ip6_frag_id. The adapter should use this ID for generating IPv6 fragments. Signed-off-by: Ananda Raju <ananda.raju@neterion.com> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (forwarded) Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-10-19 05:46:41 +07:00
struct skb_seq_state {
__u32 lower_offset;
__u32 upper_offset;
__u32 frag_idx;
__u32 stepped_offset;
struct sk_buff *root_skb;
struct sk_buff *cur_skb;
__u8 *frag_data;
};
void skb_prepare_seq_read(struct sk_buff *skb, unsigned int from,
unsigned int to, struct skb_seq_state *st);
unsigned int skb_seq_read(unsigned int consumed, const u8 **data,
struct skb_seq_state *st);
void skb_abort_seq_read(struct skb_seq_state *st);
unsigned int skb_find_text(struct sk_buff *skb, unsigned int from,
unsigned int to, struct ts_config *config,
struct ts_state *state);
/*
* Packet hash types specify the type of hash in skb_set_hash.
*
* Hash types refer to the protocol layer addresses which are used to
* construct a packet's hash. The hashes are used to differentiate or identify
* flows of the protocol layer for the hash type. Hash types are either
* layer-2 (L2), layer-3 (L3), or layer-4 (L4).
*
* Properties of hashes:
*
* 1) Two packets in different flows have different hash values
* 2) Two packets in the same flow should have the same hash value
*
* A hash at a higher layer is considered to be more specific. A driver should
* set the most specific hash possible.
*
* A driver cannot indicate a more specific hash than the layer at which a hash
* was computed. For instance an L3 hash cannot be set as an L4 hash.
*
* A driver may indicate a hash level which is less specific than the
* actual layer the hash was computed on. For instance, a hash computed
* at L4 may be considered an L3 hash. This should only be done if the
* driver can't unambiguously determine that the HW computed the hash at
* the higher layer. Note that the "should" in the second property above
* permits this.
*/
enum pkt_hash_types {
PKT_HASH_TYPE_NONE, /* Undefined type */
PKT_HASH_TYPE_L2, /* Input: src_MAC, dest_MAC */
PKT_HASH_TYPE_L3, /* Input: src_IP, dst_IP */
PKT_HASH_TYPE_L4, /* Input: src_IP, dst_IP, src_port, dst_port */
};
static inline void
skb_set_hash(struct sk_buff *skb, __u32 hash, enum pkt_hash_types type)
{
skb->l4_hash = (type == PKT_HASH_TYPE_L4);
skb->hash = hash;
}
void __skb_get_hash(struct sk_buff *skb);
static inline __u32 skb_get_hash(struct sk_buff *skb)
{
if (!skb->l4_hash)
__skb_get_hash(skb);
return skb->hash;
}
static inline __u32 skb_get_hash_raw(const struct sk_buff *skb)
{
return skb->hash;
}
static inline void skb_clear_hash(struct sk_buff *skb)
{
skb->hash = 0;
skb->l4_hash = 0;
}
static inline void skb_clear_hash_if_not_l4(struct sk_buff *skb)
{
if (!skb->l4_hash)
skb_clear_hash(skb);
}
static inline void skb_copy_hash(struct sk_buff *to, const struct sk_buff *from)
{
to->hash = from->hash;
to->l4_hash = from->l4_hash;
};
#ifdef NET_SKBUFF_DATA_USES_OFFSET
static inline unsigned char *skb_end_pointer(const struct sk_buff *skb)
{
return skb->head + skb->end;
}
static inline unsigned int skb_end_offset(const struct sk_buff *skb)
{
return skb->end;
}
#else
static inline unsigned char *skb_end_pointer(const struct sk_buff *skb)
{
return skb->end;
}
static inline unsigned int skb_end_offset(const struct sk_buff *skb)
{
return skb->end - skb->head;
}
#endif
/* Internal */
#define skb_shinfo(SKB) ((struct skb_shared_info *)(skb_end_pointer(SKB)))
static inline struct skb_shared_hwtstamps *skb_hwtstamps(struct sk_buff *skb)
{
return &skb_shinfo(skb)->hwtstamps;
}
/**
* skb_queue_empty - check if a queue is empty
* @list: queue head
*
* Returns true if the queue is empty, false otherwise.
*/
static inline int skb_queue_empty(const struct sk_buff_head *list)
{
return list->next == (const struct sk_buff *) list;
}
/**
* skb_queue_is_last - check if skb is the last entry in the queue
* @list: queue head
* @skb: buffer
*
* Returns true if @skb is the last buffer on the list.
*/
static inline bool skb_queue_is_last(const struct sk_buff_head *list,
const struct sk_buff *skb)
{
return skb->next == (const struct sk_buff *) list;
}
tcp: Try to restore large SKBs while SACK processing During SACK processing, most of the benefits of TSO are eaten by the SACK blocks that one-by-one fragment SKBs to MSS sized chunks. Then we're in problems when cleanup work for them has to be done when a large cumulative ACK comes. Try to return back to pre-split state already while more and more SACK info gets discovered by combining newly discovered SACK areas with the previous skb if that's SACKed as well. This approach has a number of benefits: 1) The processing overhead is spread more equally over the RTT 2) Write queue has less skbs to process (affect everything which has to walk in the queue past the sacked areas) 3) Write queue is consistent whole the time, so no other parts of TCP has to be aware of this (this was not the case with some other approach that was, well, quite intrusive all around). 4) Clean_rtx_queue can release most of the pages using single put_page instead of previous PAGE_SIZE/mss+1 calls In case a hole is fully filled by the new SACK block, we attempt to combine the next skb too which allows construction of skbs that are even larger than what tso split them to and it handles hole per on every nth patterns that often occur during slow start overshoot pretty nicely. Though this to be really useful also a retransmission would have to get lost since cumulative ACKs advance one hole at a time in the most typical case. TODO: handle upwards only merging. That should be rather easy when segment is fully sacked but I'm leaving that as future work item (it won't make very large difference anyway since this current approach already covers quite a lot of normal cases). I was earlier thinking of some sophisticated way of tracking timestamps of the first and the last segment but later on realized that it won't be that necessary at all to store the timestamp of the last segment. The cases that can occur are basically either: 1) ambiguous => no sensible measurement can be taken anyway 2) non-ambiguous is due to reordering => having the timestamp of the last segment there is just skewing things more off than does some good since the ack got triggered by one of the holes (besides some substle issues that would make determining right hole/skb even harder problem). Anyway, it has nothing to do with this change then. I choose to route some abnormal looking cases with goto noop, some could be handled differently (eg., by stopping the walking at that skb but again). In general, they either shouldn't happen at all or are rare enough to make no difference in practice. In theory this change (as whole) could cause some macroscale regression (global) because of cache misses that are taken over the round-trip time but it gets very likely better because of much less (local) cache misses per other write queue walkers and the big recovery clearing cumulative ack. Worth to note that these benefits would be very easy to get also without TSO/GSO being on as long as the data is in pages so that we can merge them. Currently I won't let that happen because DSACK splitting at fragment that would mess up pcounts due to sk_can_gso in tcp_set_skb_tso_segs. Once DSACKs fragments gets avoided, we have some conditions that can be made less strict. TODO: I will probably have to convert the excessive pointer passing to struct sacktag_state... :-) My testing revealed that considerable amount of skbs couldn't be shifted because they were cloned (most likely still awaiting tx reclaim)... [The rest is considering future work instead since I got repeatably EFAULT to tcpdump's recvfrom when I added pskb_expand_head to deal with clones, so I separated that into another, later patch] ...To counter that, I gave up on the fifth advantage: 5) When growing previous SACK block, less allocs for new skbs are done, basically a new alloc is needed only when new hole is detected and when the previous skb runs out of frags space ...which now only happens of if reclaim is fast enough to dispose the clone before the SACK block comes in (the window is RTT long), otherwise we'll have to alloc some. With clones being handled I got these numbers (will be somewhat worse without that), taken with fine-grained mibs: TCPSackShifted 398 TCPSackMerged 877 TCPSackShiftFallback 320 TCPSACKCOLLAPSEFALLBACKGSO 0 TCPSACKCOLLAPSEFALLBACKSKBBITS 0 TCPSACKCOLLAPSEFALLBACKSKBDATA 0 TCPSACKCOLLAPSEFALLBACKBELOW 0 TCPSACKCOLLAPSEFALLBACKFIRST 1 TCPSACKCOLLAPSEFALLBACKPREVBITS 318 TCPSACKCOLLAPSEFALLBACKMSS 1 TCPSACKCOLLAPSEFALLBACKNOHEAD 0 TCPSACKCOLLAPSEFALLBACKSHIFT 0 TCPSACKCOLLAPSENOOPSEQ 0 TCPSACKCOLLAPSENOOPSMALLPCOUNT 0 TCPSACKCOLLAPSENOOPSMALLLEN 0 TCPSACKCOLLAPSEHOLE 12 Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-25 12:20:15 +07:00
/**
* skb_queue_is_first - check if skb is the first entry in the queue
* @list: queue head
* @skb: buffer
*
* Returns true if @skb is the first buffer on the list.
*/
static inline bool skb_queue_is_first(const struct sk_buff_head *list,
const struct sk_buff *skb)
{
return skb->prev == (const struct sk_buff *) list;
tcp: Try to restore large SKBs while SACK processing During SACK processing, most of the benefits of TSO are eaten by the SACK blocks that one-by-one fragment SKBs to MSS sized chunks. Then we're in problems when cleanup work for them has to be done when a large cumulative ACK comes. Try to return back to pre-split state already while more and more SACK info gets discovered by combining newly discovered SACK areas with the previous skb if that's SACKed as well. This approach has a number of benefits: 1) The processing overhead is spread more equally over the RTT 2) Write queue has less skbs to process (affect everything which has to walk in the queue past the sacked areas) 3) Write queue is consistent whole the time, so no other parts of TCP has to be aware of this (this was not the case with some other approach that was, well, quite intrusive all around). 4) Clean_rtx_queue can release most of the pages using single put_page instead of previous PAGE_SIZE/mss+1 calls In case a hole is fully filled by the new SACK block, we attempt to combine the next skb too which allows construction of skbs that are even larger than what tso split them to and it handles hole per on every nth patterns that often occur during slow start overshoot pretty nicely. Though this to be really useful also a retransmission would have to get lost since cumulative ACKs advance one hole at a time in the most typical case. TODO: handle upwards only merging. That should be rather easy when segment is fully sacked but I'm leaving that as future work item (it won't make very large difference anyway since this current approach already covers quite a lot of normal cases). I was earlier thinking of some sophisticated way of tracking timestamps of the first and the last segment but later on realized that it won't be that necessary at all to store the timestamp of the last segment. The cases that can occur are basically either: 1) ambiguous => no sensible measurement can be taken anyway 2) non-ambiguous is due to reordering => having the timestamp of the last segment there is just skewing things more off than does some good since the ack got triggered by one of the holes (besides some substle issues that would make determining right hole/skb even harder problem). Anyway, it has nothing to do with this change then. I choose to route some abnormal looking cases with goto noop, some could be handled differently (eg., by stopping the walking at that skb but again). In general, they either shouldn't happen at all or are rare enough to make no difference in practice. In theory this change (as whole) could cause some macroscale regression (global) because of cache misses that are taken over the round-trip time but it gets very likely better because of much less (local) cache misses per other write queue walkers and the big recovery clearing cumulative ack. Worth to note that these benefits would be very easy to get also without TSO/GSO being on as long as the data is in pages so that we can merge them. Currently I won't let that happen because DSACK splitting at fragment that would mess up pcounts due to sk_can_gso in tcp_set_skb_tso_segs. Once DSACKs fragments gets avoided, we have some conditions that can be made less strict. TODO: I will probably have to convert the excessive pointer passing to struct sacktag_state... :-) My testing revealed that considerable amount of skbs couldn't be shifted because they were cloned (most likely still awaiting tx reclaim)... [The rest is considering future work instead since I got repeatably EFAULT to tcpdump's recvfrom when I added pskb_expand_head to deal with clones, so I separated that into another, later patch] ...To counter that, I gave up on the fifth advantage: 5) When growing previous SACK block, less allocs for new skbs are done, basically a new alloc is needed only when new hole is detected and when the previous skb runs out of frags space ...which now only happens of if reclaim is fast enough to dispose the clone before the SACK block comes in (the window is RTT long), otherwise we'll have to alloc some. With clones being handled I got these numbers (will be somewhat worse without that), taken with fine-grained mibs: TCPSackShifted 398 TCPSackMerged 877 TCPSackShiftFallback 320 TCPSACKCOLLAPSEFALLBACKGSO 0 TCPSACKCOLLAPSEFALLBACKSKBBITS 0 TCPSACKCOLLAPSEFALLBACKSKBDATA 0 TCPSACKCOLLAPSEFALLBACKBELOW 0 TCPSACKCOLLAPSEFALLBACKFIRST 1 TCPSACKCOLLAPSEFALLBACKPREVBITS 318 TCPSACKCOLLAPSEFALLBACKMSS 1 TCPSACKCOLLAPSEFALLBACKNOHEAD 0 TCPSACKCOLLAPSEFALLBACKSHIFT 0 TCPSACKCOLLAPSENOOPSEQ 0 TCPSACKCOLLAPSENOOPSMALLPCOUNT 0 TCPSACKCOLLAPSENOOPSMALLLEN 0 TCPSACKCOLLAPSEHOLE 12 Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-25 12:20:15 +07:00
}
/**
* skb_queue_next - return the next packet in the queue
* @list: queue head
* @skb: current buffer
*
* Return the next packet in @list after @skb. It is only valid to
* call this if skb_queue_is_last() evaluates to false.
