firefly-linux-kernel-4.4.55.git
9 years agoe1000e: Cleanup qos request in error handling of e1000_open
Jia-Ju Bai [Thu, 4 Jun 2015 13:07:27 +0000 (21:07 +0800)]
e1000e: Cleanup qos request in error handling of e1000_open

The driver lacks pm_qos_remove_request in error handling (err_req_irq) of
e1000_open, and qos request inserted by pm_qos_add_request is not removed.
This patch add pm_qos_remove_request in error handling to fix it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@163.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
9 years agoixgbevf: add support for reporting RSS key and hash table for X550
Emil Tantilov [Thu, 30 Apr 2015 18:50:55 +0000 (11:50 -0700)]
ixgbevf: add support for reporting RSS key and hash table for X550

This patch extends the reporting of the RSS key and hash table by
adding support for X550 VFs. The difference is that X550 VFs have
their own registers for RSS key and indirection table, so there is
no need to query the PF.

The RSS key and indirection table are stored in the adapter structure
during the configuration of VFRSSRK and VFRETA which in turn can be
used in ethtool for reporting.

The logic for writing VFRETA is also changed to make sure that the
indirection table is reported correctly.

In addition this patch adds defines for the VFRETA entries and number
of VFRSSRK registers as well as some whitespace cleanups.

Reported-by: Vlad Zolotarov <vladz@cloudius-systems.com>
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
9 years agoip_tunnel: Call ip_tunnel_core_init() from inet_init()
Thomas Graf [Thu, 23 Jul 2015 08:08:44 +0000 (10:08 +0200)]
ip_tunnel: Call ip_tunnel_core_init() from inet_init()

Convert the module_init() to a invocation from inet_init() since
ip_tunnel_core is part of the INET built-in.

Fixes: 3093fbe7ff4 ("route: Per route IP tunnel metadata via lightweight tunnel")
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
David S. Miller [Thu, 23 Jul 2015 07:41:16 +0000 (00:41 -0700)]
Merge git://git./linux/kernel/git/davem/net

Conflicts:
net/bridge/br_mdb.c

br_mdb.c conflict was a function call being removed to fix a bug in
'net' but whose signature was changed in 'net-next'.

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Wed, 22 Jul 2015 21:45:25 +0000 (14:45 -0700)]
Merge git://git./linux/kernel/git/davem/net

Pull networking fixes from David Miller:

 1) Don't use shared bluetooth antenna in iwlwifi driver for management
    frames, from Emmanuel Grumbach.

 2) Fix device ID check in ath9k driver, from Felix Fietkau.

 3) Off by one in xen-netback BUG checks, from Dan Carpenter.

 4) Fix IFLA_VF_PORT netlink attribute validation, from Daniel Borkmann.

 5) Fix races in setting peeked bit flag in SKBs during datagram
    receive.  If it's shared we have to clone it otherwise the value can
    easily be corrupted.  Fix from Herbert Xu.

 6) Revert fec clock handling change, causes regressions.  From Fabio
    Estevam.

 7) Fix use after free in fq_codel and sfq packet schedulers, from WANG
    Cong.

 8) ipvlan bug fixes (memory leaks, missing rcu_dereference_bh, etc.)
    from WANG Cong and Konstantin Khlebnikov.

 9) Memory leak in act_bpf packet action, from Alexei Starovoitov.

10) ARM bpf JIT bug fixes from Nicolas Schichan.

11) Fix backwards compat of ANY_LAYOUT in virtio_net driver, from
    Michael S Tsirkin.

12) Destruction of bond with different ARP header types not handled
    correctly, fix from Nikolay Aleksandrov.

13) Revert GRO receive support in ipv6 SIT tunnel driver, causes
    regressions because the GRO packets created cannot be processed
    properly on the GSO side if we forward the frame.  From Herbert Xu.

14) TCCR update race and other fixes to ravb driver from Sergei
    Shtylyov.

15) Fix SKB leaks in caif_queue_rcv_skb(), from Eric Dumazet.

16) Fix panics on packet scheduler filter replace, from Daniel Borkmann.

17) Make sure AF_PACKET sees properly IP headers in defragmented frames
    (via PACKET_FANOUT_FLAG_DEFRAG option), from Edward Hyunkoo Jee.

