summaryrefslogtreecommitdiff
path: root/net
AgeCommit message (Collapse)Author
2013-10-06Bluetooth: Simplify device type check for AMP discover responseMarcel Holtmann
When counting the number for AMP controllers, a positive check is used. To be consistent, use the same check when actually adding the data for the AMP contollers. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Use explicit check for BR/EDR device typeMarcel Holtmann
The BR/EDR and LE setup procedures apply only to BR/EDR device types and so check for that explicitly. Checking that it is not an AMP controller is dangerous in case there will be ever a third device type. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Reject enabling controllers without valid addressesMarcel Holtmann
In case of a single mode LE-only controller it is possible that no public address is used. These type of controllers require a random address to be configured. Without a configured static random address, such a controller is not functional. So reject powering on the controller in this case until it gets configured with a random address. The controller setup stage is still run since it is the only way to determinate if a public address is available or not. So it is similar on how RFKILL gets handled during initial setup of the controller. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Check for non BR/EDR controller in AMP discover responseMarcel Holtmann
Within the AMP discover response, all controllers that are not the primary BR/EDR controller are valid. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Simplify check if L2CAP connection is AMP capableMarcel Holtmann
The check if a L2CAP connection is AMP capable was a little bit complicated. This changes the code to make it simpler and more readable. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Move hci_amp_capable() function into L2CAP coreMarcel Holtmann
The hci_amp_capable() function has only a single user inside the L2CAP core. Instead of exporting the function, place it next to its user. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Remove check for number of AMP controllerMarcel Holtmann
The number of controllers for the AMP discover response has already been calculated. And since the hci_dev_list lock is held, it can not change. So there is no need for any extra checks. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Remove pointless inline functionMarcel Holtmann
The inline function for BR/EDR controller AMP discover response info is rather useless. Just include the code into the function that builds the whole response. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Rename AMP status constants and use themMarcel Holtmann
The AMP controller status constants need to be actually used to avoid crypted hardcoded numbers. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Use explicit AMP controller id value for BR/EDRMarcel Holtmann
The special AMP controller id 0 is reserved for the BR/EDR controller that has the main link. It is a fixed value and so use a constant for this throughout the code to make it more visible when the handling is for the BR/EDR channel or when it is for the AMP channel. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Separate AMP controller type from HCI device typeMarcel Holtmann
There are two defined HCI device types. One is for BR/EDR controllers and the other is for AMP controllers. The HCI device type is not the same as the AMP controller type. It just happens that currently the defined types match, but that is not guaranteed. Split the usage of AMP controller type into its own domain so that it is possible to separate between BR/EDR controllers, 802.11 AMP controllers and any other AMP technology that might be defined in the future. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Remove useless external function to count controllersMarcel Holtmann
The list of controllers can be counted ahead of time and inline inside the AMP discover handling. There is no need to export such a function at all. In addition just count the AMP controller and only allocated space for a single mandatory BR/EDR controller. No need to allocate more space than needed. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-06Bluetooth: Fix controller list for AMP discover responseMarcel Holtmann