*/
static inline struct sk_buff *skb_queue_next(const struct sk_buff_head *list,
const struct sk_buff *skb)
{
/* This BUG_ON may seem severe, but if we just return then we
* are going to dereference garbage.
*/
BUG_ON(skb_queue_is_last(list, skb));
return skb->next;
}
tcp: Try to restore large SKBs while SACK processing During SACK processing, most of the benefits of TSO are eaten by the SACK blocks that one-by-one fragment SKBs to MSS sized chunks. Then we're in problems when cleanup work for them has to be done when a large cumulative ACK comes. Try to return back to pre-split state already while more and more SACK info gets discovered by combining newly discovered SACK areas with the previous skb if that's SACKed as well. This approach has a number of benefits: 1) The processing overhead is spread more equally over the RTT 2) Write queue has less skbs to process (affect everything which has to walk in the queue past the sacked areas) 3) Write queue is consistent whole the time, so no other parts of TCP has to be aware of this (this was not the case with some other approach that was, well, quite intrusive all around). 4) Clean_rtx_queue can release most of the pages using single put_page instead of previous PAGE_SIZE/mss+1 calls In case a hole is fully filled by the new SACK block, we attempt to combine the next skb too which allows construction of skbs that are even larger than what tso split them to and it handles hole per on every nth patterns that often occur during slow start overshoot pretty nicely. Though this to be really useful also a retransmission would have to get lost since cumulative ACKs advance one hole at a time in the most typical case. TODO: handle upwards only merging. That should be rather easy when segment is fully sacked but I'm leaving that as future work item (it won't make very large difference anyway since this current approach already covers quite a lot of normal cases). I was earlier thinking of some sophisticated way of tracking timestamps of the first and the last segment but later on realized that it won't be that necessary at all to store the timestamp of the last segment. The cases that can occur are basically either: 1) ambiguous => no sensible measurement can be taken anyway 2) non-ambiguous is due to reordering => having the timestamp of the last segment there is just skewing things more off than does some good since the ack got triggered by one of the holes (besides some substle issues that would make determining right hole/skb even harder problem). Anyway, it has nothing to do with this change then. I choose to route some abnormal looking cases with goto noop, some could be handled differently (eg., by stopping the walking at that skb but again). In general, they either shouldn't happen at all or are rare enough to make no difference in practice. In theory this change (as whole) could cause some macroscale regression (global) because of cache misses that are taken over the round-trip time but it gets very likely better because of much less (local) cache misses per other write queue walkers and the big recovery clearing cumulative ack. Worth to note that these benefits would be very easy to get also without TSO/GSO being on as long as the data is in pages so that we can merge them. Currently I won't let that happen because DSACK splitting at fragment that would mess up pcounts due to sk_can_gso in tcp_set_skb_tso_segs. Once DSACKs fragments gets avoided, we have some conditions that can be made less strict. TODO: I will probably have to convert the excessive pointer passing to struct sacktag_state... :-) My testing revealed that considerable amount of skbs couldn't be shifted because they were cloned (most likely still awaiting tx reclaim)... [The rest is considering future work instead since I got repeatably EFAULT to tcpdump's recvfrom when I added pskb_expand_head to deal with clones, so I separated that into another, later patch] ...To counter that, I gave up on the fifth advantage: 5) When growing previous SACK block, less allocs for new skbs are done, basically a new alloc is needed only when new hole is detected and when the previous skb runs out of frags space ...which now only happens of if reclaim is fast enough to dispose the clone before the SACK block comes in (the window is RTT long), otherwise we'll have to alloc some. With clones being handled I got these numbers (will be somewhat worse without that), taken with fine-grained mibs: TCPSackShifted 398 TCPSackMerged 877 TCPSackShiftFallback 320 TCPSACKCOLLAPSEFALLBACKGSO 0 TCPSACKCOLLAPSEFALLBACKSKBBITS 0 TCPSACKCOLLAPSEFALLBACKSKBDATA 0 TCPSACKCOLLAPSEFALLBACKBELOW 0 TCPSACKCOLLAPSEFALLBACKFIRST 1 TCPSACKCOLLAPSEFALLBACKPREVBITS 318 TCPSACKCOLLAPSEFALLBACKMSS 1 TCPSACKCOLLAPSEFALLBACKNOHEAD 0 TCPSACKCOLLAPSEFALLBACKSHIFT 0 TCPSACKCOLLAPSENOOPSEQ 0 TCPSACKCOLLAPSENOOPSMALLPCOUNT 0 TCPSACKCOLLAPSENOOPSMALLLEN 0 TCPSACKCOLLAPSEHOLE 12 Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-25 12:20:15 +07:00
/**
* skb_queue_prev - return the prev packet in the queue
* @list: queue head
* @skb: current buffer
*
* Return the prev packet in @list before @skb. It is only valid to
* call this if skb_queue_is_first() evaluates to false.
*/
static inline struct sk_buff *skb_queue_prev(const struct sk_buff_head *list,
const struct sk_buff *skb)
{
/* This BUG_ON may seem severe, but if we just return then we
* are going to dereference garbage.
*/
BUG_ON(skb_queue_is_first(list, skb));
return skb->prev;
}
/**
* skb_get - reference buffer
* @skb: buffer to reference
*
* Makes another reference to a socket buffer and returns a pointer
* to the buffer.
*/
static inline struct sk_buff *skb_get(struct sk_buff *skb)
{
atomic_inc(&skb->users);
return skb;
}
/*
* If users == 1, we are the only owner and are can avoid redundant
* atomic change.
*/
/**
* skb_cloned - is the buffer a clone
* @skb: buffer to check
*
* Returns true if the buffer was generated with skb_clone() and is
* one of multiple shared copies of the buffer. Cloned buffers are
* shared data so must not be written to under normal circumstances.
*/
static inline int skb_cloned(const struct sk_buff *skb)
{
return skb->cloned &&
(atomic_read(&skb_shinfo(skb)->dataref) & SKB_DATAREF_MASK) != 1;
}
static inline int skb_unclone(struct sk_buff *skb, gfp_t pri)
{
might_sleep_if(pri & __GFP_WAIT);
if (skb_cloned(skb))
return pskb_expand_head(skb, 0, 0, pri);
return 0;
}
/**
* skb_header_cloned - is the header a clone
* @skb: buffer to check
*
* Returns true if modifying the header part of the buffer requires
* the data to be copied.
*/
static inline int skb_header_cloned(const struct sk_buff *skb)
{
int dataref;
if (!skb->cloned)
return 0;
dataref = atomic_read(&skb_shinfo(skb)->dataref);
dataref = (dataref & SKB_DATAREF_MASK) - (dataref >> SKB_DATAREF_SHIFT);
return dataref != 1;
}
/**
* skb_header_release - release reference to header
* @skb: buffer to operate on
*
* Drop a reference to the header part of the buffer. This is done
* by acquiring a payload reference. You must not read from the header
* part of skb->data after this.
*/
static inline void skb_header_release(struct sk_buff *skb)
{
BUG_ON(skb->nohdr);
skb->nohdr = 1;
atomic_add(1 << SKB_DATAREF_SHIFT, &skb_shinfo(skb)->dataref);
}
/**
* skb_shared - is the buffer shared
* @skb: buffer to check
*
* Returns true if more than one person has a reference to this
* buffer.
*/
static inline int skb_shared(const struct sk_buff *skb)
{
return atomic_read(&skb->users) != 1;
}
/**
* skb_share_check - check if buffer is shared and if so clone it
* @skb: buffer to check
* @pri: priority for memory allocation
*
* If the buffer is shared the buffer is cloned and the old copy
* drops a reference. A new clone with a single reference is returned.
* If the buffer is not shared the original buffer is returned. When
* being called from interrupt status or with spinlocks held pri must
* be GFP_ATOMIC.
*
* NULL is returned on a memory allocation failure.
*/
static inline struct sk_buff *skb_share_check(struct sk_buff *skb, gfp_t pri)
{
might_sleep_if(pri & __GFP_WAIT);
if (skb_shared(skb)) {
struct sk_buff *nskb = skb_clone(skb, pri);
if (likely(nskb))
consume_skb(skb);
else
kfree_skb(skb);
skb = nskb;
}
return skb;
}
/*
* Copy shared buffers into a new sk_buff. We effectively do COW on
* packets to handle cases where we have a local reader and forward
* and a couple of other messy ones. The normal one is tcpdumping
* a packet thats being forwarded.
*/
/**
* skb_unshare - make a copy of a shared buffer
* @skb: buffer to check
* @pri: priority for memory allocation
*
* If the socket buffer is a clone then this function creates a new
* copy of the data, drops a reference count on the old copy and returns
* the new copy with the reference count at 1. If the buffer is not a clone
* the original buffer is returned. When called with a spinlock held or
* from interrupt state @pri must be %GFP_ATOMIC
*
* %NULL is returned on a memory allocation failure.
*/
static inline struct sk_buff *skb_unshare(struct sk_buff *skb,
gfp_t pri)
{
might_sleep_if(pri & __GFP_WAIT);
if (skb_cloned(skb)) {
struct sk_buff *nskb = skb_copy(skb, pri);
kfree_skb(skb); /* Free our shared copy */
skb = nskb;
}
return skb;
}
/**
* skb_peek - peek at the head of an &sk_buff_head
* @list_: list to peek at
*
* Peek an &sk_buff. Unlike most other operations you _MUST_
* be careful with this one. A peek leaves the buffer on the
* list and someone else may run off with it. You must hold
* the appropriate locks or have a private queue to do this.
*
* Returns %NULL for an empty list or a pointer to the head element.
* The reference count is not incremented and the reference is therefore
* volatile. Use with caution.
*/
static inline struct sk_buff *skb_peek(const struct sk_buff_head *list_)
{
struct sk_buff *skb = list_->next;
if (skb == (struct sk_buff *)list_)
skb = NULL;
return skb;
}
/**
* skb_peek_next - peek skb following the given one from a queue
* @skb: skb to start from
* @list_: list to peek at
*
* Returns %NULL when the end of the list is met or a pointer to the
* next element. The reference count is not incremented and the
* reference is therefore volatile. Use with caution.