18) AF_NETLINK cannot hold mutex in RCU callback, fix from Florian
    Westphal.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (84 commits)
  ravb: fix ring memory allocation
  net: phy: dp83867: Fix warning check for setting the internal delay
  openvswitch: allocate nr_node_ids flow_stats instead of num_possible_nodes
  netlink: don't hold mutex in rcu callback when releasing mmapd ring
  ARM: net: fix vlan access instructions in ARM JIT.
  ARM: net: handle negative offsets in BPF JIT.
  ARM: net: fix condition for load_order > 0 when translating load instructions.
  tcp: suppress a division by zero warning
  drivers: net: cpsw: remove tx event processing in rx napi poll
  inet: frags: fix defragmented packet's IP header for af_packet
  net: mvneta: fix refilling for Rx DMA buffers
  stmmac: fix setting of driver data in stmmac_dvr_probe
  sched: cls_flow: fix panic on filter replace
  sched: cls_flower: fix panic on filter replace
  sched: cls_bpf: fix panic on filter replace
  net/mdio: fix mdio_bus_match for c45 PHY
  net: ratelimit warnings about dst entry refcount underflow or overflow
  caif: fix leaks and race in caif_queue_rcv_skb()
  qmi_wwan: add the second QMI/network interface for Sierra Wireless MC7305/MC7355
  ravb: fix race updating TCCR
  ...

9 years agoip_tunnel: Provide tunnel metadata API for CONFIG_INET=n
Thomas Graf [Wed, 22 Jul 2015 12:43:58 +0000 (14:43 +0200)]
ip_tunnel: Provide tunnel metadata API for CONFIG_INET=n

Account for the configuration FIB_RULES=y && INET=n as FIB_RULES can
be selected by IPV6 or DECNET without INET.

Fixes: e7030878fc84 ("fib: Add fib rule match on tunnel id")
Fixes: 3093fbe7ff4b ("route: Per route IP tunnel metadata via lightweight tunnel")
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoipv6: sysctl to restrict candidate source addresses
Erik Kline [Wed, 22 Jul 2015 07:38:25 +0000 (16:38 +0900)]
ipv6: sysctl to restrict candidate source addresses

Per RFC 6724, section 4, "Candidate Source Addresses":

    It is RECOMMENDED that the candidate source addresses be the set
    of unicast addresses assigned to the interface that will be used
    to send to the destination (the "outgoing" interface).

Add a sysctl to enable this behaviour.

Signed-off-by: Erik Kline <ek@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agor8152: support the new RTL8153 chip
hayeswang [Wed, 22 Jul 2015 07:27:41 +0000 (15:27 +0800)]
r8152: support the new RTL8153 chip

Support the new USB gigabit ethernet.

Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agompls_iptunnel: fix sparse warn: remove incorrect rcu_dereference
Roopa Prabhu [Wed, 22 Jul 2015 05:49:00 +0000 (22:49 -0700)]
mpls_iptunnel: fix sparse warn: remove incorrect rcu_dereference

fix for:
net/mpls/mpls_iptunnel.c:73:19: sparse: incompatible types in comparison
expression (different address spaces)

remove incorrect rcu_dereference possibly left over from
earlier revisions of the code.

Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'bnx2x-next'
David S. Miller [Wed, 22 Jul 2015 17:47:27 +0000 (10:47 -0700)]
Merge branch 'bnx2x-next'

Yuval Mintz says:

====================
bnx2x: update FW, rebrand and more

This patch series does several things - it updates the bnx2x FW into
7.12.30 which both contains some small fixes as well as opening the door
for several new features for the device - mainly vxlan/geneve offloads
and vlan filtering offload.
It then adds a new Multi-function mode [BD] which requires this FW in
order to operate.

In addition, this finally rebrands the driver from a 'broadcom' driver
into a 'qlogic' driver [although it would still reside under Broadcom's
tree in the kernel].
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agobnx2x: Bump up driver version to 1.712.30
Yuval Mintz [Wed, 22 Jul 2015 06:16:27 +0000 (09:16 +0300)]
bnx2x: Bump up driver version to 1.712.30

Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agobnx2x: Add MFW dump support
Yuval Mintz [Wed, 22 Jul 2015 06:16:26 +0000 (09:16 +0300)]
bnx2x: Add MFW dump support

Devices with up-to-date management FW will be able to store register dumps
on their persistent storage - in case management FW identifies a fatal
error it would gather and store such dumps, which could later be retrieved
using specific debug tools.

This patch adds the necessary part in the driver in order to make the
feature operational, as well as update users [under debug] during load
in case their device contains a dump of a previous crash.

Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agobnx2x: new Multi-function mode - BD
Yuval Mintz [Wed, 22 Jul 2015 06:16:25 +0000 (09:16 +0300)]
bnx2x: new Multi-function mode - BD

This adds support to a new multi-function mode, enabling driver to
initialize such devices and correctly interacting with management FW
for fully utilizing their features.

Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agobnx2x: Add 84858 phy support
Yaniv Rosner [Wed, 22 Jul 2015 06:16:24 +0000 (09:16 +0300)]
bnx2x: Add 84858 phy support

This adds support to a new copper phy.

Signed-off-by: Yaniv Rosner <Yaniv.Rosner@qlogic.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agobnx2x: Rebrand from 'broadcom' into 'qlogic'
Yuval Mintz [Wed, 22 Jul 2015 06:16:23 +0000 (09:16 +0300)]
bnx2x: Rebrand from 'broadcom' into 'qlogic'

bnx2x still appears as a Broadcom driver even though the devices it
utilizes belong to Qlogic for more than a year.

This patch changes the various headers and the device strings to indicate
the correct ownership of the device.

Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agobnx2x: Utilize FW 7.12.30
Yuval Mintz [Wed, 22 Jul 2015 06:16:22 +0000 (09:16 +0300)]
bnx2x: Utilize FW 7.12.30

This moves bnx2x into using 7.12.30 FW. Said firmware fixes the following:

 - Packets from a VF with pvid configured which were sent with a
   different vlan were transmitted instead of being discarded.

 - FCoE traffic might not recover after a failue while there's traffic
   to another function.

In addition, this FW opens the door for the driver to implement several
new features; Specifically, this enhances the device's support for
encapsulated packets and will allow vxlan/geneve offloads to be added in
the future, as well as vlan filtering offload.

Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Linus Torvalds [Wed, 22 Jul 2015 15:52:42 +0000 (08:52 -0700)]
Merge tag 'arm64-fixes' of git://git./linux/kernel/git/arm64/linux

Pull ARM64 fixes from Catalin Marinas:

 - arm64 build fix following the move of the thread_struct to the end of
   task_struct and the asm offsets becoming too large for the AArch64
   ISA

 - preparatory patch for moving irq_data struct members (applied now to
   reduce dependency for the next merging window)

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  ARM64/irq: Use access helper irq_data_get_affinity_mask()
  arm64: switch_to: calculate cpu context pointer using separate register

9 years agoARM64/irq: Use access helper irq_data_get_affinity_mask()
Jiang Liu [Mon, 13 Jul 2015 20:30:04 +0000 (20:30 +0000)]
ARM64/irq: Use access helper irq_data_get_affinity_mask()

This is a preparatory patch for moving irq_data struct members.

Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com>
Reviewed-by: Hanjun Guo <hanjun.guo@linaro.org>
Cc: linux-arm-kernel@lists.infradead.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: switch_to: calculate cpu context pointer using separate register
Will Deacon [Mon, 20 Jul 2015 14:14:53 +0000 (15:14 +0100)]
arm64: switch_to: calculate cpu context pointer using separate register

Commit 0c8c0f03e3a2 ("x86/fpu, sched: Dynamically allocate 'struct fpu'")
moved the thread_struct to the bottom of task_struct. As a result, the
offset is now too large to be used in an immediate add on arm64 with
some kernel configs:

arch/arm64/kernel/entry.S: Assembler messages:
arch/arm64/kernel/entry.S:588: Error: immediate out of range
arch/arm64/kernel/entry.S:597: Error: immediate out of range

This patch calculates the offset using an additional register instead of
an immediate offset.

Fixes: 0c8c0f03e3a2 ("x86/fpu, sched: Dynamically allocate 'struct fpu'")
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Olof Johansson <olof@lixom.net>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agonet: track success and failure of TCP PMTU probing
Rick Jones [Tue, 21 Jul 2015 23:14:13 +0000 (16:14 -0700)]
net: track success and failure of TCP PMTU probing

Track success and failure of TCP PMTU probing.

Signed-off-by: Rick Jones <rick.jones2@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoravb: fix ring memory allocation
Sergei Shtylyov [Tue, 21 Jul 2015 22:31:59 +0000 (01:31 +0300)]
ravb: fix ring memory allocation

The driver is written as if it can adapt to a low memory situation  allocating
less RX  skbs and TX aligned buffers than the respective RX/TX ring sizes.  In
reality  though  the driver  would malfunction in this case. Stop being overly
smart and just fail in such situation -- this is achieved by moving the memory
allocation from ravb_ring_format() to ravb_ring_init().

We leave dma_map_single() calls in place but make their failure non-fatal
by marking the corresponding RX descriptors  with zero data size which should
prevent DMA to an invalid addresses.

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agocxgb4: Add debugfs entry to enable backdoor access
Hariprasad Shenai [Tue, 21 Jul 2015 17:09:40 +0000 (22:39 +0530)]
cxgb4: Add debugfs entry to enable backdoor access

Add debugfs entry 'use_backdoor' to enable backdoor access to read sge
context. By default, we read sge context's via firmware. In case of FW
issues, one can enable backdoor access via debugfs to dump sge context
for debugging purpose.

Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet: phy: dp83867: Fix warning check for setting the internal delay
Dan Murphy [Tue, 21 Jul 2015 17:06:45 +0000 (12:06 -0500)]
net: phy: dp83867: Fix warning check for setting the internal delay

Fix warning: logical ‘or’ of collectively exhaustive tests is always true

Change the internal delay check from an 'or' condition to an 'and'
condition.

Reported-by: David Binderman <dcb314@hotmail.com>
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agompls: make RTA_OIF optional
Roopa Prabhu [Tue, 21 Jul 2015 16:16:24 +0000 (09:16 -0700)]
mpls: make RTA_OIF optional

If user did not specify an oif, try and get it from the via address.
If failed to get device, return with -ENODEV.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoopenvswitch: allocate nr_node_ids flow_stats instead of num_possible_nodes
Chris J Arges [Tue, 21 Jul 2015 17:36:33 +0000 (12:36 -0500)]
openvswitch: allocate nr_node_ids flow_stats instead of num_possible_nodes

Some architectures like POWER can have a NUMA node_possible_map that
contains sparse entries. This causes memory corruption with openvswitch
since it allocates flow_cache with a multiple of num_possible_nodes() and
assumes the node variable returned by for_each_node will index into
flow->stats[node].

Use nr_node_ids to allocate a maximal sparse array instead of
num_possible_nodes().

The crash was noticed after 3af229f2 was applied as it changed the
node_possible_map to match node_online_map on boot.
Fixes: 3af229f2071f5b5cb31664be6109561fbe19c861
Signed-off-by: Chris J Arges <chris.j.arges@canonical.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonetlink: don't hold mutex in rcu callback when releasing mmapd ring
Florian Westphal [Tue, 21 Jul 2015 14:33:50 +0000 (16:33 +0200)]
netlink: don't hold mutex in rcu callback when releasing mmapd ring

Kirill A. Shutemov says:

This simple test-case trigers few locking asserts in kernel:

int main(int argc, char **argv)
{
        unsigned int block_size = 16 * 4096;
        struct nl_mmap_req req = {
                .nm_block_size          = block_size,
                .nm_block_nr            = 64,
                .nm_frame_size          = 16384,
                .nm_frame_nr            = 64 * block_size / 16384,
        };
        unsigned int ring_size;
int fd;

fd = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC);
        if (setsockopt(fd, SOL_NETLINK, NETLINK_RX_RING, &req, sizeof(req)) < 0)
                exit(1);
        if (setsockopt(fd, SOL_NETLINK, NETLINK_TX_RING, &req, sizeof(req)) < 0)
                exit(1);

ring_size = req.nm_block_nr * req.nm_block_size;
mmap(NULL, 2 * ring_size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);
return 0;
}

+++ exited with 0 +++
BUG: sleeping function called from invalid context at /home/kas/git/public/linux-mm/kernel/locking/mutex.c:616
in_atomic(): 1, irqs_disabled(): 0, pid: 1, name: init
3 locks held by init/1:
 #0:  (reboot_mutex){+.+...}, at: [<ffffffff81080959>] SyS_reboot+0xa9/0x220
 #1:  ((reboot_notifier_list).rwsem){.+.+..}, at: [<ffffffff8107f379>] __blocking_notifier_call_chain+0x39/0x70
 #2:  (rcu_callback){......}, at: [<ffffffff810d32e0>] rcu_do_batch.isra.49+0x160/0x10c0
Preemption disabled at:[<ffffffff8145365f>] __delay+0xf/0x20

CPU: 1 PID: 1 Comm: init Not tainted 4.1.0-00009-gbddf4c4818e0 #253
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS Debian-1.8.2-1 04/01/2014
 ffff88017b3d8000 ffff88027bc03c38 ffffffff81929ceb 0000000000000102
 0000000000000000 ffff88027bc03c68 ffffffff81085a9d 0000000000000002
 ffffffff81ca2a20 0000000000000268 0000000000000000 ffff88027bc03c98
Call Trace:
 <IRQ>  [<ffffffff81929ceb>] dump_stack+0x4f/0x7b
 [<ffffffff81085a9d>] ___might_sleep+0x16d/0x270
 [<ffffffff81085bed>] __might_sleep+0x4d/0x90
 [<ffffffff8192e96f>] mutex_lock_nested+0x2f/0x430
 [<ffffffff81932fed>] ? _raw_spin_unlock_irqrestore+0x5d/0x80
 [<ffffffff81464143>] ? __this_cpu_preempt_check+0x13/0x20
 [<ffffffff8182fc3d>] netlink_set_ring+0x1ed/0x350
 [<ffffffff8182e000>] ? netlink_undo_bind+0x70/0x70
 [<ffffffff8182fe20>] netlink_sock_destruct+0x80/0x150
 [<ffffffff817e484d>] __sk_free+0x1d/0x160
 [<ffffffff817e49a9>] sk_free+0x19/0x20
[..]