The AMP discover response should list exactly one BR/EDR controller and ignore all other BR/EDR controller. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-05Bluetooth: Fix re-enabling advertising after a connectionJohan Hedberg
LE controllers will automatically disable advertising whenever they accept a new connection. In order not to fall out of sync with the advertising setting we need to re-enable advertising whenever the last LE connection drops. A failure to re-enable advertising should cause the setting to be disabled, so this patch also calls mgmt_new_settings() when this happens. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2013-10-05Bluetooth: Add public mgmt function to send New Settings eventJohan Hedberg
A function is needed so that the HCI event processing can ask the mgmt code to emit a new settings event. This is necessary e.g. when the event processing does updates to mgmt related states without any dependency of actual mgmt commands. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2013-10-05Bluetooth: Rename HCI_LE_PERIPHERAL to HCI_ADVERTISINGJohan Hedberg
This flag is used to indicate whether we want to have advertising enabled or not, so give it a more suitable name. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2013-10-04tcp: do not forget FIN in tcp_shifted_skb()Eric Dumazet
Yuchung found following problem : There are bugs in the SACK processing code, merging part in tcp_shift_skb_data(), that incorrectly resets or ignores the sacked skbs FIN flag. When a receiver first SACK the FIN sequence, and later throw away ofo queue (e.g., sack-reneging), the sender will stop retransmitting the FIN flag, and hangs forever. Following packetdrill test can be used to reproduce the bug. $ cat sack-merge-bug.pkt `sysctl -q net.ipv4.tcp_fack=0` // Establish a connection and send 10 MSS. 0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 +.000 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +.000 bind(3, ..., ...) = 0 +.000 listen(3, 1) = 0 +.050 < S 0:0(0) win 32792 <mss 1000,sackOK,nop,nop,nop,wscale 7> +.000 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6> +.001 < . 1:1(0) ack 1 win 1024 +.000 accept(3, ..., ...) = 4 +.100 write(4, ..., 12000) = 12000 +.000 shutdown(4, SHUT_WR) = 0 +.000 > . 1:10001(10000) ack 1 +.050 < . 1:1(0) ack 2001 win 257 +.000 > FP. 10001:12001(2000) ack 1 +.050 < . 1:1(0) ack 2001 win 257 <sack 10001:11001,nop,nop> +.050 < . 1:1(0) ack 2001 win 257 <sack 10001:12002,nop,nop> // SACK reneg +.050 < . 1:1(0) ack 12001 win 257 +0 %{ print "unacked: ",tcpi_unacked }% +5 %{ print "" }% First, a typo inverted left/right of one OR operation, then code forgot to advance end_seq if the merged skb carried FIN. Bug was added in 2.6.29 by commit 832d11c5cd076ab ("tcp: Try to restore large SKBs while SACK processing") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Acked-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-04Merge branch 'master' of ↵David S. Miller
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next Pablo Neira Ayuso says: ==================== The following patchset contains Netfilter updates for your net-next tree, mostly ipset improvements and enhancements features, they are: * Don't call ip_nest_end needlessly in the error path from me, suggested by Pablo Neira Ayuso, from Jozsef Kadlecsik. * Fixed sparse warnings about shadowed variable and missing rcu annotation and fix of "may be used uninitialized" warnings, also from Jozsef. * Renamed simple macro names to avoid namespace issues, reported by David Laight, again from Jozsef. * Use fix sized type for timeout in the extension part, and cosmetic ordering of matches and targets separatedly in xt_set.c, from Jozsef. * Support package fragments for IPv4 protos without ports from Anders K. Pedersen. For example this allows a hash:ip,port ipset containing the entry 192.168.0.1,gre:0 to match all package fragments for PPTP VPN tunnels to/from the host. Without this patch only the first package fragment (with fragment offset 0) was matched. * Introduced a new operation to get both setname and family, from