*/
static inline struct sk_buff *skb_peek_next(struct sk_buff *skb,
const struct sk_buff_head *list_)
{
struct sk_buff *next = skb->next;
if (next == (struct sk_buff *)list_)
next = NULL;
return next;
}
/**
* skb_peek_tail - peek at the tail of an &sk_buff_head
* @list_: list to peek at
*
* Peek an &sk_buff. Unlike most other operations you _MUST_
* be careful with this one. A peek leaves the buffer on the
* list and someone else may run off with it. You must hold
* the appropriate locks or have a private queue to do this.
*
* Returns %NULL for an empty list or a pointer to the tail element.
* The reference count is not incremented and the reference is therefore
* volatile. Use with caution.
*/
static inline struct sk_buff *skb_peek_tail(const struct sk_buff_head *list_)
{
struct sk_buff *skb = list_->prev;
if (skb == (struct sk_buff *)list_)
skb = NULL;
return skb;
}
/**
* skb_queue_len - get queue length
* @list_: list to measure
*
* Return the length of an &sk_buff queue.
*/
static inline __u32 skb_queue_len(const struct sk_buff_head *list_)
{
return list_->qlen;
}
/**
* __skb_queue_head_init - initialize non-spinlock portions of sk_buff_head
* @list: queue to initialize
*
* This initializes only the list and queue length aspects of
* an sk_buff_head object. This allows to initialize the list
* aspects of an sk_buff_head without reinitializing things like
* the spinlock. It can also be used for on-stack sk_buff_head
* objects where the spinlock is known to not be used.
*/
static inline void __skb_queue_head_init(struct sk_buff_head *list)
{
list->prev = list->next = (struct sk_buff *)list;
list->qlen = 0;
}
/*
* This function creates a split out lock class for each invocation;
* this is needed for now since a whole lot of users of the skb-queue
* infrastructure in drivers have different locking usage (in hardirq)
* than the networking core (in softirq only). In the long run either the
* network layer or drivers should need annotation to consolidate the
* main types of usage into 3 classes.
*/
static inline void skb_queue_head_init(struct sk_buff_head *list)
{
spin_lock_init(&list->lock);
__skb_queue_head_init(list);
}
static inline void skb_queue_head_init_class(struct sk_buff_head *list,
struct lock_class_key *class)
{
skb_queue_head_init(list);
lockdep_set_class(&list->lock, class);
}
/*
* Insert an sk_buff on a list.
*
* The "__skb_xxxx()" functions are the non-atomic ones that
* can only be called with interrupts disabled.
*/
void skb_insert(struct sk_buff *old, struct sk_buff *newsk,
struct sk_buff_head *list);
static inline void __skb_insert(struct sk_buff *newsk,
struct sk_buff *prev, struct sk_buff *next,
struct sk_buff_head *list)
{
newsk->next = next;
newsk->prev = prev;
next->prev = prev->next = newsk;
list->qlen++;
}
static inline void __skb_queue_splice(const struct sk_buff_head *list,
struct sk_buff *prev,
struct sk_buff *next)
{
struct sk_buff *first = list->next;
struct sk_buff *last = list->prev;
first->prev = prev;
prev->next = first;
last->next = next;
next->prev = last;
}
/**
* skb_queue_splice - join two skb lists, this is designed for stacks
* @list: the new list to add
* @head: the place to add it in the first list
*/
static inline void skb_queue_splice(const struct sk_buff_head *list,
struct sk_buff_head *head)
{
if (!skb_queue_empty(list)) {
__skb_queue_splice(list, (struct sk_buff *) head, head->next);
head->qlen += list->qlen;
}
}
/**
* skb_queue_splice_init - join two skb lists and reinitialise the emptied list
* @list: the new list to add
* @head: the place to add it in the first list
*
* The list at @list is reinitialised
*/
static inline void skb_queue_splice_init(struct sk_buff_head *list,
struct sk_buff_head *head)
{
if (!skb_queue_empty(list)) {
__skb_queue_splice(list, (struct sk_buff *) head, head->next);
head->qlen += list->qlen;
__skb_queue_head_init(list);
}
}
/**
* skb_queue_splice_tail - join two skb lists, each list being a queue
* @list: the new list to add
* @head: the place to add it in the first list
*/
static inline void skb_queue_splice_tail(const struct sk_buff_head *list,
struct sk_buff_head *head)
{
if (!skb_queue_empty(list)) {
__skb_queue_splice(list, head->prev, (struct sk_buff *) head);
head->qlen += list->qlen;
}
}
/**
* skb_queue_splice_tail_init - join two skb lists and reinitialise the emptied list
* @list: the new list to add
* @head: the place to add it in the first list
*
* Each of the lists is a queue.
* The list at @list is reinitialised
*/
static inline void skb_queue_splice_tail_init(struct sk_buff_head *list,
struct sk_buff_head *head)
{
if (!skb_queue_empty(list)) {
__skb_queue_splice(list, head->prev, (struct sk_buff *) head);
head->qlen += list->qlen;
__skb_queue_head_init(list);
}
}
/**
* __skb_queue_after - queue a buffer at the list head
* @list: list to use
* @prev: place after this buffer
* @newsk: buffer to queue
*
* Queue a buffer int the middle of a list. This function takes no locks
* and you must therefore hold required locks before calling it.
*
* A buffer cannot be placed on two lists at the same time.
*/
static inline void __skb_queue_after(struct sk_buff_head *list,
struct sk_buff *prev,
struct sk_buff *newsk)
{
__skb_insert(newsk, prev, prev->next, list);
}
void skb_append(struct sk_buff *old, struct sk_buff *newsk,
struct sk_buff_head *list);
static inline void __skb_queue_before(struct sk_buff_head *list,
struct sk_buff *next,
struct sk_buff *newsk)
{
__skb_insert(newsk, next->prev, next, list);
}
/**
* __skb_queue_head - queue a buffer at the list head
* @list: list to use
* @newsk: buffer to queue
*
* Queue a buffer at the start of a list. This function takes no locks
* and you must therefore hold required locks before calling it.
*
* A buffer cannot be placed on two lists at the same time.
*/
void skb_queue_head(struct sk_buff_head *list, struct sk_buff *newsk);
static inline void __skb_queue_head(struct sk_buff_head *list,
struct sk_buff *newsk)
{
__skb_queue_after(list, (struct sk_buff *)list, newsk);
}
/**
* __skb_queue_tail - queue a buffer at the list tail
* @list: list to use
* @newsk: buffer to queue
*
* Queue a buffer at the end of a list. This function takes no locks
* and you must therefore hold required locks before calling it.
*
* A buffer cannot be placed on two lists at the same time.
*/
void skb_queue_tail(struct sk_buff_head *list, struct sk_buff *newsk);
static inline void __skb_queue_tail(struct sk_buff_head *list,
struct sk_buff *newsk)
{
__skb_queue_before(list, (struct sk_buff *)list, newsk);
}
/*
* remove sk_buff from list. _Must_ be called atomically, and with
* the list known..
*/
void skb_unlink(struct sk_buff *skb, struct sk_buff_head *list);
static inline void __skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
{
struct sk_buff *next, *prev;
list->qlen--;
next = skb->next;
prev = skb->prev;
skb->next = skb->prev = NULL;
next->prev = prev;
prev->next = next;
}
/**
* __skb_dequeue - remove from the head of the queue
* @list: list to dequeue from
*
* Remove the head of the list. This function does not take any locks
* so must be used with appropriate locks held only. The head item is
* returned or %NULL if the list is empty.
*/
struct sk_buff *skb_dequeue(struct sk_buff_head *list);
static inline struct sk_buff *__skb_dequeue(struct sk_buff_head *list)
{
struct sk_buff *skb = skb_peek(list);
if (skb)
__skb_unlink(skb, list);
return skb;
}
/**
* __skb_dequeue_tail - remove from the tail of the queue
* @list: list to dequeue from
*
* Remove the tail of the list. This function does not take any locks
* so must be used with appropriate locks held only. The tail item is
* returned or %NULL if the list is empty.
*/
struct sk_buff *skb_dequeue_tail(struct sk_buff_head *list);
static inline struct sk_buff *__skb_dequeue_tail(struct sk_buff_head *list)
{
struct sk_buff *skb = skb_peek_tail(list);
if (skb)
__skb_unlink(skb, list);
return skb;
}
static inline bool skb_is_nonlinear(const struct sk_buff *skb)
{
return skb->data_len;
}
static inline unsigned int skb_headlen(const struct sk_buff *skb)
{
return skb->len - skb->data_len;
}
static inline int skb_pagelen(const struct sk_buff *skb)
{
int i, len = 0;
for (i = (int)skb_shinfo(skb)->nr_frags - 1; i >= 0; i--)
len += skb_frag_size(&skb_shinfo(skb)->frags[i]);
return len + skb_headlen(skb);
}
/**
* __skb_fill_page_desc - initialise a paged fragment in an skb
* @skb: buffer containing fragment to be initialised
* @i: paged fragment index to initialise
* @page: the page to use for this fragment
* @off: the offset to the data with @page
* @size: the length of the data
*
* Initialises the @i'th fragment of @skb to point to &size bytes at
* offset @off within @page.
*
* Does not take any additional reference on the fragment.
*/
static inline void __skb_fill_page_desc(struct sk_buff *skb, int i,
struct page *page, int off, int size)
{
skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
/*
* Propagate page->pfmemalloc to the skb if we can. The problem is
* that not all callers have unique ownership of the page. If
* pfmemalloc is set, we check the mapping as a mapping implies
* page->index is set (index and pfmemalloc share space).
* If it's a valid mapping, we cannot use page->pfmemalloc but we
* do not lose pfmemalloc information as the pages would not be
* allocated using __GFP_MEMALLOC.
*/
frag->page.p = page;
frag->page_offset = off;
skb_frag_size_set(frag, size);
skb: Propagate pfmemalloc on skb from head page only Hi. I'm trying to send big chunks of memory from application address space via TCP socket using vmsplice + splice like this mem = mmap(128Mb); vmsplice(pipe[1], mem); /* splice memory into pipe */ splice(pipe[0], tcp_socket); /* send it into network */ When I'm lucky and a huge page splices into the pipe and then into the socket _and_ client and server ends of the TCP connection are on the same host, communicating via lo, the whole connection gets stuck! The sending queue becomes full and app stops writing/splicing more into it, but the receiving queue remains empty, and that's why. The __skb_fill_page_desc observes a tail page of a huge page and erroneously propagates its page->pfmemalloc value onto socket (the pfmemalloc on tail pages contain garbage). Then this skb->pfmemalloc leaks through lo and due to the tcp_v4_rcv sk_filter if (skb->pfmemalloc && !sock_flag(sk, SOCK_MEMALLOC)) /* true */ return -ENOMEM goto release_and_discard; no packets reach the socket. Even TCP re-transmits are dropped by this, as skb cloning clones the pfmemalloc flag as well. That said, here's the proper page->pfmemalloc propagation onto socket: we must check the huge-page's head page only, other pages' pfmemalloc and mapping values do not contain what is expected in this place. However, I'm not sure whether this fix is _complete_, since pfmemalloc propagation via lo also oesn't look great. Both, bit propagation from page to skb and this check in sk_filter, were introduced by c48a11c7 (netvm: propagate page->pfmemalloc to skb), in v3.5 so Mel and stable@ are in Cc. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-14 10:29:40 +07:00
page = compound_head(page);
if (page->pfmemalloc && !page->mapping)
skb->pfmemalloc = true;
}
/**
* skb_fill_page_desc - initialise a paged fragment in an skb
* @skb: buffer containing fragment to be initialised
* @i: paged fragment index to initialise
* @page: the page to use for this fragment
* @off: the offset to the data with @page
* @size: the length of the data
*
* As per __skb_fill_page_desc() -- initialises the @i'th fragment of
* @skb to point to @size bytes at offset @off within @page. In
* addition updates @skb such that @i is the last fragment.
*
* Does not take any additional reference on the fragment.
