summaryrefslogtreecommitdiffstats
path: root/net
Commit message (Collapse)AuthorAgeFilesLines
* mac80211: Fix dynamic power save for scanning.Vivek Natarajan2009-12-072-4/+17
| | | | | | | | | | | | | | | Not only ps_sdata but also IEEE80211_CONF_PS is to be considered before restoring PS in scan_ps_disable(). For instance, when ps_sdata is set but CONF_PS is not set just because the dynamic timer is still running, a sw scan leads to setting of CONF_PS in scan_ps_disable instead of restarting the dynamic PS timer. Also for the above case, a null data frame is to be sent after returning to operating channel which was not happening with the current implementation. This patch fixes this too. Signed-off-by: Vivek Natarajan <vnatarajan@atheros.com> Reviewed-by: Kalle Valo <kalle.valo@nokia.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
* mac80211: recalculate idle later in MLMEJohannes Berg2009-12-071-2/+8
| | | | | | | | | | | | | | | hwsim testing has revealed that when the MLME recalculates the idle state of the device, it sometimes does so before sending the final deauthentication or disassociation frame. This patch changes the place where the idle state is recalculated, but of course driver transmit is typically asynchronous while configuration is expected to be synchronous, so it doesn't fix all possible cases yet. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
* Merge branch 'master' of /home/davem/src/GIT/linux-2.6/David S. Miller2009-12-052-0/+4
|\ | | | | | | | | | | | | | | Conflicts: drivers/net/pcmcia/fmvj18x_cs.c drivers/net/pcmcia/nmclan_cs.c drivers/net/pcmcia/xirc2ps_cs.c drivers/net/wireless/ray_cs.c
| * Merge branch 'core-printk-for-linus' of ↵Linus Torvalds2009-12-052-0/+4
| |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip * 'core-printk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: ratelimit: Make suppressed output messages more useful printk: Remove ratelimit.h from kernel.h ratelimit: Fix/allow use in atomic contexts ratelimit: Use per ratelimit context locking
| | * printk: Remove ratelimit.h from kernel.hIngo Molnar2009-09-222-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Decouple kernel.h from ratelimit.h: the global declaration of printk's ratelimit_state is not needed, and it leads to messy circular dependencies due to ratelimit.h's (new) adding of a spinlock_types.h include. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: David S. Miller <davem@davemloft.net> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* | | mac80211: fix reorder buffer releaseJohannes Berg2009-12-041-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | My patch "mac80211: correctly place aMPDU RX reorder code" uses an skb queue for MPDUs that were released from the buffer. I intentially didn't initialise and use the skb queue's spinlock, but in this place forgot that the code variant that doesn't touch the spinlock is needed. Thanks to Christian Lamparter for quickly spotting the bug in the backtrace Reinette reported. Reported-by: Reinette Chatre <reinette.chatre@intel.com> Bug-identified-by: Christian Lamparter <chunkeey@googlemail.com> Tested-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | Merge branch 'master' of ↵David S. Miller2009-12-0414-308/+526
|\ \ \ | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6
| * | | cfg80211: indent regulatory messages with spacesKalle Valo2009-12-041-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The regulatory messages in syslog look weird: kernel: cfg80211: Regulatory domain: US kernel: ^I(start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) kernel: ^I(2402000 KHz - 2472000 KHz @ 40000 KHz), (600 mBi, 2700 mBm) kernel: ^I(5170000 KHz - 5190000 KHz @ 40000 KHz), (600 mBi, 2300 mBm) kernel: ^I(5190000 KHz - 5210000 KHz @ 40000 KHz), (600 mBi, 2300 mBm) kernel: ^I(5210000 KHz - 5230000 KHz @ 40000 KHz), (600 mBi, 2300 mBm) kernel: ^I(5230000 KHz - 5330000 KHz @ 40000 KHz), (600 mBi, 2300 mBm) kernel: ^I(5735000 KHz - 5835000 KHz @ 40000 KHz), (600 mBi, 3000 mBm) Indent them with four spaces instead of the tab character to get prettier output. Signed-off-by: Kalle Valo <kalle.valo@nokia.com> Acked: Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | mac80211: Fix TX status reporting for injected data framesJouni Malinen2009-12-041-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | An earlier optimization on removing unnecessary traffic on cooked monitor interfaces ("mac80211: reduce the amount of unnecessary traffic on cooked monitor interfaces ") ended up removing quite a bit more than just unnecessary traffic. It was not supposed to remove TX status reporting for injected frames, but ended up doing it by checking the injected flag in skb->cb only after that field had been cleared with memset.. Fix this by taking a local copy of the injected flag before skb->cb is cleared. This broke user space applications that depend on getting TX status notifications for injected data frames. For