Jozsef. ip[6]tables set match and SET target need to know the family of the set in order to reject adding rules which refer to a set with a non-mathcing family. Currently such rules are silently accepted and then ignored instead of generating an error message to the user. * Reworked extensions support in ipset types from Jozsef. The approach of defining structures with all variations is not manageable as the number of extensions grows. Therefore a blob for the extensions is introduced, somewhat similar to conntrack. The support of extensions which need a per data destroy function is added as well. * When an element timed out in a list:set type of set, the garbage collector skipped the checking of the next element. So the purging was delayed to the next run of the gc, fixed by Jozsef. * A small Kconfig fix: NETFILTER_NETLINK cannot be selected and ipset requires it. * hash:net,net type from Oliver Smith. The type provides the ability to store pairs of subnets in a set. * Comment for ipset entries from Oliver Smith. This makes possible to annotate entries in a set with comments, for example: ipset n foo hash:net,net comment ipset a foo 10.0.0.0/21,192.168.1.0/24 comment "office nets A and B" * Fix of hash types resizing with comment extension from Jozsef. * Fix of new extensions for list:set type when an element is added into a slot from where another element was pushed away from Jozsef. * Introduction of a common function for the listing of the element extensions from Jozsef. * Net namespace support for ipset from Vitaly Lavrov. * hash:net,port,net type from Oliver Smith, which makes possible to store the triples of two subnets and a protocol, port pair in a set. * Get xt_TCPMSS working with net namespace, by Gao feng. * Use the proper net netnamespace to allocate skbs, also by Gao feng. * A couple of cleanups for the conntrack SIP helper, by Holger Eitzenberger. * Extend cttimeout to allow setting default conntrack timeouts via nfnetlink, so we can get rid of all our sysctl/proc interfaces in the future for timeout tuning, from me. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-04sched/wait: Make the __wait_event*() interface more friendlyPeter Zijlstra
Change all __wait_event*() implementations to match the corresponding wait_event*() signature for convenience. In particular this does away with the weird 'ret' logic. Since there are __wait_event*() users this requires we update them too. Reviewed-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20131002092529.042563462@infradead.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-10-04openvswitch: Simplify mega-flow APIs.Pravin B Shelar
Hides mega-flow implementation in flow_table.c rather than datapath.c. Signed-off-by: Pravin B Shelar <pshelar@nicira.com> Signed-off-by: Jesse Gross <jesse@nicira.com>
2013-10-04openvswitch: Move mega-flow list out of rehashing struct.Pravin B Shelar
ovs-flow rehash does not touch mega flow list. Following patch moves it dp struct datapath. Avoid one extra indirection for accessing mega-flow list head on every packet receive. Signed-off-by: Pravin B Shelar <pshelar@nicira.com> Signed-off-by: Jesse Gross <jesse@nicira.com>
2013-10-03Bluetooth: Initialize hci_conn fields in hci_connect_leAndre Guedes
This patch moves some hci_conn fields initialization from hci_le_ create_connection() to hci_connect_le(). It makes more sense to initialize these fields within the function that creates the hci_ conn object. Signed-off-by: Andre Guedes <andre.guedes@openbossa.org> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2013-10-03Bluetooth: Rename hci_conn variable in hci_connect_le()Andre Guedes
This patch simply rename the hci_conn variable "le" to "conn" since it is a better name. Signed-off-by: Andre Guedes <andre.guedes@openbossa.org> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2013-10-03openvswitch: Restructure datapath.c and flow.cPravin B Shelar
Over the time datapath.c and flow.c has became pretty large files. Following patch restructures functionality of component into three different components: flow.c: contains flow extract. flow_netlink.c: netlink flow api. flow_table.c: flow table api. This patch restructures code without changing logic. Signed-off-by: Pravin B Shelar <pshelar@nicira.com> Signed-off-by: Jesse Gross <jesse@nicira.com>
