Commits · 1218854afa6f659be90b748cf1bc7badee954a35 · nexedi / linux

25 Mar, 2008 13 commits

[NET] NETNS: Omit seq_net_private->net without CONFIG_NET_NS. · 1218854a

YOSHIFUJI Hideaki authored Mar 26, 2008

Without CONFIG_NET_NS, no namespace other than &init_net exists,
no need to store net in seq_net_private.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

1218854a

[NET] NETNS: Omit sock->sk_net without CONFIG_NET_NS. · 3b1e0a65

YOSHIFUJI Hideaki authored Mar 26, 2008

Introduce per-sock inlines: sock_net(), sock_net_set()
and per-inet_timewait_sock inlines: twsk_net(), twsk_net_set().
Without CONFIG_NET_NS, no namespace other than &init_net exists.
Let's explicitly define them to help compiler optimizations.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

3b1e0a65

[NET] NETNS: Omit net_device->nd_net without CONFIG_NET_NS. · c346dca1

YOSHIFUJI Hideaki authored Mar 25, 2008

Introduce per-net_device inlines: dev_net(), dev_net_set().
Without CONFIG_NET_NS, no namespace other than &init_net exists.
Let's explicitly define them to help compiler optimizations.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

c346dca1

[IPV6]: Support Source Address Selection API (RFC5014). · 7cbca67c
YOSHIFUJI Hideaki authored Mar 25, 2008
```
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
```
7cbca67c
[IPV6]: Use bitfields for hop_limit and mcast_hops. · 1d5d236d
YOSHIFUJI Hideaki authored Mar 10, 2008
```
Save some bits for future extensions.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
```
1d5d236d

[IPV6]: Optimize hop-limit determination. · 6b75d090

YOSHIFUJI Hideaki authored Mar 10, 2008

Last part of hop-limit determination is always:
    hoplimit = dst_metric(dst, RTAX_HOPLIMIT);
    if (hoplimit < 0)
        hoplimit = ipv6_get_hoplimit(dst->dev).

Let's consolidate it as ip6_dst_hoplimit(dst).
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

6b75d090

[IPV6]: Convert cork.hop_limit and cork.tclass into u8 instead of int. · 47254745

YOSHIFUJI Hideaki authored Mar 10, 2008

Values of those fields are always between 0 and 255 (inclusive),
so use u8 and save some memory on 32bit systems.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

47254745

[IPV4,IPV6]: Share cork.rt between IPv4 and IPv6. · c8cdaf99
YOSHIFUJI Hideaki authored Mar 10, 2008
```
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
```
c8cdaf99

[IPV6] ADDRCONF: Clean-up ipv6_dev_get_saddr(). · a9b05723

YOSHIFUJI Hideaki authored Mar 02, 2008

old:
|    text	   data	    bss	    dec	    hex	filename
|   28599	   1416	     96	  30111	   759f	net/ipv6/addrconf.o

new:
|    text	   data	    bss	    dec	    hex	filename
|   28007	   1416	     96	  29519	   734f	net/ipv6/addrconf.o
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

a9b05723

[XFRM] MIP6: Fix address keys for routing search. · 9bb182a7

YOSHIFUJI Hideaki authored Feb 22, 2008

Each MIPv6 XFRM state (DSTOPT/RH2) holds either destination or source
address to be mangled in the IPv6 header (that is "CoA").
On Inter-MN communication after both nodes binds each other,
they use route optimized traffic two MIPv6 states applied, and
both source and destination address in the IPv6 header
are replaced by the states respectively.
The packet format is correct, however, next-hop routing search
are not.
This patch fixes it by remembering address pairs for later states.