Cong Wang says:

We can't hold mutex lock in a rcu callback, [..]

Thomas Graf says:

The socket should be dead at this point. It might be simpler to
add a netlink_release_ring() function which doesn't require
locking at all.

Reported-by: "Kirill A. Shutemov" <kirill@shutemov.name>
Diagnosed-by: Cong Wang <cwang@twopensource.com>
Suggested-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'sfc-filter-chaining'
David S. Miller [Wed, 22 Jul 2015 05:21:33 +0000 (22:21 -0700)]
Merge branch 'sfc-filter-chaining'

Edward Cree says:

====================
sfc: support for cascaded multicast filtering

Recent versions of firmware for SFC9100 adapters add support for filter
 chaining, in which packets matching multiple filters are delivered to all
 filters' recipients, rather than only the highest match-priority filter as was
 previously the case.
This patch series enables this feature and redesigns the filter handling code
 to make use of it; in particular, subscribing to a multicast address on one
 function no longer prevents traffic to that address reaching another function
 which is in promiscuous or allmulti mode.
If the firmware does not support filter chaining, the driver will fall back to
 the old behaviour.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: clean fallbacks between promisc/normal in efx_ef10_filter_sync_rx_mode
Edward Cree [Tue, 21 Jul 2015 14:11:00 +0000 (15:11 +0100)]
sfc: clean fallbacks between promisc/normal in efx_ef10_filter_sync_rx_mode

Separate functions for inserting individual and promisc filters; explicit
 fallback logic in efx_ef10_filter_sync_rx_mode(), in order not to overload
 the 'promisc' flag as also meaning "fall back to promisc".

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: support cascaded multicast filters
Daniel Pieczko [Tue, 21 Jul 2015 14:10:44 +0000 (15:10 +0100)]
sfc: support cascaded multicast filters

If the workaround to support cascaded multicast filters ("workaround_26807") is
enabled, the broadcast filter and individual multicast filters are not inserted
when in promiscuous or allmulti mode.

There is a race while inserting and removing filters when entering and leaving
promiscuous mode.  When changing promiscuous state with cascaded multicast
filters, the old multicast filters are removed before inserting the new filters
to avoid duplicating packets; this can lead to dropped packets until all
filters have been inserted.

The efx_nic:mc_promisc flag is added to record the presence of a multicast
promiscuous filter; this gives a simple way to tell if the promiscuous state is
changing.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: re-factor efx_ef10_filter_sync_rx_mode()
Daniel Pieczko [Tue, 21 Jul 2015 14:10:27 +0000 (15:10 +0100)]
sfc: re-factor efx_ef10_filter_sync_rx_mode()

This change is only re-factoring; there are no changes to functionality
 except for a slight elaboration of an error message (on mismatch filter
 insertion failure).

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: Insert multicast filters as well as mismatch filters in promiscuous mode
Jon Cooper [Tue, 21 Jul 2015 14:10:15 +0000 (15:10 +0100)]
sfc: Insert multicast filters as well as mismatch filters in promiscuous mode

If a function is in promiscuous mode and another function has a broadcast or
 multicast filter inserted, the function in promiscuous mode won't see that
 broadcast or multicast traffic.
Most notably this breaks broadcast, which means ARP doesn't work. Less
 show-stoppingly, a function listening on a multicast address that's also in
 promiscuous mode will not see that multicast traffic if another function is
 also listening on that multicast address.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: warn if other functions have been reset by MCFW
Daniel Pieczko [Tue, 21 Jul 2015 14:10:02 +0000 (15:10 +0100)]
sfc: warn if other functions have been reset by MCFW

When enabling the workaround for cascaded multicast filters, the MC
 can reset other functions if they have already inserted filters.
 In that case, the workaround has been enabled, but print an info
 message in the log recording that other functions had to be reset.

