Skip to content
  1. Dec 10, 2010
    • Martin Willi's avatar
      xfrm: Add Traffic Flow Confidentiality padding XFRM attribute · 35d2856b
      Martin Willi authored 
      

The XFRMA_TFCPAD attribute for XFRM state installation configures
Traffic Flow Confidentiality by padding ESP packets to a specified
length.

Signed-off-by: default avatarMartin Willi <martin@strongswan.org>
Acked-by: default avatarHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      35d2856b
    • Eric Dumazet's avatar
      net: optimize INET input path further · 68835aba
      Eric Dumazet authored 
      

Followup of commit b178bb3d (net: reorder struct sock fields)

Optimize INET input path a bit further, by :

1) moving sk_refcnt close to sk_lock.

This reduces number of dirtied cache lines by one on 64bit arches (and
64 bytes cache line size).

2) moving inet_daddr & inet_rcv_saddr at the beginning of sk

(same cache line than hash / family / bound_dev_if / nulls_node)

This reduces number of accessed cache lines in lookups by one, and dont
increase size of inet and timewait socks.
inet and tw sockets now share same place-holder for these fields.

Before patch :

offsetof(struct sock, sk_refcnt) = 0x10
offsetof(struct sock, sk_lock) = 0x40
offsetof(struct sock, sk_receive_queue) = 0x60
offsetof(struct inet_sock, inet_daddr) = 0x270
offsetof(struct inet_sock, inet_rcv_saddr) = 0x274

After patch :

offsetof(struct sock, sk_refcnt) = 0x44
offsetof(struct sock, sk_lock) = 0x48
offsetof(struct sock, sk_receive_queue) = 0x68
offsetof(struct inet_sock, inet_daddr) = 0x0
offsetof(struct inet_sock, inet_rcv_saddr) = 0x4

compute_score() (udp or tcp) now use a single cache line per ignored
item, instead of two.

Signed-off-by: default avatarEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      68835aba
  3. Dec 09, 2010
    • David S. Miller's avatar
      net: Abstract away all dst_entry metrics accesses. · defb3519
      David S. Miller authored 
      

Use helper functions to hide all direct accesses, especially writes,
to dst_entry metrics values.

This will allow us to:

1) More easily change how the metrics are stored.

2) Implement COW for metrics.

In particular this will help us put metrics into the inetpeer
cache if that is what we end up doing.  We can make the _metrics
member a pointer instead of an array, initially have it point
at the read-only metrics in the FIB, and then on the first set
grab an inetpeer entry and point the _metrics member there.

Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
Acked-by: default avatarEric Dumazet <eric.dumazet@gmail.com>
      defb3519
  5. Dec 06, 2010
    • Eric Dumazet's avatar
      filter: fix sk_filter rcu handling · 46bcf14f
      Eric Dumazet authored 
      

Pavel Emelyanov tried to fix a race between sk_filter_(de|at)tach and
sk_clone() in commit 47e958ea

Problem is we can have several clones sharing a common sk_filter, and
these clones might want to sk_filter_attach() their own filters at the
same time, and can overwrite old_filter->rcu, corrupting RCU queues.

We can not use filter->rcu without being sure no other thread could do
the same thing.

Switch code to a more conventional ref-counting technique : Do the
atomic decrement immediately and queue one rcu call back when last
reference is released.

Signed-off-by: default avatarEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      46bcf14f
  7. Dec 02, 2010
  9. Dec 01, 2010
  11. Nov 30, 2010
  13. Nov 29, 2010
  15. Nov 28, 2010
  17. Nov 24, 2010
  19. Nov 22, 2010
    • Luis R. Rodriguez's avatar
      cfg80211: Fix regulatory bug with multiple cards and delays · b2e253cf
      Luis R. Rodriguez authored 
      

When two cards are connected with the same regulatory domain
if CRDA had a delayed response then cfg80211's own set regulatory
domain would still be the world regulatory domain. There was a bug
on cfg80211's logic such that it assumed that once you pegged a
request as the last request it was already the currently set
regulatory domain. This would mean we would race setting a stale
regulatory domain to secondary cards which had the same regulatory
domain since the alpha2 would match.

We fix this by processing each regulatory request atomically,
and only move on to the next one once we get it fully processed.
In the case CRDA is not present we will simply world roam.

This issue is only present when you have a slow system and the
CRDA processing is delayed. Because of this it is not a known
regression.

Without this fix when a delay is present with CRDA the second card
would end up with an intersected regulatory domain and not allow it
to use the channels it really is designed for. When two cards with
two different regulatory domains were inserted you'd end up
rejecting the second card's regulatory domain request.
This fails with mac80211_hswim's regtest=2 (two requests, same alpha2)
and regtest=3 (two requests, different alpha2) module parameter
options.

This was reproduced and tested against mac80211_hwsim using this
CRDA delayer:

       #!/bin/bash
       echo $COUNTRY >> /tmp/log
       sleep 2
       /sbin/crda.orig

And these regulatory tests:

       modprobe mac80211_hwsim regtest=2
       modprobe mac80211_hwsim regtest=3

Reported-by: default avatarMark Mentovai <mark@moxienet.com>
Signed-off-by: default avatarLuis R. Rodriguez <lrodriguez@atheros.com>
Tested-by: default avatarMark Mentovai <mark@moxienet.com>
Tested-by: default avatarBruno Randolf <br1@einfach.org>
Signed-off-by: default avatarJohn W. Linville <linville@tuxdriver.com>
      b2e253cf
  21. Nov 21, 2010
  23. Nov 18, 2010