2013-10-03tcp: shrink tcp6_timewait_sock by one cache lineEric Dumazet
While working on tcp listener refactoring, I found that it would really make things easier if sock_common could include the IPv6 addresses needed in the lookups, instead of doing very complex games to get their values (depending on sock being SYN_RECV, ESTABLISHED, TIME_WAIT) For this to happen, I need to be sure that tcp6_timewait_sock and tcp_timewait_sock consume same number of cache lines. This is possible if we only use 32bits for tw_ttd, as we remove one 32bit hole in inet_timewait_sock inet_tw_time_stamp() is defined and used, even if its current implementation looks like tcp_time_stamp : We might need finer resolution for tcp_time_stamp in the future. Before patch : sizeof(struct tcp6_timewait_sock) = 0xc8 After patch : sizeof(struct tcp6_timewait_sock) = 0xc0 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03Merge branch 'for-davem' of ↵David S. Miller
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless John W. Linville says: ==================== Here is another batch of fixes intended for the 3.12 stream... For the mac80211 bits, Johannes says: "This time I have two fixes for IBSS (including one for wext, hah), a fix for extended rates IEs, an active monitor checking fix and a sysfs registration race fix." On top of those... Amitkumar Karwar brings an mwifiex fix for an interrupt loss issue w/ SDIO devices. The problem was due to a command timeout issue introduced by an earlier patch. Felix Fietkau a stall in the ath9k driver. This patch fixes the regression introduced in the commit "ath9k: use software queues for un-aggregated data packets". Stanislaw Gruszka reverts an rt2x00 patch that was found to cause connection problems with some devices. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03Merge branch 'for-upstream' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next
2013-10-03net: heap overflow in __audit_sockaddr()Dan Carpenter
We need to cap ->msg_namelen or it leads to a buffer overflow when we to the memcpy() in __audit_sockaddr(). It requires CAP_AUDIT_CONTROL to exploit this bug. The call tree is: ___sys_recvmsg() move_addr_to_user() audit_sockaddr() __audit_sockaddr() Reported-by: Jüri Aedla <juri.aedla@gmail.com> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03Merge branch 'master' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless into for-davem
2013-10-03Merge tag 'batman-adv-fix-for-davem' of git://git.open-mesh.org/linux-mergeDavid S. Miller
Included change: - fix multi soft-interfaces setups with Network Coding enabled by registering the CODED packet type once only (instead of once per soft-if) Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03net: ipv4: Change variable type to boolPeter Senna Tschudin
The variable fully_acked is only assigned the values true and false. Change its type to bool. The simplified semantic patch that find this problem is as follows (http://coccinelle.lip6.fr/): @exists@ type T; identifier b; @@ - T + bool b = ...; ... when any b = \(true\|false\) Signed-off-by: Peter Senna Tschudin <peter.senna@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03flow_dissector: factor out the ports extraction in skb_flow_get_portsNikolay Aleksandrov
Factor out the code that extracts the ports from skb_flow_dissect and add a new function skb_flow_get_ports which can be re-used. Suggested-by: Veaceslav Falico <vfalico@redhat.com> Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Veaceslav Falico <vfalico@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03inet: consolidate INET_TW_MATCHEric Dumazet
TCP listener refactoring, part 2 : We can use a generic lookup, sockets being in whatever state, if we are sure all relevant fields are at the same place in all socket types (ESTABLISH, TIME_WAIT, SYN_RECV) This patch removes these macros : inet_addrpair, inet_addrpair, tw_addrpair, tw_portpair And adds : sk_portpair, sk_addrpair, sk_daddr, sk_rcv_saddr Then, INET_TW_MATCH() is really the same than INET_MATCH() Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03Bluetooth: Only one command per L2CAP LE signalling is supportedMarcel Holtmann