*/
static inline void skb_fill_page_desc(struct sk_buff *skb, int i,
struct page *page, int off, int size)
{
__skb_fill_page_desc(skb, i, page, off, size);
skb_shinfo(skb)->nr_frags = i + 1;
}
void skb_add_rx_frag(struct sk_buff *skb, int i, struct page *page, int off,
int size, unsigned int truesize);
void skb_coalesce_rx_frag(struct sk_buff *skb, int i, int size,
unsigned int truesize);
#define SKB_PAGE_ASSERT(skb) BUG_ON(skb_shinfo(skb)->nr_frags)
#define SKB_FRAG_ASSERT(skb) BUG_ON(skb_has_frag_list(skb))
#define SKB_LINEAR_ASSERT(skb) BUG_ON(skb_is_nonlinear(skb))
#ifdef NET_SKBUFF_DATA_USES_OFFSET
static inline unsigned char *skb_tail_pointer(const struct sk_buff *skb)
{
return skb->head + skb->tail;
}
static inline void skb_reset_tail_pointer(struct sk_buff *skb)
{
skb->tail = skb->data - skb->head;
}
static inline void skb_set_tail_pointer(struct sk_buff *skb, const int offset)
{
skb_reset_tail_pointer(skb);
skb->tail += offset;
}
#else /* NET_SKBUFF_DATA_USES_OFFSET */
static inline unsigned char *skb_tail_pointer(const struct sk_buff *skb)
{
return skb->tail;
}
static inline void skb_reset_tail_pointer(struct sk_buff *skb)
{
skb->tail = skb->data;
}
static inline void skb_set_tail_pointer(struct sk_buff *skb, const int offset)
{
skb->tail = skb->data + offset;
}
#endif /* NET_SKBUFF_DATA_USES_OFFSET */
/*
* Add data to an sk_buff
*/
unsigned char *pskb_put(struct sk_buff *skb, struct sk_buff *tail, int len);
unsigned char *skb_put(struct sk_buff *skb, unsigned int len);
static inline unsigned char *__skb_put(struct sk_buff *skb, unsigned int len)
{
unsigned char *tmp = skb_tail_pointer(skb);
SKB_LINEAR_ASSERT(skb);
skb->tail += len;
skb->len += len;
return tmp;
}
unsigned char *skb_push(struct sk_buff *skb, unsigned int len);
static inline unsigned char *__skb_push(struct sk_buff *skb, unsigned int len)
{
skb->data -= len;
skb->len += len;
return skb->data;
}
unsigned char *skb_pull(struct sk_buff *skb, unsigned int len);
static inline unsigned char *__skb_pull(struct sk_buff *skb, unsigned int len)
{
skb->len -= len;
BUG_ON(skb->len < skb->data_len);
return skb->data += len;
}
static inline unsigned char *skb_pull_inline(struct sk_buff *skb, unsigned int len)
{
return unlikely(len > skb->len) ? NULL : __skb_pull(skb, len);
}
unsigned char *__pskb_pull_tail(struct sk_buff *skb, int delta);
static inline unsigned char *__pskb_pull(struct sk_buff *skb, unsigned int len)
{
if (len > skb_headlen(skb) &&
!__pskb_pull_tail(skb, len - skb_headlen(skb)))
return NULL;
skb->len -= len;
return skb->data += len;
}
static inline unsigned char *pskb_pull(struct sk_buff *skb, unsigned int len)
{
return unlikely(len > skb->len) ? NULL : __pskb_pull(skb, len);
}
static inline int pskb_may_pull(struct sk_buff *skb, unsigned int len)
{
if (likely(len <= skb_headlen(skb)))
return 1;
if (unlikely(len > skb->len))
return 0;
return __pskb_pull_tail(skb, len - skb_headlen(skb)) != NULL;
}
/**
* skb_headroom - bytes at buffer head
* @skb: buffer to check
*
* Return the number of bytes of free space at the head of an &sk_buff.
*/
static inline unsigned int skb_headroom(const struct sk_buff *skb)
{
return skb->data - skb->head;
}
/**
* skb_tailroom - bytes at buffer end
* @skb: buffer to check
*
* Return the number of bytes of free space at the tail of an sk_buff
*/
static inline int skb_tailroom(const struct sk_buff *skb)
{
return skb_is_nonlinear(skb) ? 0 : skb->end - skb->tail;
}
/**
* skb_availroom - bytes at buffer end
* @skb: buffer to check
*
* Return the number of bytes of free space at the tail of an sk_buff
* allocated by sk_stream_alloc()
*/
static inline int skb_availroom(const struct sk_buff *skb)
{
if (skb_is_nonlinear(skb))
return 0;
return skb->end - skb->tail - skb->reserved_tailroom;
}
/**
* skb_reserve - adjust headroom
* @skb: buffer to alter
* @len: bytes to move
*
* Increase the headroom of an empty &sk_buff by reducing the tail
* room. This is only allowed for an empty buffer.
*/
static inline void skb_reserve(struct sk_buff *skb, int len)
{
skb->data += len;
skb->tail += len;
}
static inline void skb_reset_inner_headers(struct sk_buff *skb)
{
skb->inner_mac_header = skb->mac_header;
skb->inner_network_header = skb->network_header;
skb->inner_transport_header = skb->transport_header;
}
static inline void skb_reset_mac_len(struct sk_buff *skb)
{
skb->mac_len = skb->network_header - skb->mac_header;
}
static inline unsigned char *skb_inner_transport_header(const struct sk_buff
*skb)
{
return skb->head + skb->inner_transport_header;
}
static inline void skb_reset_inner_transport_header(struct sk_buff *skb)
{
skb->inner_transport_header = skb->data - skb->head;
}
static inline void skb_set_inner_transport_header(struct sk_buff *skb,
const int offset)
{
skb_reset_inner_transport_header(skb);
skb->inner_transport_header += offset;
}
static inline unsigned char *skb_inner_network_header(const struct sk_buff *skb)
{
return skb->head + skb->inner_network_header;
}
static inline void skb_reset_inner_network_header(struct sk_buff *skb)
{
skb->inner_network_header = skb->data - skb->head;
}
static inline void skb_set_inner_network_header(struct sk_buff *skb,
const int offset)
{
skb_reset_inner_network_header(skb);
skb->inner_network_header += offset;
}
static inline unsigned char *skb_inner_mac_header(const struct sk_buff *skb)
{
return skb->head + skb->inner_mac_header;
}
static inline void skb_reset_inner_mac_header(struct sk_buff *skb)
{
skb->inner_mac_header = skb->data - skb->head;
}
static inline void skb_set_inner_mac_header(struct sk_buff *skb,
const int offset)
{
skb_reset_inner_mac_header(skb);
skb->inner_mac_header += offset;
}
static inline bool skb_transport_header_was_set(const struct sk_buff *skb)
{
return skb->transport_header != (typeof(skb->transport_header))~0U;
}
static inline unsigned char *skb_transport_header(const struct sk_buff *skb)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
return skb->head + skb->transport_header;
}
static inline void skb_reset_transport_header(struct sk_buff *skb)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
skb->transport_header = skb->data - skb->head;
}
static inline void skb_set_transport_header(struct sk_buff *skb,
const int offset)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
skb_reset_transport_header(skb);
skb->transport_header += offset;
}
static inline unsigned char *skb_network_header(const struct sk_buff *skb)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
return skb->head + skb->network_header;
}
static inline void skb_reset_network_header(struct sk_buff *skb)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
skb->network_header = skb->data - skb->head;
}
static inline void skb_set_network_header(struct sk_buff *skb, const int offset)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
skb_reset_network_header(skb);
skb->network_header += offset;
}
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
static inline unsigned char *skb_mac_header(const struct sk_buff *skb)
{
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
return skb->head + skb->mac_header;
}
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
static inline int skb_mac_header_was_set(const struct sk_buff *skb)
{
return skb->mac_header != (typeof(skb->mac_header))~0U;
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
}
static inline void skb_reset_mac_header(struct sk_buff *skb)
{
skb->mac_header = skb->data - skb->head;
}
static inline void skb_set_mac_header(struct sk_buff *skb, const int offset)
{
skb_reset_mac_header(skb);
skb->mac_header += offset;
}
static inline void skb_pop_mac_header(struct sk_buff *skb)
{
skb->mac_header = skb->network_header;
}
static inline void skb_probe_transport_header(struct sk_buff *skb,
const int offset_hint)
{
struct flow_keys keys;
if (skb_transport_header_was_set(skb))
return;
else if (skb_flow_dissect(skb, &keys))
skb_set_transport_header(skb, keys.thoff);
else
skb_set_transport_header(skb, offset_hint);
}
static inline void skb_mac_header_rebuild(struct sk_buff *skb)
{
if (skb_mac_header_was_set(skb)) {
const unsigned char *old_mac = skb_mac_header(skb);
skb_set_mac_header(skb, -skb->mac_len);
memmove(skb_mac_header(skb), old_mac, skb->mac_len);
}
}
static inline int skb_checksum_start_offset(const struct sk_buff *skb)
{
return skb->csum_start - skb_headroom(skb);
}
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
static inline int skb_transport_offset(const struct sk_buff *skb)
{
return skb_transport_header(skb) - skb->data;
}
static inline u32 skb_network_header_len(const struct sk_buff *skb)
{
return skb->transport_header - skb->network_header;
}
static inline u32 skb_inner_network_header_len(const struct sk_buff *skb)
{
return skb->inner_transport_header - skb->inner_network_header;
}
[SK_BUFF]: Use offsets for skb->{mac,network,transport}_header on 64bit architectures With this we save 8 bytes per network packet, leaving a 4 bytes hole to be used in further shrinking work, likely with the offsetization of other pointers, such as ->{data,tail,end}, at the cost of adds, that were minimized by the usual practice of setting skb->{mac,nh,n}.raw to a local variable that is then accessed multiple times in each function, it also is not more expensive than before with regards to most of the handling of such headers, like setting one of these headers to another (transport to network, etc), or subtracting, adding to/from it, comparing them, etc. Now we have this layout for sk_buff on a x86_64 machine: [acme@mica net-2.6.22]$ pahole vmlinux sk_buff struct sk_buff { struct sk_buff * next; /* 0 8 */ struct sk_buff * prev; /* 8 8 */ struct rb_node rb; /* 16 24 */ struct sock * sk; /* 40 8 */ ktime_t tstamp; /* 48 8 */ struct net_device * dev; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct net_device * input_dev; /* 64 8 */ sk_buff_data_t transport_header; /* 72 4 */ sk_buff_data_t network_header; /* 76 4 */ sk_buff_data_t mac_header; /* 80 4 */ /* XXX 4 bytes hole, try to pack */ struct dst_entry * dst; /* 88 8 */ struct sec_path * sp; /* 96 8 */ char cb[48]; /* 104 48 */ /* cacheline 2 boundary (128 bytes) was 24 bytes ago*/ unsigned int len; /* 152 4 */ unsigned int data_len; /* 156 4 */ unsigned int mac_len; /* 160 4 */ union { __wsum csum; /* 4 */ __u32 csum_offset; /* 4 */ }; /* 164 4 */ __u32 priority; /* 168 4 */ __u8 local_df:1; /* 172 1 */ __u8 cloned:1; /* 172 1 */ __u8 ip_summed:2; /* 172 1 */ __u8 nohdr:1; /* 172 1 */ __u8 nfctinfo:3; /* 172 1 */ __u8 pkt_type:3; /* 173 1 */ __u8 fclone:2; /* 173 1 */ __u8 ipvs_property:1; /* 173 1 */ /* XXX 2 bits hole, try to pack */ __be16 protocol; /* 174 2 */ void (*destructor)(struct sk_buff *); /* 176 8 */ struct nf_conntrack * nfct; /* 184 8 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct sk_buff * nfct_reasm; /* 192 8 */ struct nf_bridge_info *nf_bridge; /* 200 8 */ __u16 tc_index; /* 208 2 */ __u16 tc_verd; /* 210 2 */ dma_cookie_t dma_cookie; /* 212 4 */ __u32 secmark; /* 216 4 */ __u32 mark; /* 220 4 */ unsigned int truesize; /* 224 4 */ atomic_t users; /* 228 4 */ unsigned char * head; /* 232 8 */ unsigned char * data; /* 240 8 */ unsigned char * tail; /* 248 8 */ /* --- cacheline 4 boundary (256 bytes) --- */ unsigned char * end; /* 256 8 */ }; /* size: 264, cachelines: 5 */ /* sum members: 260, holes: 1, sum holes: 4 */ /* bit holes: 1, sum bit holes: 2 bits */ /* last cacheline: 8 bytes */ On 32 bits nothing changes, and pointers continue to be used with the compiler turning all this abstraction layer into dust. But there are some sk_buff validation tricks that are now possible, humm... :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 11:22:35 +07:00
static inline int skb_network_offset(const struct sk_buff *skb)
{
return skb_network_header(skb) - skb->data;
}
static inline int skb_inner_network_offset(const struct sk_buff *skb)
{
return skb_inner_network_header(skb) - skb->data;
}
static inline int pskb_network_may_pull(struct sk_buff *skb, unsigned int len)
{
return pskb_may_pull(skb, skb_network_offset(skb) + len);
}
static inline void skb_pop_rcv_encapsulation(struct sk_buff *skb)
{
/* Only continue with checksum unnecessary if device indicated
* it is valid across encapsulation (skb->encapsulation was set).