Based on patch from Masahide NAKAMURA <nakam@linux-ipv6.org>.
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

9bb182a7

[XFRM] IPV6: Optimize __xfrm_tunnel_alloc_spi(). · df8ea19b

YOSHIFUJI Hideaki authored Feb 19, 2008

| % size old/net/ipv6/xfrm6_tunnel.o new/net/ipv6/xfrm6_tunnel.o
|    text	   data	    bss	    dec	    hex	filename
|    1606	     40	   2080	   3726	    e8e	old/net/ipv6/xfrm6_tunnel.o
|    1574	     40	   2080	   3694	    e6e	new/net/ipv6/xfrm6_tunnel.o
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

df8ea19b

[XFRM] IPV6: Optimize xfrm6_input_addr(). · a002c6fd

YOSHIFUJI Hideaki authored Feb 19, 2008

| % size old/net/ipv6/xfrm6_input.o new/net/ipv6/xfrm6_input.o
|    text	   data	    bss	    dec	    hex	filename
|    1026	      0	      0	   1026	    402	old/net/ipv6/xfrm6_input.o
|     947	      0	      0	    947	    3b3	new/net/ipv6/xfrm6_input.o
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

a002c6fd

[XFRM] IPV6: Use distribution counting sort for xfrm_state/xfrm_tmpl chain. · 3b6cdf94
YOSHIFUJI Hideaki authored Feb 19, 2008
```
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
```
3b6cdf94

24 Mar, 2008 27 commits

[NETNS]: Enable TCP/UDP/ICMP inside namespace. · 92f1fecb

Denis V. Lunev authored Mar 24, 2008

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

92f1fecb

[NETNS]: Allow to create sockets in non-initial namespace. · 2342fd7e

Denis V. Lunev authored Mar 24, 2008

Allow to create sockets in the namespace if the protocol ok with this.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

2342fd7e

[NETNS]: Drop packets in the non-initial namespace on the per/protocol basis. · f145049a

Denis V. Lunev authored Mar 24, 2008

IP layer now can handle multiple namespaces normally. So, process such
packets normally and drop them only if the transport layer is not
aware about namespaces.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

f145049a

[NETNS]: Process netfilter hooks in initial namespace only. · 0be43f82

Denis V. Lunev authored Mar 24, 2008

There were no packets in the namespace other than initial
previously. This will be changed in the neareast future. Netfilters
are not namespace aware and should be processed in the initial
namespace only for now.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

0be43f82

[NETNS]: Process INET socket layer in the correct namespace. · 05cf89d4

Denis V. Lunev authored Mar 24, 2008

Replace all the reast of the init_net with a proper net on the socket
layer.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

05cf89d4

[NETNS]: Process IP layer in the context of the correct namespace. · cb84663e

Denis V. Lunev authored Mar 24, 2008

Replace all the rest of the init_net with a proper net on the IP layer.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

cb84663e

[NETNS]: Add namespace parameter to ip_cmsg_send. · 7a6adb92

Denis V. Lunev authored Mar 24, 2008

Pass the init_net there for now.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

7a6adb92

[NETNS]: Add namespace parameter to ip_options_get(...). · f2c4802b

Denis V. Lunev authored Mar 24, 2008

Pass the init_net there for now.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

f2c4802b

[NETNS]: Add namespace parameter to ip_options_compile. · 0e6bd4a1

Denis V. Lunev authored Mar 24, 2008

ip_options_compile uses inet_addr_type which requires a namespace. The
packet argument is optional, so parameter is the only way to obtain
it. Pass the init_net there for now.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

0e6bd4a1

[NETNS]: /proc/net/arp namespacing. · ffc31d3d

Denis V. Lunev authored Mar 24, 2008

Seqfile operation showing /proc/net/arp are already namespace
aware. All we need is to register this file for each namespace.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

ffc31d3d

[NETNS]: Process ARP in the context of the correct namespace. · 49e8a279

Denis V. Lunev authored Mar 24, 2008

Get namespace from a device and pass it to the routing engine. Enable
ARP packet processing and device notifiers after that.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

49e8a279

[NETNS]: Minor information leak via /proc/net/ptype file. · 2feb27db

Pavel Emelyanov authored Mar 24, 2008

This file displays the registered packet types, but some of them
(packet sockets creates such) can be bound to a net device and showing
them in a wrong namespace is not correct.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

