[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
/*
|
|
|
|
* NET Generic infrastructure for Network protocols.
|
|
|
|
*
|
|
|
|
* Definitions for request_sock
|
|
|
|
*
|
|
|
|
* Authors: Arnaldo Carvalho de Melo <acme@conectiva.com.br>
|
|
|
|
*
|
|
|
|
* From code originally in include/net/tcp.h
|
|
|
|
*
|
|
|
|
* 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 _REQUEST_SOCK_H
|
|
|
|
#define _REQUEST_SOCK_H
|
|
|
|
|
|
|
|
#include <linux/slab.h>
|
2005-06-19 13:47:59 +08:00
|
|
|
#include <linux/spinlock.h>
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
#include <linux/types.h>
|
2005-06-19 13:47:59 +08:00
|
|
|
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
#include <net/sock.h>
|
|
|
|
|
2005-06-19 13:47:21 +08:00
|
|
|
struct request_sock;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
struct sk_buff;
|
|
|
|
struct dst_entry;
|
|
|
|
struct proto;
|
|
|
|
|
2005-06-19 13:47:21 +08:00
|
|
|
struct request_sock_ops {
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
int family;
|
|
|
|
int obj_size;
|
2006-12-07 12:33:20 +08:00
|
|
|
struct kmem_cache *slab;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
int (*rtx_syn_ack)(struct sock *sk,
|
2008-03-01 03:43:03 +08:00
|
|
|
struct request_sock *req);
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
void (*send_ack)(struct sk_buff *skb,
|
2005-06-19 13:47:21 +08:00
|
|
|
struct request_sock *req);
|
2006-11-15 11:07:45 +08:00
|
|
|
void (*send_reset)(struct sock *sk,
|
|
|
|
struct sk_buff *skb);
|
2005-06-19 13:47:21 +08:00
|
|
|
void (*destructor)(struct request_sock *req);
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
};
|
|
|
|
|
2005-06-19 13:47:21 +08:00
|
|
|
/* struct request_sock - mini sock to represent a connection request
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
*/
|
2005-06-19 13:47:21 +08:00
|
|
|
struct request_sock {
|
|
|
|
struct request_sock *dl_next; /* Must be first member! */
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
u16 mss;
|
|
|
|
u8 retrans;
|
2008-04-10 18:12:40 +08:00
|
|
|
u8 cookie_ts; /* syncookie: encode tcpopts in timestamp */
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
/* The following two fields can be easily recomputed I think -AK */
|
|
|
|
u32 window_clamp; /* window clamp at creation time */
|
|
|
|
u32 rcv_wnd; /* rcv_wnd offered first time */
|
|
|
|
u32 ts_recent;
|
|
|
|
unsigned long expires;
|
2006-11-16 18:30:37 +08:00
|
|
|
const struct request_sock_ops *rsk_ops;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
struct sock *sk;
|
2006-07-25 14:32:50 +08:00
|
|
|
u32 secid;
|
2006-11-09 07:04:09 +08:00
|
|
|
u32 peer_secid;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
};
|
|
|
|
|
2006-11-16 18:30:37 +08:00
|
|
|
static inline struct request_sock *reqsk_alloc(const struct request_sock_ops *ops)
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
{
|
2006-12-07 12:33:16 +08:00
|
|
|
struct request_sock *req = kmem_cache_alloc(ops->slab, GFP_ATOMIC);
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
|
|
|
|
if (req != NULL)
|
2005-06-19 13:47:21 +08:00
|
|
|
req->rsk_ops = ops;
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
|
|
|
|
return req;
|
|
|
|
}
|
|
|
|
|
2005-06-19 13:47:21 +08:00
|
|
|
static inline void __reqsk_free(struct request_sock *req)
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
{
|
2005-06-19 13:47:21 +08:00
|
|
|
kmem_cache_free(req->rsk_ops->slab, req);