*/
if (skb->ip_summed == CHECKSUM_UNNECESSARY && !skb->encapsulation)
skb->ip_summed = CHECKSUM_NONE;
skb->encapsulation = 0;
skb->csum_valid = 0;
}
/*
* CPUs often take a performance hit when accessing unaligned memory
* locations. The actual performance hit varies, it can be small if the
* hardware handles it or large if we have to take an exception and fix it
* in software.
*
* Since an ethernet header is 14 bytes network drivers often end up with
* the IP header at an unaligned offset. The IP header can be aligned by
* shifting the start of the packet by 2 bytes. Drivers should do this
* with:
*
* skb_reserve(skb, NET_IP_ALIGN);
*
* The downside to this alignment of the IP header is that the DMA is now
* unaligned. On some architectures the cost of an unaligned DMA is high
* and this cost outweighs the gains made by aligning the IP header.
*
* Since this trade off varies between architectures, we allow NET_IP_ALIGN
* to be overridden.
*/
#ifndef NET_IP_ALIGN
#define NET_IP_ALIGN 2
#endif
/*
* The networking layer reserves some headroom in skb data (via
* dev_alloc_skb). This is used to avoid having to reallocate skb data when
* the header has to grow. In the default case, if the header has to grow
* 32 bytes or less we avoid the reallocation.
*
* Unfortunately this headroom changes the DMA alignment of the resulting
* network packet. As for NET_IP_ALIGN, this unaligned DMA is expensive
* on some architectures. An architecture can override this value,
* perhaps setting it to a cacheline in size (since that will maintain
* cacheline alignment of the DMA). It must be a power of 2.
*
* Various parts of the networking layer expect at least 32 bytes of
* headroom, you should not reduce this.
*
* Using max(32, L1_CACHE_BYTES) makes sense (especially with RPS)
* to reduce average number of cache lines per packet.
* get_rps_cpus() for example only access one 64 bytes aligned block :
* NET_IP_ALIGN(2) + ethernet_header(14) + IP_header(20/40) + ports(8)
*/
#ifndef NET_SKB_PAD
#define NET_SKB_PAD max(32, L1_CACHE_BYTES)
#endif
int ___pskb_trim(struct sk_buff *skb, unsigned int len);
static inline void __skb_trim(struct sk_buff *skb, unsigned int len)
{
if (unlikely(skb_is_nonlinear(skb))) {
WARN_ON(1);
return;
}
skb->len = len;
skb_set_tail_pointer(skb, len);
}
void skb_trim(struct sk_buff *skb, unsigned int len);
static inline int __pskb_trim(struct sk_buff *skb, unsigned int len)
{
if (skb->data_len)
return ___pskb_trim(skb, len);
__skb_trim(skb, len);
return 0;
}
static inline int pskb_trim(struct sk_buff *skb, unsigned int len)
{
return (len < skb->len) ? __pskb_trim(skb, len) : 0;
}
/**
* pskb_trim_unique - remove end from a paged unique (not cloned) buffer
* @skb: buffer to alter
* @len: new length
*
* This is identical to pskb_trim except that the caller knows that
* the skb is not cloned so we should never get an error due to out-
* of-memory.
*/
static inline void pskb_trim_unique(struct sk_buff *skb, unsigned int len)
{
int err = pskb_trim(skb, len);
BUG_ON(err);
}
/**
* skb_orphan - orphan a buffer
* @skb: buffer to orphan
*
* If a buffer currently has an owner then we call the owner's
* destructor function and make the @skb unowned. The buffer continues
* to exist but is no longer charged to its former owner.
*/
static inline void skb_orphan(struct sk_buff *skb)
{
if (skb->destructor) {
skb->destructor(skb);
skb->destructor = NULL;
skb->sk = NULL;
} else {
BUG_ON(skb->sk);
}
}
/**
* skb_orphan_frags - orphan the frags contained in a buffer
* @skb: buffer to orphan frags from
* @gfp_mask: allocation mask for replacement pages
*
* For each frag in the SKB which needs a destructor (i.e. has an
* owner) create a copy of that frag and release the original
* page by calling the destructor.
*/
static inline int skb_orphan_frags(struct sk_buff *skb, gfp_t gfp_mask)
{
if (likely(!(skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY)))
return 0;
return skb_copy_ubufs(skb, gfp_mask);
}
/**
* __skb_queue_purge - empty a list
* @list: list to empty
*
* Delete all buffers on an &sk_buff list. Each buffer is removed from
* the list and one reference dropped. This function does not take the
* list lock and the caller must hold the relevant locks to use it.
*/
void skb_queue_purge(struct sk_buff_head *list);
static inline void __skb_queue_purge(struct sk_buff_head *list)
{
struct sk_buff *skb;
while ((skb = __skb_dequeue(list)) != NULL)
kfree_skb(skb);
}
#define NETDEV_FRAG_PAGE_MAX_ORDER get_order(32768)
#define NETDEV_FRAG_PAGE_MAX_SIZE (PAGE_SIZE << NETDEV_FRAG_PAGE_MAX_ORDER)
#define NETDEV_PAGECNT_MAX_BIAS NETDEV_FRAG_PAGE_MAX_SIZE
void *netdev_alloc_frag(unsigned int fragsz);
struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int length,
gfp_t gfp_mask);
/**
* netdev_alloc_skb - allocate an skbuff for rx on a specific device
* @dev: network device to receive on
* @length: length to allocate
*
* Allocate a new &sk_buff and assign it a usage count of one. The
* buffer has unspecified headroom built in. Users should allocate
* the headroom they think they need without accounting for the
* built in space. The built in space is used for optimisations.
*
* %NULL is returned if there is no free memory. Although this function
* allocates memory it can be called from an interrupt.
*/
static inline struct sk_buff *netdev_alloc_skb(struct net_device *dev,
unsigned int length)
{
return __netdev_alloc_skb(dev, length, GFP_ATOMIC);
}
/* legacy helper around __netdev_alloc_skb() */
static inline struct sk_buff *__dev_alloc_skb(unsigned int length,
gfp_t gfp_mask)
{
return __netdev_alloc_skb(NULL, length, gfp_mask);
}
/* legacy helper around netdev_alloc_skb() */
static inline struct sk_buff *dev_alloc_skb(unsigned int length)
{
return netdev_alloc_skb(NULL, length);
}
static inline struct sk_buff *__netdev_alloc_skb_ip_align(struct net_device *dev,
unsigned int length, gfp_t gfp)
{
struct sk_buff *skb = __netdev_alloc_skb(dev, length + NET_IP_ALIGN, gfp);
if (NET_IP_ALIGN && skb)
skb_reserve(skb, NET_IP_ALIGN);
return skb;
}
static inline struct sk_buff *netdev_alloc_skb_ip_align(struct net_device *dev,
unsigned int length)
{
return __netdev_alloc_skb_ip_align(dev, length, GFP_ATOMIC);
}
/**
* __skb_alloc_pages - allocate pages for ps-rx on a skb and preserve pfmemalloc data
2012-08-01 06:44:24 +07:00
* @gfp_mask: alloc_pages_node mask. Set __GFP_NOMEMALLOC if not for network packet RX
* @skb: skb to set pfmemalloc on if __GFP_MEMALLOC is used
* @order: size of the allocation
*
* Allocate a new page.
*
* %NULL is returned if there is no free memory.
*/
static inline struct page *__skb_alloc_pages(gfp_t gfp_mask,
struct sk_buff *skb,
unsigned int order)
{
struct page *page;
gfp_mask |= __GFP_COLD;
if (!(gfp_mask & __GFP_NOMEMALLOC))
gfp_mask |= __GFP_MEMALLOC;
page = alloc_pages_node(NUMA_NO_NODE, gfp_mask, order);
if (skb && page && page->pfmemalloc)
skb->pfmemalloc = true;
return page;
}
/**
* __skb_alloc_page - allocate a page for ps-rx for a given skb and preserve pfmemalloc data
* @gfp_mask: alloc_pages_node mask. Set __GFP_NOMEMALLOC if not for network packet RX
* @skb: skb to set pfmemalloc on if __GFP_MEMALLOC is used
*
* Allocate a new page.
*
* %NULL is returned if there is no free memory.
*/
static inline struct page *__skb_alloc_page(gfp_t gfp_mask,
struct sk_buff *skb)
{
return __skb_alloc_pages(gfp_mask, skb, 0);
}
/**
* skb_propagate_pfmemalloc - Propagate pfmemalloc if skb is allocated after RX page
* @page: The page that was allocated from skb_alloc_page
* @skb: The skb that may need pfmemalloc set
*/
static inline void skb_propagate_pfmemalloc(struct page *page,
struct sk_buff *skb)
{
if (page && page->pfmemalloc)
skb->pfmemalloc = true;
}
/**
* skb_frag_page - retrieve the page referred to by a paged fragment
* @frag: the paged fragment
*
* Returns the &struct page associated with @frag.
*/
static inline struct page *skb_frag_page(const skb_frag_t *frag)
{
return frag->page.p;
}
/**
* __skb_frag_ref - take an addition reference on a paged fragment.
* @frag: the paged fragment
*
* Takes an additional reference on the paged fragment @frag.
*/
static inline void __skb_frag_ref(skb_frag_t *frag)
{
get_page(skb_frag_page(frag));
}
/**
* skb_frag_ref - take an addition reference on a paged fragment of an skb.
* @skb: the buffer
* @f: the fragment offset.
*
* Takes an additional reference on the @f'th paged fragment of @skb.
*/
static inline void skb_frag_ref(struct sk_buff *skb, int f)
{
__skb_frag_ref(&skb_shinfo(skb)->frags[f]);
}
/**
* __skb_frag_unref - release a reference on a paged fragment.
* @frag: the paged fragment
*
* Releases a reference on the paged fragment @frag.
*/
static inline void __skb_frag_unref(skb_frag_t *frag)
{
put_page(skb_frag_page(frag));
}
/**
* skb_frag_unref - release a reference on a paged fragment of an skb.
* @skb: the buffer
* @f: the fragment offset
*
* Releases a reference on the @f'th paged fragment of @skb.
*/
static inline void skb_frag_unref(struct sk_buff *skb, int f)
{
__skb_frag_unref(&skb_shinfo(skb)->frags[f]);
}
/**
* skb_frag_address - gets the address of the data contained in a paged fragment
* @frag: the paged fragment buffer
*
* Returns the address of the data within @frag. The page must already
* be mapped.
*/
static inline void *skb_frag_address(const skb_frag_t *frag)
{
return page_address(skb_frag_page(frag)) + frag->page_offset;
}
/**
* skb_frag_address_safe - gets the address of the data contained in a paged fragment
* @frag: the paged fragment buffer
*
* Returns the address of the data within @frag. Checks that the page
* is mapped and returns %NULL otherwise.