example, STA inactivity poll from hostapd did not work and ended up kicking out stations even if they were still present. Signed-off-by: Jouni Malinen <j@w1.fi> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | WE: Fix set events not propagatedJean Tourrilhes2009-12-041-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | I've just noticed that some events are no longer propagated for some wireless drivers. Basically, SET request with a extra payload for driver without commit handler. The fix is pretty simple, see attached. Actually, a few lines below this line, you will see that the event generation for simple SET (iwpoint-less ?) is done properly, and this other event generation does not need fixing. Signed-off-by: Jean Tourrilhes <jt@hpl.hp.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | cfg80211: Add PMKSA wext compatibility handlerSamuel Ortiz2009-11-281-0/+42
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | With the addition of the *_pmksa cfg80211 ops, we can now add the corresponding wireless extensions compatibility handler. Signed-off-by: Samuel Ortiz <sameo@linux.intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | nl80211: PMKSA caching supportSamuel Ortiz2009-11-281-0/+120
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is an interface to set, delete and flush PMKIDs through nl80211. Main users would be fullmac devices which firmwares are capable of generating the RSN IEs for the re-association requests, e.g. iwmc3200wifi. Signed-off-by: Samuel Ortiz <sameo@linux.intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | mac80211: log more data when tracingJohannes Berg2009-11-281-2/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Enable logging of more configuration data when tracing is enabled. Except for the channel frequency this is only useful with the binary trace format, but that can be recorded and replayed with trace-cmd and I will be working on a plugin that reports all the information. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | mac80211: correctly place aMPDU RX reorder codeJohannes Berg2009-11-282-73/+96
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As indicated by the comment, the aMPDU RX reorder code should logically be after ieee80211_rx_h_check(). The previous patch moved the code there, and this patch now hooks it up in that place by introducing a list of skbs that are then processed by the remaining handlers. The list may be empty if the function is buffering the skb to release it later. The only change needed to the RX data is that the crypto handler needs to clear the key that may be set from a previous loop iteration, and that not everything can be in the rx flags now. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | mac80211: move aMPDU RX reorder codeJohannes Berg2009-11-281-230/+229
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This code should be part of RX handlers, so move it to the place where it belongs without changing it. A follow-up patch will do the changes to hook it up. The sole purpose of this code move is to make the other patch readable, it doesn't change the code at all except that it now requires a different static function declaration (which will go away too). Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | mac80211: move cmntr flag out of rx flagsJohannes Berg2009-11-282-4/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The RX flags should soon be used only for flags that cannot change within an a-MPDU, so move the cooked monitor flag into the RX status flags. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | wireless : use a dedicated workqueue for cfg80211.Alban Browaeys2009-11-285-6/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch moves the works cleanup, scan and events to a cfg80211 dedicated workqueue. Platform driver like eeepc-laptop ought to use works to rfkill (as new rfkill does lock in rfkill_unregister and the platform driver is called from rfkill_switch_all which also lock the same mutex). This raise a new issue in itself that the work scheduled by the platform driver to the global worqueue calls wiphy_unregister which flush_work scan and event works (which thus flush works on the global workqueue inside a work on the global workqueue) and also put on hold the wdev_cleanup_work (which prevents the dev_put on netdev thus indefinite Usage count error on wifi device). Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: Alban Browaeys <prahal@yahoo.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * | | mac80211: Speedup ieee80211_remove_interfaces()Eric Dumazet2009-11-251-9/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Speedup ieee80211_remove_interfaces() by factorizing synchronize_rcu() calls Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