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
}
|
|
|
|
|
2005-06-19 13:47:21 +08:00
|
|
|
static inline void reqsk_free(struct request_sock *req)
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
{
|
2005-06-19 13:47:21 +08:00
|
|
|
req->rsk_ops->destructor(req);
|
|
|
|
__reqsk_free(req);
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
}
|
|
|
|
|
2005-06-19 13:47:59 +08:00
|
|
|
extern int sysctl_max_syn_backlog;
|
|
|
|
|
2005-06-19 13:48:55 +08:00
|
|
|
/** struct listen_sock - listen state
|
2005-06-19 13:47:59 +08:00
|
|
|
*
|
|
|
|
* @max_qlen_log - log_2 of maximal queued SYNs/REQUESTs
|
|
|
|
*/
|
2005-06-19 13:48:55 +08:00
|
|
|
struct listen_sock {
|
2005-06-19 13:47:59 +08:00
|
|
|
u8 max_qlen_log;
|
|
|
|
/* 3 bytes hole, try to use */
|
|
|
|
int qlen;
|
|
|
|
int qlen_young;
|
|
|
|
int clock_hand;
|
|
|
|
u32 hash_rnd;
|
2005-08-10 10:33:31 +08:00
|
|
|
u32 nr_table_entries;
|
2005-06-19 13:47:59 +08:00
|
|
|
struct request_sock *syn_table[0];
|
|
|
|
};
|
|
|
|
|
|
|
|
/** struct request_sock_queue - queue of request_socks
|
|
|
|
*
|
|
|
|
* @rskq_accept_head - FIFO head of established children
|
|
|
|
* @rskq_accept_tail - FIFO tail of established children
|
2005-08-10 11:11:56 +08:00
|
|
|
* @rskq_defer_accept - User waits for some data after accept()
|
2005-06-19 13:47:59 +08:00
|
|
|
* @syn_wait_lock - serializer
|
|
|
|
*
|
|
|
|
* %syn_wait_lock is necessary only to avoid proc interface having to grab the main
|
|
|
|
* lock sock while browsing the listening hash (otherwise it's deadlock prone).
|
|
|
|
*
|
|
|
|
* This lock is acquired in read mode only from listening_get_next() seq_file
|
|
|
|
* op and it's acquired in write mode _only_ from code that is actively
|
|
|
|
* changing rskq_accept_head. All readers that are holding the master sock lock
|
|
|
|
* don't need to grab this lock in read mode too as rskq_accept_head. writes
|
|
|
|
* are always protected from the main sock lock.
|
|
|
|
*/
|
|
|
|
struct request_sock_queue {
|
|
|
|
struct request_sock *rskq_accept_head;
|
|
|
|
struct request_sock *rskq_accept_tail;
|
|
|
|
rwlock_t syn_wait_lock;
|
2008-03-22 07:33:01 +08:00
|
|
|
u16 rskq_defer_accept;
|
|
|
|
/* 2 bytes hole, try to pack */
|
2005-06-19 13:48:55 +08:00
|
|
|
struct listen_sock *listen_opt;
|
2005-06-19 13:47:59 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
extern int reqsk_queue_alloc(struct request_sock_queue *queue,
|
2006-11-16 18:30:37 +08:00
|
|
|
unsigned int nr_table_entries);
|
2005-06-19 13:47:59 +08:00
|
|
|
|
2007-11-15 18:57:06 +08:00
|
|
|
extern void __reqsk_queue_destroy(struct request_sock_queue *queue);
|
2005-08-10 10:33:31 +08:00
|
|
|
extern void reqsk_queue_destroy(struct request_sock_queue *queue);
|
|
|
|
|
2005-06-19 13:47:59 +08:00
|
|
|
static inline struct request_sock *
|
|
|
|
reqsk_queue_yank_acceptq(struct request_sock_queue *queue)
|
|
|
|
{
|
|
|
|
struct request_sock *req = queue->rskq_accept_head;
|
|
|
|
|
2006-03-27 09:39:55 +08:00
|
|
|
queue->rskq_accept_head = NULL;
|
2005-06-19 13:47:59 +08:00
|
|
|
return req;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int reqsk_queue_empty(struct request_sock_queue *queue)
|
|
|
|
{
|
|
|
|
return queue->rskq_accept_head == NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void reqsk_queue_unlink(struct request_sock_queue *queue,
|
|
|
|
struct request_sock *req,
|
|
|
|
struct request_sock **prev_req)
|
|
|
|
{
|
|
|
|
write_lock(&queue->syn_wait_lock);