*/
static inline void *skb_frag_address_safe(const skb_frag_t *frag)
{
void *ptr = page_address(skb_frag_page(frag));
if (unlikely(!ptr))
return NULL;
return ptr + frag->page_offset;
}
/**
* __skb_frag_set_page - sets the page contained in a paged fragment
* @frag: the paged fragment
* @page: the page to set
*
* Sets the fragment @frag to contain @page.
*/
static inline void __skb_frag_set_page(skb_frag_t *frag, struct page *page)
{
frag->page.p = page;
}
/**
* skb_frag_set_page - sets the page contained in a paged fragment of an skb
* @skb: the buffer
* @f: the fragment offset
* @page: the page to set
*
* Sets the @f'th fragment of @skb to contain @page.
*/
static inline void skb_frag_set_page(struct sk_buff *skb, int f,
struct page *page)
{
__skb_frag_set_page(&skb_shinfo(skb)->frags[f], page);
}
bool skb_page_frag_refill(unsigned int sz, struct page_frag *pfrag, gfp_t prio);
/**
* skb_frag_dma_map - maps a paged fragment via the DMA API
* @dev: the device to map the fragment to
* @frag: the paged fragment to map
* @offset: the offset within the fragment (starting at the
* fragment's own offset)
* @size: the number of bytes to map
* @dir: the direction of the mapping (%PCI_DMA_*)
*
* Maps the page associated with @frag to @device.
*/
static inline dma_addr_t skb_frag_dma_map(struct device *dev,
const skb_frag_t *frag,
size_t offset, size_t size,
enum dma_data_direction dir)
{
return dma_map_page(dev, skb_frag_page(frag),
frag->page_offset + offset, size, dir);
}
static inline struct sk_buff *pskb_copy(struct sk_buff *skb,
gfp_t gfp_mask)
{
return __pskb_copy(skb, skb_headroom(skb), gfp_mask);
}
static inline struct sk_buff *pskb_copy_for_clone(struct sk_buff *skb,
gfp_t gfp_mask)
{
return __pskb_copy_fclone(skb, skb_headroom(skb), gfp_mask, true);
}
[SKBUFF]: Keep track of writable header len of headerless clones Currently NAT (and others) that want to modify cloned skbs copy them, even if in the vast majority of cases its not necessary because the skb is a clone made by TCP and the portion NAT wants to modify is actually writable because TCP release the header reference before cloning. The problem is that there is no clean way for NAT to find out how long the writable header area is, so this patch introduces skb->hdr_len to hold this length. When a headerless skb is cloned skb->hdr_len is set to the current headroom, for regular clones it is copied from the original. A new function skb_clone_writable(skb, len) returns whether the skb is writable up to len bytes from skb->data. To avoid enlarging the skb the mac_len field is reduced to 16 bit and the new hdr_len field is put in the remaining 16 bit. I've done a few rough benchmarks of NAT (not with this exact patch, but a very similar one). As expected it saves huge amounts of system time in case of sendfile, bringing it down to basically the same amount as without NAT, with sendmsg it only helps on loopback, probably because of the large MTU. Transmit a 1GB file using sendfile/sendmsg over eth0/lo with and without NAT: - sendfile eth0, no NAT: sys 0m0.388s - sendfile eth0, NAT: sys 0m1.835s - sendfile eth0: NAT + path: sys 0m0.370s (~ -80%) - sendfile lo, no NAT: sys 0m0.258s - sendfile lo, NAT: sys 0m2.609s - sendfile lo, NAT + patch: sys 0m0.260s (~ -90%) - sendmsg eth0, no NAT: sys 0m2.508s - sendmsg eth0, NAT: sys 0m2.539s - sendmsg eth0, NAT + patch: sys 0m2.445s (no change) - sendmsg lo, no NAT: sys 0m2.151s - sendmsg lo, NAT: sys 0m3.557s - sendmsg lo, NAT + patch: sys 0m2.159s (~ -40%) I expect other users can see a similar performance improvement, packet mangling iptables targets, ipip and ip_gre come to mind .. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-06-25 18:35:20 +07:00
/**
* skb_clone_writable - is the header of a clone writable
* @skb: buffer to check
* @len: length up to which to write
*
* Returns true if modifying the header part of the cloned buffer
* does not requires the data to be copied.
*/
static inline int skb_clone_writable(const struct sk_buff *skb, unsigned int len)
[SKBUFF]: Keep track of writable header len of headerless clones Currently NAT (and others) that want to modify cloned skbs copy them, even if in the vast majority of cases its not necessary because the skb is a clone made by TCP and the portion NAT wants to modify is actually writable because TCP release the header reference before cloning. The problem is that there is no clean way for NAT to find out how long the writable header area is, so this patch introduces skb->hdr_len to hold this length. When a headerless skb is cloned skb->hdr_len is set to the current headroom, for regular clones it is copied from the original. A new function skb_clone_writable(skb, len) returns whether the skb is writable up to len bytes from skb->data. To avoid enlarging the skb the mac_len field is reduced to 16 bit and the new hdr_len field is put in the remaining 16 bit. I've done a few rough benchmarks of NAT (not with this exact patch, but a very similar one). As expected it saves huge amounts of system time in case of sendfile, bringing it down to basically the same amount as without NAT, with sendmsg it only helps on loopback, probably because of the large MTU. Transmit a 1GB file using sendfile/sendmsg over eth0/lo with and without NAT: - sendfile eth0, no NAT: sys 0m0.388s - sendfile eth0, NAT: sys 0m1.835s - sendfile eth0: NAT + path: sys 0m0.370s (~ -80%) - sendfile lo, no NAT: sys 0m0.258s - sendfile lo, NAT: sys 0m2.609s - sendfile lo, NAT + patch: sys 0m0.260s (~ -90%) - sendmsg eth0, no NAT: sys 0m2.508s - sendmsg eth0, NAT: sys 0m2.539s - sendmsg eth0, NAT + patch: sys 0m2.445s (no change) - sendmsg lo, no NAT: sys 0m2.151s - sendmsg lo, NAT: sys 0m3.557s - sendmsg lo, NAT + patch: sys 0m2.159s (~ -40%) I expect other users can see a similar performance improvement, packet mangling iptables targets, ipip and ip_gre come to mind .. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-06-25 18:35:20 +07:00
{
return !skb_header_cloned(skb) &&
skb_headroom(skb) + len <= skb->hdr_len;
}
static inline int __skb_cow(struct sk_buff *skb, unsigned int headroom,
int cloned)
{
int delta = 0;
if (headroom > skb_headroom(skb))
delta = headroom - skb_headroom(skb);
if (delta || cloned)
return pskb_expand_head(skb, ALIGN(delta, NET_SKB_PAD), 0,
GFP_ATOMIC);
return 0;
}
/**
* skb_cow - copy header of skb when it is required
* @skb: buffer to cow
* @headroom: needed headroom
*
* If the skb passed lacks sufficient headroom or its data part
* is shared, data is reallocated. If reallocation fails, an error
* is returned and original skb is not changed.
*
* The result is skb with writable area skb->head...skb->tail
* and at least @headroom of space at head.
*/
static inline int skb_cow(struct sk_buff *skb, unsigned int headroom)
{
return __skb_cow(skb, headroom, skb_cloned(skb));
}
/**
* skb_cow_head - skb_cow but only making the head writable
* @skb: buffer to cow
* @headroom: needed headroom
*
* This function is identical to skb_cow except that we replace the
* skb_cloned check by skb_header_cloned. It should be used when
* you only need to push on some header and do not need to modify
* the data.
*/
static inline int skb_cow_head(struct sk_buff *skb, unsigned int headroom)
{
return __skb_cow(skb, headroom, skb_header_cloned(skb));
}
/**
* skb_padto - pad an skbuff up to a minimal size
* @skb: buffer to pad
* @len: minimal length
*
* Pads up a buffer to ensure the trailing bytes exist and are
* blanked. If the buffer already contains sufficient data it
* is untouched. Otherwise it is extended. Returns zero on
* success. The skb is freed on error.
*/
static inline int skb_padto(struct sk_buff *skb, unsigned int len)
{
unsigned int size = skb->len;
if (likely(size >= len))
return 0;
return skb_pad(skb, len - size);
}
static inline int skb_add_data(struct sk_buff *skb,
char __user *from, int copy)
{
const int off = skb->len;
if (skb->ip_summed == CHECKSUM_NONE) {
int err = 0;
__wsum csum = csum_and_copy_from_user(from, skb_put(skb, copy),
copy, 0, &err);
if (!err) {
skb->csum = csum_block_add(skb->csum, csum, off);
return 0;
}
} else if (!copy_from_user(skb_put(skb, copy), from, copy))
return 0;
__skb_trim(skb, off);
return -EFAULT;
}
static inline bool skb_can_coalesce(struct sk_buff *skb, int i,
const struct page *page, int off)
{
if (i) {
const struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[i - 1];
return page == skb_frag_page(frag) &&
off == frag->page_offset + skb_frag_size(frag);
}
return false;
}
static inline int __skb_linearize(struct sk_buff *skb)
{
return __pskb_pull_tail(skb, skb->data_len) ? 0 : -ENOMEM;
}
/**
* skb_linearize - convert paged skb to linear one
* @skb: buffer to linarize
*
* If there is no free memory -ENOMEM is returned, otherwise zero
* is returned and the old skb data released.
*/
static inline int skb_linearize(struct sk_buff *skb)
{
return skb_is_nonlinear(skb) ? __skb_linearize(skb) : 0;
}
/**
* skb_has_shared_frag - can any frag be overwritten
* @skb: buffer to test
*
* Return true if the skb has at least one frag that might be modified
* by an external entity (as in vmsplice()/sendfile())
*/
static inline bool skb_has_shared_frag(const struct sk_buff *skb)
{
return skb_is_nonlinear(skb) &&
skb_shinfo(skb)->tx_flags & SKBTX_SHARED_FRAG;
}
/**
* skb_linearize_cow - make sure skb is linear and writable
* @skb: buffer to process
*
* If there is no free memory -ENOMEM is returned, otherwise zero
* is returned and the old skb data released.
*/
static inline int skb_linearize_cow(struct sk_buff *skb)
{
return skb_is_nonlinear(skb) || skb_cloned(skb) ?
__skb_linearize(skb) : 0;
}
/**
* skb_postpull_rcsum - update checksum for received skb after pull
* @skb: buffer to update
* @start: start of data before pull
* @len: length of data pulled
*
* After doing a pull on a received packet, you need to call this to
* update the CHECKSUM_COMPLETE checksum, or set ip_summed to
* CHECKSUM_NONE so that it can be recomputed from scratch.
*/
static inline void skb_postpull_rcsum(struct sk_buff *skb,
const void *start, unsigned int len)
{
if (skb->ip_summed == CHECKSUM_COMPLETE)
skb->csum = csum_sub(skb->csum, csum_partial(start, len, 0));
}
unsigned char *skb_pull_rcsum(struct sk_buff *skb, unsigned int len);
/**
* pskb_trim_rcsum - trim received skb and update checksum
* @skb: buffer to trim
* @len: new length
*
* This is exactly the same as pskb_trim except that it ensures the
* checksum of received packets are still valid after the operation.