As other functions were reset, the MC will have incremented its boot
 count, so also increment the warm_boot_count on the function which
 enabled the workaround, as that function won't have received an MC
 reboot event and does not need to reset.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: add output flag decoding to efx_mcdi_set_workaround
Daniel Pieczko [Tue, 21 Jul 2015 14:09:43 +0000 (15:09 +0100)]
sfc: add output flag decoding to efx_mcdi_set_workaround

The initial use of this will be to check a flag reporting if an FLR was
performed on other functions when enabling cascaded multicast filters.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: cope with ENOSYS from efx_mcdi_get_workarounds()
Edward Cree [Tue, 21 Jul 2015 14:09:31 +0000 (15:09 +0100)]
sfc: cope with ENOSYS from efx_mcdi_get_workarounds()

GET_WORKAROUNDS was only introduced in May 2014, not all firmware
 will have it.  So call sites need to handle ENOSYS.
In this case we're probing the bug26807 workaround, which is not
 implemented in any firmware that doesn't have GET_WORKAROUNDS.
 So interpret ENOSYS as 'false'.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: enable cascaded multicast filters in MCFW
Daniel Pieczko [Tue, 21 Jul 2015 14:09:18 +0000 (15:09 +0100)]
sfc: enable cascaded multicast filters in MCFW

After creating event queue 0, check to see if the workaround is enabled,
 and enable it if necessary.  This will be called during PCI probe and
 also when coming back up after a reset.  The nic_data->workaround_26807
 will be used in the future to control the filter insertion behaviour
 based on this workaround.

Only the primary PF can enable this workaround, so tolerate an EPERM
 error and continue.  Otherwise, if any step in the checking and enabling
 of the workaround fails, the event queue must be removed.

We check that workaround is implemented before trying to enable it,
 and store the current workaround setting before trying to change it.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosfc: update MCDI protocol definitions
Edward Cree [Tue, 21 Jul 2015 14:08:56 +0000 (15:08 +0100)]
sfc: update MCDI protocol definitions

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'arm-bpf-fixes'
David S. Miller [Wed, 22 Jul 2015 05:19:55 +0000 (22:19 -0700)]
Merge branch 'arm-bpf-fixes'

Nicolas Schichan says:

====================
BPF JIT fixes for ARM

These patches are fixing bugs in the ARM JIT and should probably find
their way to a stable kernel. All 60 test_bpf tests in Linux 4.1 release
are now passing OK (was 54 out of 60 before).
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoARM: net: fix vlan access instructions in ARM JIT.
Nicolas Schichan [Tue, 21 Jul 2015 12:14:14 +0000 (14:14 +0200)]
ARM: net: fix vlan access instructions in ARM JIT.

This makes BPF_ANC | SKF_AD_VLAN_TAG and BPF_ANC | SKF_AD_VLAN_TAG_PRESENT
have the same behaviour as the in kernel VM and makes the test_bpf LD_VLAN_TAG
and LD_VLAN_TAG_PRESENT tests pass.

Signed-off-by: Nicolas Schichan <nschichan@freebox.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoARM: net: handle negative offsets in BPF JIT.
Nicolas Schichan [Tue, 21 Jul 2015 12:14:13 +0000 (14:14 +0200)]
ARM: net: handle negative offsets in BPF JIT.

Previously, the JIT would reject negative offsets known during code
generation and mishandle negative offsets provided at runtime.

Fix that by calling bpf_internal_load_pointer_neg_helper()
appropriately in the jit_get_skb_{b,h,w} slow path helpers and by forcing
the execution flow to the slow path helpers when the offset is
negative.

Signed-off-by: Nicolas Schichan <nschichan@freebox.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoARM: net: fix condition for load_order > 0 when translating load instructions.
Nicolas Schichan [Tue, 21 Jul 2015 12:14:12 +0000 (14:14 +0200)]
ARM: net: fix condition for load_order > 0 when translating load instructions.

To check whether the load should take the fast path or not, the code
would check that (r_skb_hlen - load_order) is greater than the offset
of the access using an "Unsigned higher or same" condition. For
halfword accesses and an skb length of 1 at offset 0, that test is
valid, as we end up comparing 0xffffffff(-1) and 0, so the fast path
is taken and the filter allows the load to wrongly succeed. A similar
issue exists for word loads at offset 0 and an skb length of less than
4.

Fix that by using the condition "Signed greater than or equal"
condition for the fast path code for load orders greater than 0.

Signed-off-by: Nicolas Schichan <nschichan@freebox.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agotcp: suppress a division by zero warning
Eric Dumazet [Wed, 22 Jul 2015 05:02:00 +0000 (07:02 +0200)]
tcp: suppress a division by zero warning

Andrew Morton reported following warning on one ARM build
with gcc-4.4 :

net/ipv4/inet_hashtables.c: In function 'inet_ehash_locks_alloc':
net/ipv4/inet_hashtables.c:617: warning: division by zero

Even guarded with a test on sizeof(spinlock_t), compiler does not
like current construct on a !CONFIG_SMP build.

