Commit f5f80e32 authored by Eric Dumazet's avatar Eric Dumazet Committed by David S. Miller
Browse files

ipv6: remove hard coded limitation on ipv6_pinfo 



IPv6 inet sockets are supposed to have a "struct ipv6_pinfo"
field at the end of their definition, so that inet6_sk_generic()
can derive from socket size the offset of the "struct ipv6_pinfo".

This is very fragile, and prevents adding bigger alignment
in sockets, because inet6_sk_generic() does not work
if the compiler adds padding after the ipv6_pinfo component.

We are currently working on a patch series to reorganize
TCP structures for better data locality and found issues
similar to the one fixed in commit f5d54767
("tcp: fix tcp_inet6_sk() for 32bit kernels")

Alternative would be to force an alignment on "struct ipv6_pinfo",
greater or equal to __alignof__(any ipv6 sock) to ensure there is
no padding. This does not look great.

v2: fix typo in mptcp_proto_v6_init() (Paolo)

Signed-off-by: default avatarEric Dumazet <edumazet@google.com>
Cc: Chao Wu <wwchao@google.com>
Cc: Wei Wang <weiwan@google.com>
Cc: Coco Li <lixiaoyan@google.com>
Cc: YiFei Zhu <zhuyifei@google.com>
Reviewed-by: default avatarSimon Horman <simon.horman@corigine.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 1671bcfd

include/linux/ipv6.h

+4 −11
Original line number Diff line number Diff line
@@ -200,14 +200,7 @@ struct inet6_cork { 
	u8 tclass;

};



/**

 * struct ipv6_pinfo - ipv6 private area

 *

 * In the struct sock hierarchy (tcp6_sock, upd6_sock, etc)

 * this _must_ be the last member, so that inet6_sk_generic

 * is able to calculate its offset from the base struct sock

 * by using the struct proto->slab_obj_size member. -acme

 */

/* struct ipv6_pinfo - ipv6 private area */

struct ipv6_pinfo {

	struct in6_addr 	saddr;

	struct in6_pktinfo	sticky_pktinfo;
@@ -307,19 +300,19 @@ struct raw6_sock { 
	__u32			offset;		/* checksum offset  */

	struct icmp6_filter	filter;

	__u32			ip6mr_table;

	/* ipv6_pinfo has to be the last member of raw6_sock, see inet6_sk_generic */



	struct ipv6_pinfo	inet6;

};



struct udp6_sock {

	struct udp_sock	  udp;

	/* ipv6_pinfo has to be the last member of udp6_sock, see inet6_sk_generic */



	struct ipv6_pinfo inet6;

};



struct tcp6_sock {

	struct tcp_sock	  tcp;

	/* ipv6_pinfo has to be the last member of tcp6_sock, see inet6_sk_generic */



	struct ipv6_pinfo inet6;

};

include/net/sock.h

+1 −0
Original line number Diff line number Diff line
@@ -1339,6 +1339,7 @@ struct proto { 


	struct kmem_cache	*slab;

	unsigned int		obj_size;

	unsigned int		ipv6_pinfo_offset;

	slab_flags_t		slab_flags;

	unsigned int		useroffset;	/* Usercopy region offset */

	unsigned int		usersize;	/* Usercopy region size */

net/dccp/ipv6.c

+1 −0
Original line number Diff line number Diff line
@@ -1056,6 +1056,7 @@ static struct proto dccp_v6_prot = { 
	.orphan_count	   = &dccp_orphan_count,

	.max_header	   = MAX_DCCP_HEADER,

	.obj_size	   = sizeof(struct dccp6_sock),

	.ipv6_pinfo_offset = offsetof(struct dccp6_sock, inet6),

	.slab_flags	   = SLAB_TYPESAFE_BY_RCU,

	.rsk_prot	   = &dccp6_request_sock_ops,

	.twsk_prot	   = &dccp6_timewait_sock_ops,

net/dccp/ipv6.h

+0 −4
Original line number Diff line number Diff line
@@ -13,10 +13,6 @@ 


struct dccp6_sock {

	struct dccp_sock  dccp;

	/*

	 * ipv6_pinfo has to be the last member of dccp6_sock,

	 * see inet6_sk_generic.

	 */

	struct ipv6_pinfo inet6;

};

net/ipv6/af_inet6.c

+2 −2
Original line number Diff line number Diff line
@@ -102,9 +102,9 @@ bool ipv6_mod_enabled(void) 
}

EXPORT_SYMBOL_GPL(ipv6_mod_enabled);



static __inline__ struct ipv6_pinfo *inet6_sk_generic(struct sock *sk)

static struct ipv6_pinfo *inet6_sk_generic(struct sock *sk)

{

	const int offset = sk->sk_prot->obj_size - sizeof(struct ipv6_pinfo);

	const int offset = sk->sk_prot->ipv6_pinfo_offset;



	return (struct ipv6_pinfo *)(((u8 *)sk) + offset);

}