|
|
|
|
*prev_req = req->dl_next;
|
|
|
|
write_unlock(&queue->syn_wait_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void reqsk_queue_add(struct request_sock_queue *queue,
|
|
|
|
struct request_sock *req,
|
|
|
|
struct sock *parent,
|
|
|
|
struct sock *child)
|
|
|
|
{
|
|
|
|
req->sk = child;
|
|
|
|
sk_acceptq_added(parent);
|
|
|
|
|
|
|
|
if (queue->rskq_accept_head == NULL)
|
|
|
|
queue->rskq_accept_head = req;
|
|
|
|
else
|
|
|
|
queue->rskq_accept_tail->dl_next = req;
|
|
|
|
|
|
|
|
queue->rskq_accept_tail = req;
|
|
|
|
req->dl_next = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue *queue)
|
|
|
|
{
|
|
|
|
struct request_sock *req = queue->rskq_accept_head;
|
|
|
|
|
|
|
|
BUG_TRAP(req != NULL);
|
|
|
|
|
|
|
|
queue->rskq_accept_head = req->dl_next;
|
|
|
|
if (queue->rskq_accept_head == NULL)
|
|
|
|
queue->rskq_accept_tail = NULL;
|
|
|
|
|
|
|
|
return req;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline struct sock *reqsk_queue_get_child(struct request_sock_queue *queue,
|
|
|
|
struct sock *parent)
|
|
|
|
{
|
|
|
|
struct request_sock *req = reqsk_queue_remove(queue);
|
|
|
|
struct sock *child = req->sk;
|
|
|
|
|
|
|
|
BUG_TRAP(child != NULL);
|
|
|
|
|
|
|
|
sk_acceptq_removed(parent);
|
|
|
|
__reqsk_free(req);
|
|
|
|
return child;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int reqsk_queue_removed(struct request_sock_queue *queue,
|
|
|
|
struct request_sock *req)
|
|
|
|
{
|
2005-06-19 13:48:55 +08:00
|
|
|
struct listen_sock *lopt = queue->listen_opt;
|
2005-06-19 13:47:59 +08:00
|
|
|
|
|
|
|
if (req->retrans == 0)
|
|
|
|
--lopt->qlen_young;
|
|
|
|
|
|
|
|
return --lopt->qlen;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int reqsk_queue_added(struct request_sock_queue *queue)
|
|
|
|
{
|
2005-06-19 13:48:55 +08:00
|
|
|
struct listen_sock *lopt = queue->listen_opt;
|
2005-06-19 13:47:59 +08:00
|
|
|
const int prev_qlen = lopt->qlen;
|
|
|
|
|
|
|
|
lopt->qlen_young++;
|
|
|
|
lopt->qlen++;
|
|
|
|
return prev_qlen;
|
|
|
|
}
|
|
|
|
|
2005-08-10 11:10:42 +08:00
|
|
|
static inline int reqsk_queue_len(const struct request_sock_queue *queue)
|
2005-06-19 13:47:59 +08:00
|
|
|
{
|
|
|
|
return queue->listen_opt != NULL ? queue->listen_opt->qlen : 0;
|
|
|
|
}
|
|
|
|
|
2005-08-10 11:10:42 +08:00
|
|
|
static inline int reqsk_queue_len_young(const struct request_sock_queue *queue)
|
2005-06-19 13:47:59 +08:00
|
|
|
{
|
|
|
|
return queue->listen_opt->qlen_young;
|
|
|
|
}
|
|
|
|
|
2005-08-10 11:10:42 +08:00
|
|
|
static inline int reqsk_queue_is_full(const struct request_sock_queue *queue)
|
2005-06-19 13:47:59 +08:00
|
|
|
{
|
|
|
|
return queue->listen_opt->qlen >> queue->listen_opt->max_qlen_log;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void reqsk_queue_hash_req(struct request_sock_queue *queue,
|
|
|
|
u32 hash, struct request_sock *req,
|
2005-12-14 15:15:24 +08:00
|
|
|
unsigned long timeout)
|
2005-06-19 13:47:59 +08:00
|
|
|
{
|
2005-06-19 13:48:55 +08:00
|
|
|
struct listen_sock *lopt = queue->listen_opt;
|
2005-06-19 13:47:59 +08:00
|
|
|
|
|
|
|
req->expires = jiffies + timeout;
|
|
|
|
req->retrans = 0;
|
|
|
|
req->sk = NULL;
|
|
|
|
req->dl_next = lopt->syn_table[hash];
|
|
|
|
|
|
|
|
write_lock(&queue->syn_wait_lock);
|
|
|
|
lopt->syn_table[hash] = req;
|
|
|
|
write_unlock(&queue->syn_wait_lock);
|
|
|
|
}
|
|
|
|
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
#endif /* _REQUEST_SOCK_H */
|