* | | | tcp: fix a timewait refcnt raceEric Dumazet2009-12-031-3/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | After TCP RCU conversion, tw->tw_refcnt should not be set to 1 in inet_twsk_alloc(). It allows a RCU reader to get this timewait socket, while we not yet stabilized it. Only choice we have is to set tw_refcnt to 0 in inet_twsk_alloc(), then atomic_add() it later, once everything is done. Location of this atomic_add() is tricky, because we dont want another writer to find this timewait in ehash, while tw_refcnt is still zero ! Thanks to Kapil Dakhane tests and reports. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | tcp: connect() race with timewait reuseEric Dumazet2009-12-033-18/+45
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Its currently possible that several threads issuing a connect() find the same timewait socket and try to reuse it, leading to list corruptions. Condition for bug is that these threads bound their socket on same address/port of to-be-find timewait socket, and connected to same target. (SO_REUSEADDR needed) To fix this problem, we could unhash timewait socket while holding ehash lock, to make sure lookups/changes will be serialized. Only first thread finds the timewait socket, other ones find the established socket and return an EADDRNOTAVAIL error. This second version takes into account Evgeniy's review and makes sure inet_twsk_put() is called outside of locked sections. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | tcp: diag: Dont report negative values for rx queueEric Dumazet2009-12-032-3/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Both netlink and /proc/net/tcp interfaces can report transient negative values for rx queue. ss -> State Recv-Q Send-Q Local Address:Port Peer Address:Port ESTAB -6 6 127.0.0.1:45956 127.0.0.1:3333 netstat -> tcp 4294967290 6 127.0.0.1:37784 127.0.0.1:3333 ESTABLISHED This is because we dont lock socket while computing tp->rcv_nxt - tp->copied_seq, and another CPU can update copied_seq before rcv_next in RX path. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | netdevice: provide common routine for macvlan and vlan operstate managementPatrick Mullaney2009-12-032-25/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Provide common routine for the transition of operational state for a leaf device during a root device transition. Signed-off-by: Patrick Mullaney <pmullaney@novell.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | Merge branch 'master' of ↵David S. Miller2009-12-037-37/+149
|\ \ \ \ | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/holtmann/bluetooth-next-2.6
| * | | | Bluetooth: Add RFCOMM option to use L2CAP ERTM modeMarcel Holtmann2009-12-031-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | By default the RFCOMM layer would still use L2CAP basic mode. For testing purposes this option enables RFCOMM to select enhanced retransmission mode. Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Add L2CAP option for max transmit valueMarcel Holtmann2009-12-031-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For testing purposes it is important to modify the max transmit value. Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Fix 'SendRRorRNR' to send the ReqSeq valueGustavo F. Padovan2009-12-031-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | SendRRorRNR needs to acknowledge received I-frames (actually every packet needs to acknowledge received I-frames by sending the proper packet sequence number), so ReqSeq is set to the next I-frame number sequence to be pulled by the reassembly function. SendRRorRNR tells the remote side about local busy conditions, it sends a Receiver Ready frame if local busy is false or a Receiver Not Ready if local busy is true. ReqSeq is the packet's field to send the number of the acknowledged packets. Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Implement RejActioned flagGustavo F. Padovan2009-12-031-3/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | RejActioned is used to prevent retransmission when a entity is on the WAIT_F state, i.e., waiting for a frame with F-bit set due local busy condition or a expired retransmission timer. (When these two events raise they send a frame with the Poll bit set and enters in the WAIT_F state to wait for a frame with the Final bit set.) The local entity doesn't send I-frames(the data frames) until the receipt of a frame with F-bit set. When that happens it also set RejActioned to false. RejActioned is a mandatory feature of ERTM spec. Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Fix sending ReqSeq on I-framesGustavo F. Padovan2009-12-031-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As specified by ERTM spec an ERTM channel can acknowledge received I-frames(the data frames) by sending an I-frame with the proper ReqSeq value (i.e. ReqSeq is set to BufferSeq). Until now we aren't setting the ReqSeq value on I-frame control bits. That way we can save sending S-frames(Supervise frames) only to acknowledge receipt of I-frames. It is very helpful to the full-duplex channel. ReqSeq is the packet sequence number sent in an acknowledgement frame to acknowledge receipt of frames up to (ReqSeq - 1). BufferSeq controls the receiver buffer, it is used to delay acknowledgement of new frames to not cause buffer overflow. BufferSeq value is not increased until frames are pulled by reassembly function. Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Fix unset of SrejActioned flagGustavo F. Padovan2009-12-031-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | SrejActioned is a flag that when set prevents local side to retransmit a I-frame(the data frame) already retransmitted. The local entity can retransmit again only when it receives a SREJ frame with the F-bit set. SREJ frame - Selective Reject frame - is sent when an entity wants the retransmission of a specific I-frame that was lost or corrupted. This bug can put ERTM in an unknown state once the entity can't retransmit. A frame with the Final bit set is expected when the local side sends a frame with the Poll bit set due to a local busy condition or a retransmission timer expired. (Receipt of P-bit shall always be replied by a frame with the F-bit set). pi->conn_state keeps informations about many ERTM flags including SrejActioned. Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Initialize variables and timers for both channel's sidesGustavo F. Padovan2009-12-031-18/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix ERTM's full-duplex channel to work as specified by ERTM spec. ERTM needs to handle state vars, timers and counters to send and receive I-frames(the data frames), i.e., for both sides of data communication. We initialize all of them to the default values here. Full-duplex channel is a mandatory feature of ERTM spec. Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Fix handling of BNEP setup connection requestsVikram Kandukuri2009-12-031-1/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | According to BNEP test specification the proper response should be sent for a setup connection request message after the BNEP connection setup has been completed. Signed-off-by: Vikram Kandukuri <vikram.kandukuri@atheros.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Unobfuscate tasklet_schedule usageMarcel Holtmann2009-12-033-12/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The tasklet schedule function helpers are just an obfuscation. So remove them and call the schedule functions directly. Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Turn hci_recv_frame into an exported functionMarcel Holtmann2009-12-031-0/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For future simplification it is important that the hci_recv_frame function is no longer an inline function. So move it into the module itself and export it. Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Return ENETDOWN when interface is downMarcel Holtmann2009-12-031-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Sending commands to a down interface results in a timeout while clearly it should just return ENETDOWN. When using the ioctls this works fine, but not when using the HCI sockets sendmsg interface. Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * | | | Bluetooth: Implement raw output support for HIDP layerJiri Kosina2009-12-031-0/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Implement raw output callback which is used by hidraw to send raw data to the underlying device. Without this patch, the userspace hidraw-based applications can't send output reports to HID Bluetooth devices. Reported-and-tested-by: Brian Gunn <bgunn@solekai.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
* | | | | Merge branch 'master' of ↵David S. Miller2009-12-0331-257/+250
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/kaber/nf-next-2.6
| * | | | | netfilter: net/ipv[46]/netfilter: Move && and || to end of previous lineJoe Perches2009-11-2322-174/+163
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Compile tested only. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: xtables: fix conntrack match v1 ipt-save outputFlorian Westphal2009-11-231-44/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit d6d3f08b0fd998b647a05540cedd11a067b72867 (netfilter: xtables: conntrack match revision 2) does break the v1 conntrack match iptables-save output in a subtle way. Problem is as follows: up = kmalloc(sizeof(*up), GFP_KERNEL); [..] /* * The strategy here is to minimize the overhead of v1 matching, * by prebuilding a v2 struct and putting the pointer into the * v1 dataspace. */ memcpy(up, info, offsetof(typeof(*info), state_mask)); [..] *(void **)info = up; As the v2 struct pointer is saved in the match data space, it clobbers the first structure member (->origsrc_addr). Because the _v1 match function grabs this pointer and does not actually look at the v1 origsrc, run time functionality does not break. But iptables -nvL (or iptables-save) cannot know that v1 origsrc_addr has been overloaded in this way: $ iptables -p tcp -A OUTPUT -m conntrack --ctorigsrc 10.0.0.1 -j ACCEPT $ iptables-save -A OUTPUT -p tcp -m conntrack --ctorigsrc 128.173.134.206 -j ACCEPT (128.173... is the address to the v2 match structure). To fix this, we take advantage of the fact that the v1 and v2 structures are identical with exception of the last two structure members (u8 in v1, u16 in v2). We extract them as early as possible and prevent the v2 matching function from looking at those two members directly. Previously reported by Michel Messerschmidt via Ben Hutchings, also see Debian Bug tracker #556587. Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: nf_ct_tcp: improve out-of-sync situation in TCP trackingPablo Neira Ayuso2009-11-231-10/+41