Remove the warning by using a temporary variable.

Fixes: 095dc8e0c368 ("tcp: fix/cleanup inet_ehash_locks_alloc()")
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agotipc: fix compatibility bug
Jon Paul Maloy [Tue, 21 Jul 2015 10:42:28 +0000 (06:42 -0400)]
tipc: fix compatibility bug

In commit d999297c3dbbe7fdd832f7fa4ec84301e170b3e6
("tipc: reduce locking scope during packet reception") we introduced
a new function tipc_link_proto_rcv(). This function contains a bug,
so that it sometimes by error sends out a non-zero link priority value
in created protocol messages.

The bug may lead to an extra link reset at initial link establising
with older nodes. This will never happen more than once, whereafter
the link will work as intended.

We fix this bug in this commit.

Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'explicit-inbound-link-state'
David S. Miller [Tue, 21 Jul 2015 23:12:56 +0000 (16:12 -0700)]
Merge branch 'explicit-inbound-link-state'

Florian Fainelli says:

====================
net: enable inband link state negotiation only when explicitly requested

Changes in v5:

- removed an invalid use of the link_update callback in the SF2 driver
  was appeared after merging "net: phy: fixed_phy: handle link-down case"

- reworded the commit message for patch 2 to make it clear what it fixes and
  why this is required

Initial cover letter from Stas:

Hello.

Currently the link status auto-negotiation is enabled
for any SGMII link with fixed-link DT binding.
The regression was reported:
https://lkml.org/lkml/2015/7/8/865
Apparently not all HW that implements SGMII protocol, generates the
inband status for the auto-negotiation to work.
More details here:
https://lkml.org/lkml/2015/7/10/206

The following patches reverts to the old behavior by default,
which is to not enable the auto-negotiation for fixed-link.
The new DT property is added that allows to explicitly request
the auto-negotiation.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agomvneta: use inband status only when explicitly enabled
Stas Sergeev [Tue, 21 Jul 2015 00:49:58 +0000 (17:49 -0700)]
mvneta: use inband status only when explicitly enabled

The commit 898b2970e2c9 ("mvneta: implement SGMII-based in-band link state
signaling") implemented the link parameters auto-negotiation unconditionally.
Unfortunately it appears that some HW that implements SGMII protocol,
doesn't generate the inband status, so it is not possible to auto-negotiate
anything with such HW.

This patch enables the auto-negotiation only if explicitly requested with
the 'managed' DT property.

This patch fixes the following regression:
https://lkml.org/lkml/2015/7/8/865

Signed-off-by: Stas Sergeev <stsp@users.sourceforge.net>
CC: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
CC: netdev@vger.kernel.org
CC: linux-kernel@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoof_mdio: add new DT property 'managed' to specify the PHY management type
Stas Sergeev [Tue, 21 Jul 2015 00:49:57 +0000 (17:49 -0700)]
of_mdio: add new DT property 'managed' to specify the PHY management type

Currently the PHY management type is selected by the MAC driver arbitrary.
The decision is based on the presence of the "fixed-link" node and on a
will of the driver's authors.
This caused a regression recently, when mvneta driver suddenly started
to use the in-band status for auto-negotiation on fixed links.
It appears the auto-negotiation may not work when expected by the MAC driver.
Sebastien Rannou explains:
<< Yes, I confirm that my HW does not generate an in-band status. AFAIK, it's
a PHY that aggregates 4xSGMIIs to 1xQSGMII ; the MAC side of the PHY (with
inband status) is connected to the switch through QSGMII, and in this context
we are on the media side of the PHY. >>
https://lkml.org/lkml/2015/7/10/206

This patch introduces the new string property 'managed' that allows
the user to set the management type explicitly.
The supported values are:
"auto" - default. Uses either MDIO or nothing, depending on the presence
of the fixed-link node
"in-band-status" - use in-band status

Signed-off-by: Stas Sergeev <stsp@users.sourceforge.net>
CC: Rob Herring <robh+dt@kernel.org>
CC: Pawel Moll <pawel.moll@arm.com>
CC: Mark Rutland <mark.rutland@arm.com>
CC: Ian Campbell <ijc+devicetree@hellion.org.uk>
CC: Kumar Gala <galak@codeaurora.org>
CC: Florian Fainelli <f.fainelli@gmail.com>
CC: Grant Likely <grant.likely@linaro.org>
CC: devicetree@vger.kernel.org
CC: linux-kernel@vger.kernel.org
CC: netdev@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet: phy: fixed_phy: handle link-down case
Stas Sergeev [Tue, 21 Jul 2015 00:49:56 +0000 (17:49 -0700)]
net: phy: fixed_phy: handle link-down case

fixed_phy_register() currently hardcodes the fixed PHY link to 1, and
expects to find a "speed" parameter to provide correct information
towards the fixed PHY consumer.