The Bluetooth specification makes it clear that only one command should be present in the L2CAP LE signalling packet. So tighten the checks here and restrict it to exactly one command. This is different from L2CAP BR/EDR signalling where multiple commands can be part of the same packet. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Check minimum length of SMP packetsMarcel Holtmann
When SMP packets are received, make sure they contain at least 1 byte header for the opcode. If not, drop the packet and disconnect the link. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Drop packets on ATT fixed channel on BR/EDRMarcel Holtmann
The ATT fixed channel is only valid when using LE connections. On BR/EDR it is required to go through L2CAP connection oriented channel for ATT. Drop ATT packets when they are received on a BR/EDR connection. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: L2CAP connectionless channels are only valid for BR/EDRMarcel Holtmann
When receiving connectionless packets on a LE connection, just drop the packet. There is no concept of connectionless channels for LE. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: SMP packets are only valid on LE connectionsMarcel Holtmann
When receiving SMP packets on a BR/EDR connection, then just drop the packet and do not try to process it. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Don't copy L2CAP LE signalling to raw socketsMarcel Holtmann
The L2CAP raw sockets are only used for BR/EDR signalling. Packets on LE links should not be forwarded there. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Fix switch statement order for L2CAP fixed channelsMarcel Holtmann
The switch statement for the various L2CAP fixed channel handlers is not really ordered. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Allow changing device class when BR/EDR is disabledMarcel Holtmann
Changing the device class when BR/EDR is disabled has no visible effect for remote devices. However to simplify the logic allow it as long as the controller supports BR/EDR operations. If it is not allowed, then the overall logic becomes rather complicated since the class of device values would need clearing or restoring when BR/EDR setting changes. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Restrict loading of long term keys to LE capable controllersMarcel Holtmann
Loading long term keys into a BR/EDR only controller make no sense. The kernel would never use any of these keys. So instead of allowing userspace to waste memory, reject such operation with a not supported error message. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Restrict loading of link keys to BR/EDR capable controllersMarcel Holtmann
Loading link keys into a LE only controller make no sense. The kernel would never use any of these keys. So instead of allowing userspace to waste memory, reject such operation with a not supported error message. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Allow setting static address even if LE is disabledMarcel Holtmann
Setting the static address does not depend on LE beeing enabled. It only depends on a controller with LE support. When depending on LE enabled this command becomes really complicated since in case LE gets disabled, it would be required to clear the static address and also its random address representation inside the controller. With future support for private addresses such complex setup should be avoided. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-03Bluetooth: Restrict SSP setting changes to BR/EDR enabled controllersMarcel Holtmann
Only when BR/EDR is supported and enabled, allow changing of the SSP setting. Just checking if the hardware supports SSP is not enough since it might be the case that BR/EDR is disabled. In the case that BR/EDR is disabled, but SSP supported by the controller the not supported error message is now returned. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
2013-10-02l2tp: fix kernel panic when using IPv4-mapped IPv6 addressesFrançois Cachereul