*/
static inline int pskb_trim_rcsum(struct sk_buff *skb, unsigned int len)
{
if (likely(len >= skb->len))
return 0;
if (skb->ip_summed == CHECKSUM_COMPLETE)
skb->ip_summed = CHECKSUM_NONE;
return __pskb_trim(skb, len);
}
#define skb_queue_walk(queue, skb) \
for (skb = (queue)->next; \
skb != (struct sk_buff *)(queue); \
skb = skb->next)
#define skb_queue_walk_safe(queue, skb, tmp) \
for (skb = (queue)->next, tmp = skb->next; \
skb != (struct sk_buff *)(queue); \
skb = tmp, tmp = skb->next)
#define skb_queue_walk_from(queue, skb) \
for (; skb != (struct sk_buff *)(queue); \
skb = skb->next)
#define skb_queue_walk_from_safe(queue, skb, tmp) \
for (tmp = skb->next; \
skb != (struct sk_buff *)(queue); \
skb = tmp, tmp = skb->next)
#define skb_queue_reverse_walk(queue, skb) \
for (skb = (queue)->prev; \
skb != (struct sk_buff *)(queue); \
skb = skb->prev)
#define skb_queue_reverse_walk_safe(queue, skb, tmp) \
for (skb = (queue)->prev, tmp = skb->prev; \
skb != (struct sk_buff *)(queue); \
skb = tmp, tmp = skb->prev)
#define skb_queue_reverse_walk_from_safe(queue, skb, tmp) \
for (tmp = skb->prev; \
skb != (struct sk_buff *)(queue); \
skb = tmp, tmp = skb->prev)
static inline bool skb_has_frag_list(const struct sk_buff *skb)
{
return skb_shinfo(skb)->frag_list != NULL;
}
static inline void skb_frag_list_init(struct sk_buff *skb)
{
skb_shinfo(skb)->frag_list = NULL;
}
static inline void skb_frag_add_head(struct sk_buff *skb, struct sk_buff *frag)
{
frag->next = skb_shinfo(skb)->frag_list;
skb_shinfo(skb)->frag_list = frag;
}
#define skb_walk_frags(skb, iter) \
for (iter = skb_shinfo(skb)->frag_list; iter; iter = iter->next)
struct sk_buff *__skb_recv_datagram(struct sock *sk, unsigned flags,
int *peeked, int *off, int *err);
struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
int *err);
unsigned int datagram_poll(struct file *file, struct socket *sock,
struct poll_table_struct *wait);
int skb_copy_datagram_iovec(const struct sk_buff *from, int offset,
struct iovec *to, int size);
int skb_copy_and_csum_datagram_iovec(struct sk_buff *skb, int hlen,
struct iovec *iov);
int skb_copy_datagram_from_iovec(struct sk_buff *skb, int offset,
const struct iovec *from, int from_offset,
int len);
int zerocopy_sg_from_iovec(struct sk_buff *skb, const struct iovec *frm,
int offset, size_t count);
int skb_copy_datagram_const_iovec(const struct sk_buff *from, int offset,
const struct iovec *to, int to_offset,
int size);
void skb_free_datagram(struct sock *sk, struct sk_buff *skb);
void skb_free_datagram_locked(struct sock *sk, struct sk_buff *skb);
int skb_kill_datagram(struct sock *sk, struct sk_buff *skb, unsigned int flags);
int skb_copy_bits(const struct sk_buff *skb, int offset, void *to, int len);
int skb_store_bits(struct sk_buff *skb, int offset, const void *from, int len);
__wsum skb_copy_and_csum_bits(const struct sk_buff *skb, int offset, u8 *to,
int len, __wsum csum);
int skb_splice_bits(struct sk_buff *skb, unsigned int offset,
struct pipe_inode_info *pipe, unsigned int len,
unsigned int flags);
void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to);
unsigned int skb_zerocopy_headlen(const struct sk_buff *from);
int skb_zerocopy(struct sk_buff *to, struct sk_buff *from,
int len, int hlen);
void skb_split(struct sk_buff *skb, struct sk_buff *skb1, const u32 len);
int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen);
void skb_scrub_packet(struct sk_buff *skb, bool xnet);
unsigned int skb_gso_transport_seglen(const struct sk_buff *skb);
struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features);
struct skb_checksum_ops {
__wsum (*update)(const void *mem, int len, __wsum wsum);
__wsum (*combine)(__wsum csum, __wsum csum2, int offset, int len);
};
__wsum __skb_checksum(const struct sk_buff *skb, int offset, int len,
__wsum csum, const struct skb_checksum_ops *ops);
__wsum skb_checksum(const struct sk_buff *skb, int offset, int len,
__wsum csum);
static inline void *skb_header_pointer(const struct sk_buff *skb, int offset,
int len, void *buffer)
{
int hlen = skb_headlen(skb);
if (hlen - offset >= len)
return skb->data + offset;
if (skb_copy_bits(skb, offset, buffer, len) < 0)
return NULL;
return buffer;
}
/**
* skb_needs_linearize - check if we need to linearize a given skb
* depending on the given device features.
* @skb: socket buffer to check
* @features: net device features
*
* Returns true if either:
* 1. skb has frag_list and the device doesn't support FRAGLIST, or
* 2. skb is fragmented and the device does not support SG.
*/
static inline bool skb_needs_linearize(struct sk_buff *skb,
netdev_features_t features)
{
return skb_is_nonlinear(skb) &&
((skb_has_frag_list(skb) && !(features & NETIF_F_FRAGLIST)) ||
(skb_shinfo(skb)->nr_frags && !(features & NETIF_F_SG)));
}
static inline void skb_copy_from_linear_data(const struct sk_buff *skb,
void *to,
const unsigned int len)
{
memcpy(to, skb->data, len);
}
static inline void skb_copy_from_linear_data_offset(const struct sk_buff *skb,
const int offset, void *to,
const unsigned int len)
{
memcpy(to, skb->data + offset, len);
}
static inline void skb_copy_to_linear_data(struct sk_buff *skb,
const void *from,
const unsigned int len)
{
memcpy(skb->data, from, len);
}
static inline void skb_copy_to_linear_data_offset(struct sk_buff *skb,
const int offset,
const void *from,
const unsigned int len)
{
memcpy(skb->data + offset, from, len);
}
void skb_init(void);
static inline ktime_t skb_get_ktime(const struct sk_buff *skb)
{
return skb->tstamp;
}
/**
* skb_get_timestamp - get timestamp from a skb
* @skb: skb to get stamp from
* @stamp: pointer to struct timeval to store stamp in
*
* Timestamps are stored in the skb as offsets to a base timestamp.
* This function converts the offset back to a struct timeval and stores
* it in stamp.
*/
static inline void skb_get_timestamp(const struct sk_buff *skb,
struct timeval *stamp)
{
*stamp = ktime_to_timeval(skb->tstamp);
}
static inline void skb_get_timestampns(const struct sk_buff *skb,
struct timespec *stamp)
{
*stamp = ktime_to_timespec(skb->tstamp);
}
static inline void __net_timestamp(struct sk_buff *skb)
{
skb->tstamp = ktime_get_real();
}
static inline ktime_t net_timedelta(ktime_t t)
{
return ktime_sub(ktime_get_real(), t);
}
static inline ktime_t net_invalid_timestamp(void)
{
return ktime_set(0, 0);
}
#ifdef CONFIG_NETWORK_PHY_TIMESTAMPING
void skb_clone_tx_timestamp(struct sk_buff *skb);
bool skb_defer_rx_timestamp(struct sk_buff *skb);
#else /* CONFIG_NETWORK_PHY_TIMESTAMPING */
static inline void skb_clone_tx_timestamp(struct sk_buff *skb)
{
}
static inline bool skb_defer_rx_timestamp(struct sk_buff *skb)
{
return false;
}
#endif /* !CONFIG_NETWORK_PHY_TIMESTAMPING */
/**
* skb_complete_tx_timestamp() - deliver cloned skb with tx timestamps
*
* PHY drivers may accept clones of transmitted packets for
* timestamping via their phy_driver.txtstamp method. These drivers
* must call this function to return the skb back to the stack, with
* or without a timestamp.
*
* @skb: clone of the the original outgoing packet
* @hwtstamps: hardware time stamps, may be NULL if not available
*
*/
void skb_complete_tx_timestamp(struct sk_buff *skb,
struct skb_shared_hwtstamps *hwtstamps);
/**
* skb_tstamp_tx - queue clone of skb with send time stamps
* @orig_skb: the original outgoing packet
* @hwtstamps: hardware time stamps, may be NULL if not available
*
* If the skb has a socket associated, then this function clones the
* skb (thus sharing the actual data and optional structures), stores
* the optional hardware time stamping information (if non NULL) or
* generates a software time stamp (otherwise), then queues the clone
* to the error queue of the socket. Errors are silently ignored.
*/
void skb_tstamp_tx(struct sk_buff *orig_skb,
struct skb_shared_hwtstamps *hwtstamps);
static inline void sw_tx_timestamp(struct sk_buff *skb)
{
if (skb_shinfo(skb)->tx_flags & SKBTX_SW_TSTAMP &&
!(skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS))
skb_tstamp_tx(skb, NULL);
}
/**
* skb_tx_timestamp() - Driver hook for transmit timestamping
*
* Ethernet MAC Drivers should call this function in their hard_xmit()
* function immediately before giving the sk_buff to the MAC hardware.
*
* Specifically, one should make absolutely sure that this function is
* called before TX completion of this packet can trigger. Otherwise
* the packet could potentially already be freed.
*
* @skb: A socket buffer.
*/
static inline void skb_tx_timestamp(struct sk_buff *skb)
{
skb_clone_tx_timestamp(skb);
sw_tx_timestamp(skb);
}
/**
* skb_complete_wifi_ack - deliver skb with wifi status
*
* @skb: the original outgoing packet
* @acked: ack status
*
*/
void skb_complete_wifi_ack(struct sk_buff *skb, bool acked);
__sum16 __skb_checksum_complete_head(struct sk_buff *skb, int len);
__sum16 __skb_checksum_complete(struct sk_buff *skb);
static inline int skb_csum_unnecessary(const struct sk_buff *skb)
{
return ((skb->ip_summed & CHECKSUM_UNNECESSARY) || skb->csum_valid);
}
/**
* skb_checksum_complete - Calculate checksum of an entire packet
* @skb: packet to process
*
* This function calculates the checksum over the entire packet plus
* the value of skb->csum. The latter can be used to supply the
* checksum of a pseudo header as used by TCP/UDP. It returns the
* checksum.
*
* For protocols that contain complete checksums such as ICMP/TCP/UDP,
* this function can be used to verify that checksum on received
* packets. In that case the function should return zero if the
* checksum is correct. In particular, this function will return zero
* if skb->ip_summed is CHECKSUM_UNNECESSARY which indicates that the
* hardware has already verified the correctness of the checksum.
*/
static inline __sum16 skb_checksum_complete(struct sk_buff *skb)
{
return skb_csum_unnecessary(skb) ?
0 : __skb_checksum_complete(skb);
}
/* Check if we need to perform checksum complete validation.
*
* Returns true if checksum complete is needed, false otherwise
* (either checksum is unnecessary or zero checksum is allowed).
*/
static inline bool __skb_checksum_validate_needed(struct sk_buff *skb,
bool zero_okay,
__sum16 check)
{
if (skb_csum_unnecessary(skb) || (zero_okay && !check)) {
skb->csum_valid = 1;
return false;
}
return true;
}
/* For small packets <= CHECKSUM_BREAK peform checksum complete directly
* in checksum_init.
*/
#define CHECKSUM_BREAK 76
/* Validate (init) checksum based on checksum complete.
*
* Return values:
* 0: checksum is validated or try to in skb_checksum_complete. In the latter
* case the ip_summed will not be CHECKSUM_UNNECESSARY and the pseudo
* checksum is stored in skb->csum for use in __skb_checksum_complete
* non-zero: value of invalid checksum
*
*/
static inline __sum16 __skb_checksum_validate_complete(struct sk_buff *skb,
bool complete,
__wsum psum)
{
if (skb->ip_summed == CHECKSUM_COMPLETE) {
if (!csum_fold(csum_add(psum, skb->csum))) {
skb->csum_valid = 1;
return 0;
}
}
skb->csum = psum;
if (complete || skb->len <= CHECKSUM_BREAK) {
__sum16 csum;
csum = __skb_checksum_complete(skb);
skb->csum_valid = !csum;
return csum;
}
return 0;
}
static inline __wsum null_compute_pseudo(struct sk_buff *skb, int proto)
{
return 0;
}
/* Perform checksum validate (init). Note that this is a macro since we only
* want to calculate the pseudo header which is an input function if necessary.
* First we try to validate without any computation (checksum unnecessary) and
* then calculate based on checksum complete calling the function to compute
* pseudo header.