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Without this patch, if we receive a SYN packet from the client while the firewall is out-of-sync, we let it go through. Then, if we see the SYN/ACK reply coming from the server, we destroy the conntrack entry and drop the packet to trigger a new retransmission. Then, the retransmision from the client is used to start a new clean session. This patch improves the current handling. Basically, if we see an unexpected SYN packet, we annotate the TCP options. Then, if we see the reply SYN/ACK, this means that the firewall was indeed out-of-sync. Therefore, we set a clean new session from the existing entry based on the annotated values. This patch adds two new 8-bits fields that fit in a 16-bits gap of the ip_ct_tcp structure. This patch is particularly useful for conntrackd since the asynchronous nature of the state-synchronization allows to have backup nodes that are not perfect copies of the master. This helps to improve the recovery under some worst-case scenarios. I have tested this by creating lots of conntrack entries in wrong state: for ((i=1024;i<65535;i++)); do conntrack -I -p tcp -s 192.168.2.101 -d 192.168.2.2 --sport $i --dport 80 -t 800 --state ESTABLISHED -u ASSURED,SEEN_REPLY; done Then, I make some TCP connections: $ echo GET / | nc 192.168.2.2 80 The events show the result: [UPDATE] tcp 6 60 SYN_RECV src=192.168.2.101 dst=192.168.2.2 sport=33220 dport=80 src=192.168.2.2 dst=192.168.2.101 sport=80 dport=33220 [ASSURED] [UPDATE] tcp 6 432000 ESTABLISHED src=192.168.2.101 dst=192.168.2.2 sport=33220 dport=80 src=192.168.2.2 dst=192.168.2.101 sport=80 dport=33220 [ASSURED] [UPDATE] tcp 6 120 FIN_WAIT src=192.168.2.101 dst=192.168.2.2 sport=33220 dport=80 src=192.168.2.2 dst=192.168.2.101 sport=80 dport=33220 [ASSURED] [UPDATE] tcp 6 30 LAST_ACK src=192.168.2.101 dst=192.168.2.2 sport=33220 dport=80 src=192.168.2.2 dst=192.168.2.101 sport=80 dport=33220 [ASSURED] [UPDATE] tcp 6 120 TIME_WAIT src=192.168.2.101 dst=192.168.2.2 sport=33220 dport=80 src=192.168.2.2 dst=192.168.2.101 sport=80 dport=33220 [ASSURED] and tcpdump shows no retransmissions: 20:47:57.271951 IP 192.168.2.101.33221 > 192.168.2.2.www: S 435402517:435402517(0) win 5840 <mss 1460,sackOK,timestamp 4294961827 0,nop,wscale 6> 20:47:57.273538 IP 192.168.2.2.www > 192.168.2.101.33221: S 3509927945:3509927945(0) ack 435402518 win 5792 <mss 1460,sackOK,timestamp 235681024 4294961827,nop,wscale 4> 20:47:57.273608 IP 192.168.2.101.33221 > 192.168.2.2.www: . ack 3509927946 win 92 <nop,nop,timestamp 4294961827 235681024> 20:47:57.273693 IP 192.168.2.101.33221 > 192.168.2.2.www: P 435402518:435402524(6) ack 3509927946 win 92 <nop,nop,timestamp 4294961827 235681024> 20:47:57.275492 IP 192.168.2.2.www > 192.168.2.101.33221: . ack 435402524 win 362 <nop,nop,timestamp 235681024 4294961827> 20:47:57.276492 IP 192.168.2.2.www > 192.168.2.101.33221: P 3509927946:3509928082(136) ack 435402524 win 362 <nop,nop,timestamp 235681025 4294961827> 20:47:57.276515 IP 192.168.2.101.33221 > 192.168.2.2.www: . ack 3509928082 win 108 <nop,nop,timestamp 4294961828 235681025> 20:47:57.276521 IP 192.168.2.2.www > 192.168.2.101.33221: F 3509928082:3509928082(0) ack 435402524 win 362 <nop,nop,timestamp 235681025 4294961827> 20:47:57.277369 IP 192.168.2.101.33221 > 192.168.2.2.www: F 435402524:435402524(0) ack 3509928083 win 108 <nop,nop,timestamp 4294961828 235681025> 20:47:57.279491 IP 192.168.2.2.www > 192.168.2.101.33221: . ack 435402525 win 362 <nop,nop,timestamp 235681025 4294961828> I also added a rule to log invalid packets, with no occurrences :-) . Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Acked-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu> Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: remove unneccessary checks from netlink notifiersPatrick McHardy2009-11-064-8/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The NETLINK_URELEASE notifier is only invoked for bound sockets, so there is no need to check ->pid again. Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: nf_nat_helper: tidy up adjust_tcp_sequenceHannes Eder2009-11-051-13/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The variable 'other_way' gets initialized but is not read afterwards, so remove it. Pass the right arguments to a pr_debug call. While being at tidy up a bit and it fix this checkpatch warning: WARNING: suspect code indent for conditional statements Signed-off-by: Hannes Eder <heder@google.com> Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: nf_conntrack: avoid additional compare.Changli Gao2009-11-051-4/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Signed-off-by: Changli Gao <xiaosuo@gmail.com> Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: remove synchronize_net() calls in ip_queue/ip6_queueEric Dumazet2009-11-042-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | nf_unregister_queue_handlers() already does a synchronize_rcu() call, we dont need to do it again in callers. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Patrick McHardy <kaber@trash.net>
| * | | | | netfilter: xt_socket: make module available for INPUT chainJan Engelhardt2009-10-291-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This should make it possible to test for the existence of local sockets in the INPUT path. References: http://marc.info/?l=netfilter-devel&m=125380481517129&w=2 Signed-off-by: Jan Engelhardt <jengelh@medozas.de> Signed-off-by: Balazs Scheidler <bazsi@balabit.hu> Signed-off-by: Patrick McHardy <kaber@trash.net>
* | | | | | net: Batch inet_twsk_purgeEric W. Biederman2009-12-033-11/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This function walks the whole hashtable so there is no point in passing it a network namespace. Instead I purge all timewait sockets from dead network namespaces that I find. If the namespace is one of the once I am trying to purge I am guaranteed no new timewait sockets can be formed so this will get them all. If the namespace is one I am not acting for it might form a few more but I will call inet_twsk_purge again and shortly to get rid of them. In any even if the network namespace is dead timewait sockets are useless. Move the calls of inet_twsk_purge into batch_exit routines so that if I am killing a bunch of namespaces at once I will just call inet_twsk_purge once and save a lot of redundant unnecessary work. My simple 4k network namespace exit test the cleanup time dropped from roughly 8.2s to 1.6s. While the time spent running inet_twsk_purge fell to about 2ms. 1ms for ipv4 and 1ms for ipv6. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | | | net: Use rcu lookups in inet_twsk_purge.Eric W. Biederman2009-12-031-15/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | While we are looking up entries to free there is no reason to take the lock in inet_twsk_purge. We have to drop locks and restart occassionally anyway so adding a few more in case we get on the wrong list because of a timewait move is no big deal. At the same time not taking the lock for long periods of time is much more polite to the rest of the users of the hash table. In my test configuration of killing 4k network namespaces this change causes 4k back to back runs of inet_twsk_purge on an empty hash table to go from roughly 20.7s to 3.3s, and the total time to destroy 4k network namespaces goes from roughly 44s to 3.3s. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | | | net: Allow fib_rule_unregister to batchEric W. Biederman2009-12-034-37/+55
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Refactor the code so fib_rules_register always takes a template instead of the actual fib_rules_ops structure that will be used. This is required for network namespace support so 2 out of the 3 callers already do this, it allows the error handling to be made common, and it allows fib_rules_unregister to free the template for hte caller. Modify fib_rules_unregister to use call_rcu instead of syncrhonize_rcu to allw multiple namespaces to be cleaned up in the same rcu grace period. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | | | netns: Add an explicit rcu_barrier to unregister_pernet_{device|subsys}Eric W. Biederman2009-12-031-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows namespace exit methods to batch work that comes requires an rcu barrier using call_rcu without having to treat the unregister_pernet_operations cases specially. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | | | net: Allow xfrm_user_net_exit to batch efficiently.Eric W. Biederman2009-12-031-8/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | xfrm.nlsk is provided by the xfrm_user module and is access via rcu from other parts of the xfrm code. Add xfrm.nlsk_stash a copy of xfrm.nlsk that will never be set to NULL. This allows the synchronize_net and netlink_kernel_release to be deferred until a whole batch of xfrm.nlsk sockets have been set to NULL. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | | | net: Move network device exit batchingEric W. Biederman2009-12-032-24/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Move network device exit batching from a special case in net_namespace.c to using common mechanisms in dev.c Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
OpenPOWER on IntegriCloud