IPv4 mapped addresses cause kernel panic. The patch juste check whether the IPv6 address is an IPv4 mapped address. If so, use IPv4 API instead of IPv6. [ 940.026915] general protection fault: 0000 [#1] [ 940.026915] Modules linked in: l2tp_ppp l2tp_netlink l2tp_core pppox ppp_generic slhc loop psmouse [ 940.026915] CPU: 0 PID: 3184 Comm: memcheck-amd64- Not tainted 3.11.0+ #1 [ 940.026915] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007 [ 940.026915] task: ffff880007130e20 ti: ffff88000737e000 task.ti: ffff88000737e000 [ 940.026915] RIP: 0010:[<ffffffff81333780>] [<ffffffff81333780>] ip6_xmit+0x276/0x326 [ 940.026915] RSP: 0018:ffff88000737fd28 EFLAGS: 00010286 [ 940.026915] RAX: c748521a75ceff48 RBX: ffff880000c30800 RCX: 0000000000000000 [ 940.026915] RDX: ffff88000075cc4e RSI: 0000000000000028 RDI: ffff8800060e5a40 [ 940.026915] RBP: ffff8800060e5a40 R08: 0000000000000000 R09: ffff88000075cc90 [ 940.026915] R10: 0000000000000000 R11: 0000000000000000 R12: ffff88000737fda0 [ 940.026915] R13: 0000000000000000 R14: 0000000000002000 R15: ffff880005d3b580 [ 940.026915] FS: 00007f163dc5e800(0000) GS:ffffffff81623000(0000) knlGS:0000000000000000 [ 940.026915] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 940.026915] CR2: 00000004032dc940 CR3: 0000000005c25000 CR4: 00000000000006f0 [ 940.026915] Stack: [ 940.026915] ffff88000075cc4e ffffffff81694e90 ffff880000c30b38 0000000000000020 [ 940.026915] 11000000523c4bac ffff88000737fdb4 0000000000000000 ffff880000c30800 [ 940.026915] ffff880005d3b580 ffff880000c30b38 ffff8800060e5a40 0000000000000020 [ 940.026915] Call Trace: [ 940.026915] [<ffffffff81356cc3>] ? inet6_csk_xmit+0xa4/0xc4 [ 940.026915] [<ffffffffa0038535>] ? l2tp_xmit_skb+0x503/0x55a [l2tp_core] [ 940.026915] [<ffffffff812b8d3b>] ? pskb_expand_head+0x161/0x214 [ 940.026915] [<ffffffffa003e91d>] ? pppol2tp_xmit+0xf2/0x143 [l2tp_ppp] [ 940.026915] [<ffffffffa00292e0>] ? ppp_channel_push+0x36/0x8b [ppp_generic] [ 940.026915] [<ffffffffa00293fe>] ? ppp_write+0xaf/0xc5 [ppp_generic] [ 940.026915] [<ffffffff8110ead4>] ? vfs_write+0xa2/0x106 [ 940.026915] [<ffffffff8110edd6>] ? SyS_write+0x56/0x8a [ 940.026915] [<ffffffff81378ac0>] ? system_call_fastpath+0x16/0x1b [ 940.026915] Code: 00 49 8b 8f d8 00 00 00 66 83 7c 11 02 00 74 60 49 8b 47 58 48 83 e0 fe 48 8b 80 18 01 00 00 48 85 c0 74 13 48 8b 80 78 02 00 00 <48> ff 40 28 41 8b 57 68 48 01 50 30 48 8b 54 24 08 49 c7 c1 51 [ 940.026915] RIP [<ffffffff81333780>] ip6_xmit+0x276/0x326 [ 940.026915] RSP <ffff88000737fd28> [ 940.057945] ---[ end trace be8aba9a61c8b7f3 ]--- [ 940.058583] Kernel panic - not syncing: Fatal exception in interrupt Signed-off-by: François CACHEREUL <f.cachereul@alphalink.fr> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-02net: do not call sock_put() on TIMEWAIT socketsEric Dumazet
commit 3ab5aee7fe84 ("net: Convert TCP & DCCP hash tables to use RCU / hlist_nulls") incorrectly used sock_put() on TIMEWAIT sockets. We should instead use inet_twsk_put() Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-02net:drivers/net: Miscellaneous conversions to ETH_ALENJoe Perches
Convert the memset/memcpy uses of 6 to ETH_ALEN where appropriate. Also convert some struct definitions and u8 array declarations of [6] to ETH_ALEN. Signed-off-by: Joe Perches <joe@perches.com> Acked-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-02tcp: sndbuf autotuning improvementsEric Dumazet
tcp_fixup_sndbuf() is underestimating initial send buffer requirements. It was not noticed because big GSO packets were escaping the limitation, but with smaller TSO packets (or TSO/GSO/SG off), application hits sk_sndbuf before having a chance to fill enough packets in socket write queue. - initial cwnd can be bigger than 10 for specific routes - SKB_TRUESIZE() is a bit under real needs in some cases, because of power-of-two rounding in kmalloc() - Fast Recovery (RFC 5681 3.2) : Cubic needs 70% factor - Extra cushion (application might react slowly to POLLOUT) tcp_v4_conn_req_fastopen() needs to call tcp_init_metrics() before calling tcp_init_buffer_space() Then we realize tcp_new_space() should call tcp_fixup_sndbuf() instead of duplicating this stuff. Rename tcp_fixup_sndbuf() to tcp_sndbuf_expand() to be more descriptive. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Acked-by: Maciej Żenczykowski <maze@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-02fib_trie: avoid a redundant bit judgement in inflatebaker.zhang