*
* Return values:
* 0: checksum is validated or try to in skb_checksum_complete
* non-zero: value of invalid checksum
*/
#define __skb_checksum_validate(skb, proto, complete, \
zero_okay, check, compute_pseudo) \
({ \
__sum16 __ret = 0; \
skb->csum_valid = 0; \
if (__skb_checksum_validate_needed(skb, zero_okay, check)) \
__ret = __skb_checksum_validate_complete(skb, \
complete, compute_pseudo(skb, proto)); \
__ret; \
})
#define skb_checksum_init(skb, proto, compute_pseudo) \
__skb_checksum_validate(skb, proto, false, false, 0, compute_pseudo)
#define skb_checksum_init_zero_check(skb, proto, check, compute_pseudo) \
__skb_checksum_validate(skb, proto, false, true, check, compute_pseudo)
#define skb_checksum_validate(skb, proto, compute_pseudo) \
__skb_checksum_validate(skb, proto, true, false, 0, compute_pseudo)
#define skb_checksum_validate_zero_check(skb, proto, check, \
compute_pseudo) \
__skb_checksum_validate_(skb, proto, true, true, check, compute_pseudo)
#define skb_checksum_simple_validate(skb) \
__skb_checksum_validate(skb, 0, true, false, 0, null_compute_pseudo)
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
void nf_conntrack_destroy(struct nf_conntrack *nfct);
static inline void nf_conntrack_put(struct nf_conntrack *nfct)
{
if (nfct && atomic_dec_and_test(&nfct->use))
nf_conntrack_destroy(nfct);
}
static inline void nf_conntrack_get(struct nf_conntrack *nfct)
{
if (nfct)
atomic_inc(&nfct->use);
}
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
static inline void nf_bridge_put(struct nf_bridge_info *nf_bridge)
{
if (nf_bridge && atomic_dec_and_test(&nf_bridge->use))
kfree(nf_bridge);
}
static inline void nf_bridge_get(struct nf_bridge_info *nf_bridge)
{
if (nf_bridge)
atomic_inc(&nf_bridge->use);
}
#endif /* CONFIG_BRIDGE_NETFILTER */
static inline void nf_reset(struct sk_buff *skb)
{
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
nf_conntrack_put(skb->nfct);
skb->nfct = NULL;
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
nf_bridge_put(skb->nf_bridge);
skb->nf_bridge = NULL;
#endif
}
static inline void nf_reset_trace(struct sk_buff *skb)
{
#if IS_ENABLED(CONFIG_NETFILTER_XT_TARGET_TRACE) || defined(CONFIG_NF_TABLES)
skb->nf_trace = 0;
#endif
}
/* Note: This doesn't put any conntrack and bridge info in dst. */
static inline void __nf_copy(struct sk_buff *dst, const struct sk_buff *src)
{
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
dst->nfct = src->nfct;
nf_conntrack_get(src->nfct);
dst->nfctinfo = src->nfctinfo;
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
dst->nf_bridge = src->nf_bridge;
nf_bridge_get(src->nf_bridge);
#endif
#if IS_ENABLED(CONFIG_NETFILTER_XT_TARGET_TRACE) || defined(CONFIG_NF_TABLES)
dst->nf_trace = src->nf_trace;
#endif
}
static inline void nf_copy(struct sk_buff *dst, const struct sk_buff *src)
{
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
nf_conntrack_put(dst->nfct);
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
nf_bridge_put(dst->nf_bridge);
#endif
__nf_copy(dst, src);
}
#ifdef CONFIG_NETWORK_SECMARK
static inline void skb_copy_secmark(struct sk_buff *to, const struct sk_buff *from)
{
to->secmark = from->secmark;
}
static inline void skb_init_secmark(struct sk_buff *skb)
{
skb->secmark = 0;
}
#else
static inline void skb_copy_secmark(struct sk_buff *to, const struct sk_buff *from)
{ }
static inline void skb_init_secmark(struct sk_buff *skb)
{ }
#endif
static inline bool skb_irq_freeable(const struct sk_buff *skb)
{
return !skb->destructor &&
#if IS_ENABLED(CONFIG_XFRM)
!skb->sp &&
#endif
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
!skb->nfct &&
#endif
!skb->_skb_refdst &&
!skb_has_frag_list(skb);
}
static inline void skb_set_queue_mapping(struct sk_buff *skb, u16 queue_mapping)
{
skb->queue_mapping = queue_mapping;
}
static inline u16 skb_get_queue_mapping(const struct sk_buff *skb)
{
return skb->queue_mapping;
}
static inline void skb_copy_queue_mapping(struct sk_buff *to, const struct sk_buff *from)
{
to->queue_mapping = from->queue_mapping;
}
static inline void skb_record_rx_queue(struct sk_buff *skb, u16 rx_queue)
{
skb->queue_mapping = rx_queue + 1;
}
static inline u16 skb_get_rx_queue(const struct sk_buff *skb)
{
return skb->queue_mapping - 1;
}
static inline bool skb_rx_queue_recorded(const struct sk_buff *skb)
{
return skb->queue_mapping != 0;
}
u16 __skb_tx_hash(const struct net_device *dev, const struct sk_buff *skb,
unsigned int num_tx_queues);
static inline struct sec_path *skb_sec_path(struct sk_buff *skb)
{
#ifdef CONFIG_XFRM
return skb->sp;
#else
return NULL;
#endif
}
/* Keeps track of mac header offset relative to skb->head.
* It is useful for TSO of Tunneling protocol. e.g. GRE.
* For non-tunnel skb it points to skb_mac_header() and for
* tunnel skb it points to outer mac header.
* Keeps track of level of encapsulation of network headers.
*/
struct skb_gso_cb {
int mac_offset;
int encap_level;
__u16 csum_start;
};
#define SKB_GSO_CB(skb) ((struct skb_gso_cb *)(skb)->cb)
static inline int skb_tnl_header_len(const struct sk_buff *inner_skb)
{
return (skb_mac_header(inner_skb) - inner_skb->head) -
SKB_GSO_CB(inner_skb)->mac_offset;
}
static inline int gso_pskb_expand_head(struct sk_buff *skb, int extra)
{
int new_headroom, headroom;
int ret;
headroom = skb_headroom(skb);
ret = pskb_expand_head(skb, extra, 0, GFP_ATOMIC);
if (ret)
return ret;
new_headroom = skb_headroom(skb);
SKB_GSO_CB(skb)->mac_offset += (new_headroom - headroom);
return 0;
}
/* Compute the checksum for a gso segment. First compute the checksum value
* from the start of transport header to SKB_GSO_CB(skb)->csum_start, and
* then add in skb->csum (checksum from csum_start to end of packet).
* skb->csum and csum_start are then updated to reflect the checksum of the
* resultant packet starting from the transport header-- the resultant checksum
* is in the res argument (i.e. normally zero or ~ of checksum of a pseudo
* header.
*/
static inline __sum16 gso_make_checksum(struct sk_buff *skb, __wsum res)
{
int plen = SKB_GSO_CB(skb)->csum_start - skb_headroom(skb) -
skb_transport_offset(skb);
__u16 csum;
csum = csum_fold(csum_partial(skb_transport_header(skb),
plen, skb->csum));
skb->csum = res;
SKB_GSO_CB(skb)->csum_start -= plen;
return csum;
}
static inline bool skb_is_gso(const struct sk_buff *skb)
{
return skb_shinfo(skb)->gso_size;
}
/* Note: Should be called only if skb_is_gso(skb) is true */
static inline bool skb_is_gso_v6(const struct sk_buff *skb)
{
return skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6;
}
void __skb_warn_lro_forwarding(const struct sk_buff *skb);
static inline bool skb_warn_if_lro(const struct sk_buff *skb)
{
/* LRO sets gso_size but not gso_type, whereas if GSO is really
* wanted then gso_type will be set. */
const struct skb_shared_info *shinfo = skb_shinfo(skb);
if (skb_is_nonlinear(skb) && shinfo->gso_size != 0 &&
unlikely(shinfo->gso_type == 0)) {
__skb_warn_lro_forwarding(skb);
return true;
}
return false;
}
static inline void skb_forward_csum(struct sk_buff *skb)
{
/* Unfortunately we don't support this one. Any brave souls? */
if (skb->ip_summed == CHECKSUM_COMPLETE)
skb->ip_summed = CHECKSUM_NONE;
}
/**
* skb_checksum_none_assert - make sure skb ip_summed is CHECKSUM_NONE
* @skb: skb to check
*
* fresh skbs have their ip_summed set to CHECKSUM_NONE.
* Instead of forcing ip_summed to CHECKSUM_NONE, we can
* use this helper, to document places where we make this assertion.
*/
static inline void skb_checksum_none_assert(const struct sk_buff *skb)
{
#ifdef DEBUG
BUG_ON(skb->ip_summed != CHECKSUM_NONE);
#endif
}
bool skb_partial_csum_set(struct sk_buff *skb, u16 start, u16 off);
int skb_checksum_setup(struct sk_buff *skb, bool recalculate);
u32 __skb_get_poff(const struct sk_buff *skb);
/**
* skb_head_is_locked - Determine if the skb->head is locked down
* @skb: skb to check
*
* The head on skbs build around a head frag can be removed if they are
* not cloned. This function returns true if the skb head is locked down
* due to either being allocated via kmalloc, or by being a clone with
* multiple references to the head.
*/
static inline bool skb_head_is_locked(const struct sk_buff *skb)
{
return !skb->head_frag || skb_cloned(skb);
}
net: ip, ipv6: handle gso skbs in forwarding path Marcelo Ricardo Leitner reported problems when the forwarding link path has a lower mtu than the incoming one if the inbound interface supports GRO. Given: Host <mtu1500> R1 <mtu1200> R2 Host sends tcp stream which is routed via R1 and R2. R1 performs GRO. In this case, the kernel will fail to send ICMP fragmentation needed messages (or pkt too big for ipv6), as GSO packets currently bypass dstmtu checks in forward path. Instead, Linux tries to send out packets exceeding the mtu. When locking route MTU on Host (i.e., no ipv4 DF bit set), R1 does not fragment the packets when forwarding, and again tries to send out packets exceeding R1-R2 link mtu. This alters the forwarding dstmtu checks to take the individual gso segment lengths into account. For ipv6, we send out pkt too big error for gso if the individual segments are too big. For ipv4, we either send icmp fragmentation needed, or, if the DF bit is not set, perform software segmentation and let the output path create fragments when the packet is leaving the machine. It is not 100% correct as the error message will contain the headers of the GRO skb instead of the original/segmented one, but it seems to work fine in my (limited) tests. Eric Dumazet suggested to simply shrink mss via ->gso_size to avoid sofware segmentation. However it turns out that skb_segment() assumes skb nr_frags is related to mss size so we would BUG there. I don't want to mess with it considering Herbert and Eric disagree on what the correct behavior should be. Hannes Frederic Sowa notes that when we would shrink gso_size skb_segment would then also need to deal with the case where SKB_MAX_FRAGS would be exceeded. This uses sofware segmentation in the forward path when we hit ipv4 non-DF packets and the outgoing link mtu is too small. Its not perfect, but given the lack of bug reports wrt. GRO fwd being broken this is a rare case anyway. Also its not like this could not be improved later once the dust settles. Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Reported-by: Marcelo Ricardo Leitner <mleitner@redhat.com> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-14 05:09:12 +07:00
/**
* skb_gso_network_seglen - Return length of individual segments of a gso packet
*
* @skb: GSO skb
*
* skb_gso_network_seglen is used to determine the real size of the
* individual segments, including Layer3 (IP, IPv6) and L4 headers (TCP/UDP).
*
* The MAC/L2 header is not accounted for.
*/
static inline unsigned int skb_gso_network_seglen(const struct sk_buff *skb)
{
unsigned int hdr_len = skb_transport_header(skb) -
skb_network_header(skb);
return hdr_len + skb_gso_transport_seglen(skb);
}
#endif /* __KERNEL__ */
#endif /* _LINUX_SKBUFF_H */