2feb27db

[NETNS][UDP-Lite]: Register /proc/net/udplite(6) in a namespace. · 84c375af

Pavel Emelyanov authored Mar 24, 2008

UDP-Lite sockets are displayed in another files, rather than
UDP ones, so make the present in namespaces as well.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

84c375af

[UDP-Lite]: Clean up proc creation a bit. · ff2bac6a

Pavel Emelyanov authored Mar 24, 2008

Just introduce a helper to remove ifdefs from inside the
udplite4_register function. This will help to make the next patch
nicer.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

ff2bac6a

[NETNS][TCP]: Register /proc/net/tcp in a namespace. · 757764f6

Pavel Emelyanov authored Mar 24, 2008

After the commit f40c8174 ([NETNS][IPV4] 
tcp - make proc handle the network namespaces) it is now possible to make
this file present in newly created namespaces.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

757764f6

[NETNS][UDP]: Register /proc/net/udp in a namespace. · 15439feb

Pavel Emelyanov authored Mar 24, 2008

After the commit a91275ef ([NETNS][IPV6]
udp - make proc handle the network namespace) it is now possible to make
this file present in newly created namespaces.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

15439feb

Merge branch 'master' of ../net-2.6/ · 06802a81
David S. Miller authored Mar 23, 2008
```
Conflicts:

	net/ipv6/ndisc.c
```
06802a81
[CASSINI]: Use shorter list_splice_init() macro for brevity. · 9bd512f6
Robert P. J. Day authored Mar 23, 2008
```
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
```
9bd512f6

[SCTP]: Remove redundant wrapper functions. · 80445cfb

Florian Westphal authored Mar 23, 2008

sctp_datamsg_free and sctp_datamsg_track are just aliases for
sctp_datamsg_put and sctp_chunk_hold, respectively.

Saves 32 Bytes on x86.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

80445cfb

[SCTP]: Replace char msg[] with static const char[]. · 2444844c

Florian Westphal authored Mar 23, 2008

133886    2004     220  136110   213ae sctp.new/sctp.o
134018    2004     220  136242   21432 sctp.old/sctp.o
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

2444844c

fib_trie: print information on all routing tables · 3d3b2d25

Stephen Hemminger authored Mar 23, 2008

Make /proc/net/fib_trie and /proc/net/fib_triestat display
all routing tables, not just local and main.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

3d3b2d25

[AF_PACKET]: Remove unused variable. · 2a706ec1

Jiri Olsa authored Mar 23, 2008

Signed-off-by: Jiri Olsa <olsajiri@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

2a706ec1

[TCP]: Shrink syncookie_secret by 8 byte. · 2051f11f

Florian Westphal authored Mar 23, 2008

the first u32 copied from syncookie_secret is overwritten by the
minute-counter four lines below.  After adjusting the destination
address, the size of syncookie_secret can be reduced accordingly.

AFAICS, the only other user of syncookie_secret[] is the ipv6
syncookie support.  Because ipv6 syncookies only grab 44 bytes from
syncookie_secret[], this shouldn't affect them in any way.

With fixes from Glenn Griffin.
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Glenn Griffin <ggriffin.kernel@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

2051f11f

[NET]: include/linux/udp.h - remove duplicate include · 310afe86

Joe Perches authored Mar 23, 2008

Remove duplicate #include <linux/types.h>
Combine #ifdef __KERNEL__ blocks
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

310afe86

[NET]: include/linux/igmp.h - remove duplicate include · cc32e054

Joe Perches authored Mar 23, 2008

Removed duplicate #include <linux/skbuff.h>	
Combined #ifdef __KERNEL__ blocks
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

cc32e054

[NET]: include/linux/atalk.h - remove duplicate include · 414f69d8
Joe Perches authored Mar 23, 2008
```
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
```
414f69d8
[NET]: include/net/route.h - remove duplicate include · 7d164be8
Joe Perches authored Mar 23, 2008
```
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
```
7d164be8