Because 'node' is the i'st child of 'oldnode', thus, here 'i' equals tkey_extract_bits(node->key, oldtnode->pos, oldtnode->bits) we just get 1 more bit, and need not care the detail value of this bits. I apologize for the mistake. I generated the patch on a branch version, and did not notice the put_child has been changed. I have redone the test on HEAD version with my patch. two cases are used. case 1. inflate a node which has a leaf child node. case 2: inflate a node which has a an child node with skipped bits test env: ip link set eth0 up ip a add dev eth0 192.168.11.1/32 here, we just focus on route table(MAIN), so I use a "192.168.11.1/32" address to simplify the test case. call trace: + fib_insert_node + + trie_rebalance + + + resize + + + + inflate Test case 1: inflate a node which has a leaf child node. =========================================================== step 1. prepare a fib trie ------------------------------------------ ip r a 192.168.0.0/24 via 192.168.11.1 ip r a 192.168.1.0/24 via 192.168.11.1 we get a fib trie. root@baker:~# cat /proc/net/fib_trie Main: +-- 192.168.0.0/23 1 0 0 |-- 192.168.0.0 /24 universe UNICAST |-- 192.168.1.0 /24 universe UNICAST Local: ..... step 2. Add the third route ------------------------------------------ root@baker:~# ip r a 192.168.2.0/24 via 192.168.11.1 A fib_trie leaf will be inserted in fib_insert_node before trie_rebalance. For function 'inflate': 'inflate' is called with following trie. +-- 192.168.0.0/22 1 1 0 <=== tn node +-- 192.168.0.0/23 1 0 0 <== node a |-- 192.168.0.0 /24 universe UNICAST |-- 192.168.1.0 /24 universe UNICAST |-- 192.168.2.0 <== leaf(node b) When process node b, which is a leaf. here: i is 1, node key "192.168.2.0" oldnode is (pos:22, bits:1) unpatch source: tkey_extract_bits(node->key, oldtnode->pos + oldtnode->bits, 1) it equals: tkey_extract_bits("192.168,2,0", 22 + 1, 1) thus got 0, and call put_child(tn, 2*i, node); <== 2*i=2. patched source: tkey_extract_bits(node->key, oldtnode->pos, oldtnode->bits + 1), tkey_extract_bits("192.168,2,0", 22, 1 + 1) <== get 2. Test case 2: inflate a node which has a an child node with skipped bits ========================================================================== step 1. prepare a fib trie. ip link set eth0 up ip a add dev eth0 192.168.11.1/32 ip r a 192.168.128.0/24 via 192.168.11.1 ip r a 192.168.0.0/24 via 192.168.11.1 ip r a 192.168.16.0/24 via 192.168.11.1 ip r a 192.168.32.0/24 via 192.168.11.1 ip r a 192.168.48.0/24 via 192.168.11.1 ip r a 192.168.144.0/24 via 192.168.11.1 ip r a 192.168.160.0/24 via 192.168.11.1 ip r a 192.168.176.0/24 via 192.168.11.1 check: root@baker:~# cat /proc/net/fib_trie Main: +-- 192.168.0.0/16 1 0 0 +-- 192.168.0.0/18 2 0 0 |-- 192.168.0.0 /24 universe UNICAST |-- 192.168.16.0 /24 universe UNICAST |-- 192.168.32.0 /24 universe UNICAST |-- 192.168.48.0 /24 universe UNICAST +-- 192.168.128.0/18 2 0 0 |-- 192.168.128.0 /24 universe UNICAST |-- 192.168.144.0 /24 universe UNICAST |-- 192.168.160.0 /24 universe UNICAST |-- 192.168.176.0 /24 universe UNICAST Local: ... step 2. add a route to trigger inflate. ip r a 192.168.96.0/24 via 192.168.11.1 This command will call serveral times inflate. In the first time, the fib_trie is: ________________________ +-- 192.168.128.0/(16, 1) <== tn node +-- 192.168.0.0/(17, 1) <== node a +-- 192.168.0.0/(18, 2) |-- 192.168.0.0 |-- 192.168.16.0 |-- 192.168.32.0 |-- 192.168.48.0 |-- 192.168.96.0 +-- 192.168.128.0/(18, 2) <== node b. |-- 192.168.128.0 |-- 192.168.144.0 |-- 192.168.160.0 |-- 192.168.176.0 NOTE: node b is a interal node with skipped bits. here, i:1, node->key "192.168.128.0", oldnode:(pos:16, bits:1) so tkey_extract_bits(node->key, oldtnode->pos + oldtnode->bits, 1) it equals: tkey_extract_bits("192.168,128,0", 16 + 1, 1) <=== 0 tkey_extract_bits(node->key, oldtnode->pos, oldtnode->bits, 1) it equals: tkey_extract_bits("192.168,128,0", 16, 1+1) <=== 2 2*i + 0 == 2, so the result is same. Signed-off-by: baker.zhang <baker.kernel@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit (git 2.34.1) at 2025-09-13 23:21:21 +0000