linux/net
Herbert Xu f845172531 cls_cgroup: Store classid in struct sock
Up until now cls_cgroup has relied on fetching the classid out of
the current executing thread.  This runs into trouble when a packet
processing is delayed in which case it may execute out of another
thread's context.

Furthermore, even when a packet is not delayed we may fail to
classify it if soft IRQs have been disabled, because this scenario
is indistinguishable from one where a packet unrelated to the
current thread is processed by a real soft IRQ.

In fact, the current semantics is inherently broken, as a single
skb may be constructed out of the writes of two different tasks.
A different manifestation of this problem is when the TCP stack
transmits in response of an incoming ACK.  This is currently
unclassified.

As we already have a concept of packet ownership for accounting
purposes in the skb->sk pointer, this is a natural place to store
the classid in a persistent manner.

This patch adds the cls_cgroup classid in struct sock, filling up
an existing hole on 64-bit :)

The value is set at socket creation time.  So all sockets created
via socket(2) automatically gains the ID of the thread creating it.
Whenever another process touches the socket by either reading or
writing to it, we will change the socket classid to that of the
process if it has a valid (non-zero) classid.

For sockets created on inbound connections through accept(2), we
inherit the classid of the original listening socket through
sk_clone, possibly preceding the actual accept(2) call.

In order to minimise risks, I have not made this the authoritative
classid.  For now it is only used as a backup when we execute
with soft IRQs disabled.  Once we're completely happy with its
semantics we can use it as the sole classid.

Footnote: I have rearranged the error path on cls_group module
creation.  If we didn't do this, then there is a window where
someone could create a tc rule using cls_group before the cgroup
subsystem has been registered.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-24 00:12:34 -07:00
..
9p net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
802 Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-04-11 14:53:53 -07:00
8021q net: Remove unnecessary semicolons after switch statements 2010-05-17 17:44:35 -07:00
appletalk Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-04-11 14:53:53 -07:00
atm net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
ax25 net: sk_sleep() helper 2010-04-20 16:37:13 -07:00
bluetooth net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
bridge bridge: fix build for CONFIG_SYSFS disabled 2010-05-18 12:26:27 -07:00
caif caif: Bugfix - use MSG_TRUNC in receive 2010-05-23 23:57:43 -07:00
can net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
core cls_cgroup: Store classid in struct sock 2010-05-24 00:12:34 -07:00
dcb include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
dccp net: sock_def_readable() and friends RCU conversion 2010-05-01 15:00:15 -07:00
decnet net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
dsa Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-04-11 14:53:53 -07:00
econet include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
ethernet net: Inline skb_pull() in eth_type_trans(). 2010-05-02 02:21:44 -07:00
ieee802154 ieee802154: Fix possible NULL pointer dereference in wpan_phy_alloc 2010-05-23 23:11:07 -07:00
ipv4 Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6 2010-05-20 21:04:44 -07:00
ipv6 ipv6: Never schedule DAD timer on dead address 2010-05-18 15:56:06 -07:00
ipx include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
irda net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
iucv net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
key pfkey: add severity to printk 2010-05-17 23:23:13 -07:00
l2tp l2tp_eth: fix memory allocation 2010-04-23 16:37:33 -07:00
lapb include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
llc Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-05-12 00:05:35 -07:00
mac80211 Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6 2010-05-20 21:04:44 -07:00
netfilter Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/kaber/nf-next-2.6 2010-05-20 23:12:18 -07:00
netlabel net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
netlink Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-04-11 14:53:53 -07:00
netrom net: sk_sleep() helper 2010-04-20 16:37:13 -07:00
packet Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-04-21 01:14:25 -07:00
phonet net: sock_def_readable() and friends RCU conversion 2010-05-01 15:00:15 -07:00
rds net: Remove unnecessary semicolons after switch statements 2010-05-17 17:44:35 -07:00
rfkill Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2010-04-11 14:53:53 -07:00
rose net: sk_sleep() helper 2010-04-20 16:37:13 -07:00
rxrpc net: sock_def_readable() and friends RCU conversion 2010-05-01 15:00:15 -07:00
sched cls_cgroup: Store classid in struct sock 2010-05-24 00:12:34 -07:00
sctp net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
sunrpc Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6 2010-05-20 21:04:44 -07:00
tipc tipc: Reduce footprint by un-inlining tipc_msg_* routines 2010-05-12 23:02:29 -07:00
unix unix/garbage: kill copy of the skb queue walker 2010-05-03 15:39:58 -07:00
wanrouter
wimax Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6 2010-05-20 21:04:44 -07:00
wireless Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6 into for-davem 2010-05-11 14:24:55 -04:00
x25 X25: Remove bkl in sockopts 2010-05-17 17:39:28 -07:00
xfrm net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
compat.c include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
Kconfig l2tp: Split pppol2tp patch into separate l2tp and ppp parts 2010-04-03 14:56:02 -07:00
Makefile l2tp: Split pppol2tp patch into separate l2tp and ppp parts 2010-04-03 14:56:02 -07:00
nonet.c
socket.c cls_cgroup: Store classid in struct sock 2010-05-24 00:12:34 -07:00
sysctl_net.c net: Remove unnecessary returns from void function()s 2010-05-17 23:23:14 -07:00
TUNABLE