In a subsequent change, where we allow "managed" (e.g: (RS)GMII in-band
status auto-negotiation) fixed PHYs, none of these parameters can be
provided since they will be auto-negotiated, hence, we just provide a
zero-initialized fixed_phy_status to fixed_phy_register() which makes it
fail when we call fixed_phy_update_regs() since status.speed = 0 which
makes us hit the "default" label and error out.

Without this change, we would also see potentially inconsistent
speed/duplex parameters for fixed PHYs when the link is DOWN.

CC: netdev@vger.kernel.org
CC: linux-kernel@vger.kernel.org
Signed-off-by: Stas Sergeev <stsp@users.sourceforge.net>
[florian: add more background to why this is correct and desirable]
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet: dsa: bcm_sf2: Do not override speed settings
Florian Fainelli [Tue, 21 Jul 2015 00:49:55 +0000 (17:49 -0700)]
net: dsa: bcm_sf2: Do not override speed settings

The SF2 driver currently overrides speed settings for its port
configured using a fixed PHY, this is both unnecessary and incorrect,
because we keep feedback to the hardware parameters that we read from
the PHY device, which in the case of a fixed PHY cannot possibly change
speed.

This is a required change to allow the fixed PHY code to allow
registering a PHY with a link configured as DOWN by default and avoid
some sort of circular dependency where we require the link_update
callback to run to program the hardware, and we then utilize the fixed
PHY parameters to program the hardware with the same settings.

Fixes: 246d7f773c13 ("net: dsa: add Broadcom SF2 switch driver")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoRevert "fsnotify: fix oops in fsnotify_clear_marks_by_group_flags()"
Linus Torvalds [Tue, 21 Jul 2015 23:06:53 +0000 (16:06 -0700)]
Revert "fsnotify: fix oops in fsnotify_clear_marks_by_group_flags()"

This reverts commit a2673b6e040663bf16a552f8619e6bde9f4b9acf.

Kinglong Mee reports a memory leak with that patch, and Jan Kara confirms:

 "Thanks for report! You are right that my patch introduces a race
  between fsnotify kthread and fsnotify_destroy_group() which can result
  in leaking inotify event on group destruction.

  I haven't yet decided whether the right fix is not to queue events for
  dying notification group (as that is pointless anyway) or whether we
  should just fix the original problem differently...  Whenever I look
  at fsnotify code mark handling I get lost in the maze of locks, lists,
  and subtle differences between how different notification systems
  handle notification marks :( I'll think about it over night"

and after thinking about it, Jan says:

 "OK, I have looked into the code some more and I found another
  relatively simple way of fixing the original oops.  It will be IMHO
  better than trying to fixup this issue which has more potential for
  breakage.  I'll ask Linus to revert the fsnotify fix he already merged
  and send a new fix"

Reported-by: Kinglong Mee <kinglongmee@gmail.com>
Requested-by: Jan Kara <jack@suse.cz>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoMerge tag 'wireless-drivers-for-davem-2015-07-20' of git://git.kernel.org/pub/scm...
David S. Miller [Tue, 21 Jul 2015 23:06:39 +0000 (16:06 -0700)]
Merge tag 'wireless-drivers-for-davem-2015-07-20' of git://git./linux/kernel/git/kvalo/wireless-drivers

Kalle Valo says:

====================
ath9k:

* fix device ID check for AR956x

iwlwifi:

* bug fixes specific for 8000 series
* fix a crash in time events
* fix a crash in PCIe transport
* fix BT Coex code that prevented association on certain
  devices (3160).
* revert the new RBD allocation model because it introduced
  a bug when running on weak VM setups.
* new device IDs
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet: #ifdefify sk_classid member of struct sock
Mathias Krause [Sun, 19 Jul 2015 20:21:13 +0000 (22:21 +0200)]
net: #ifdefify sk_classid member of struct sock

The sk_classid member is only required when CONFIG_CGROUP_NET_CLASSID is
enabled. #ifdefify it to reduce the size of struct sock on 32 bit
systems, at least.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge tag 'pinctrl-v4.2-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw...
Linus Torvalds [Tue, 21 Jul 2015 22:27:27 +0000 (15:27 -0700)]
Merge tag 'pinctrl-v4.2-2' of git://git./linux/kernel/git/linusw/linux-pinctrl

Pull pin control fixes from Linus Walleij:
 "Here are some overly ripe pin control fixes for the v4.2 series.

  They got delayed because of various crap commits and having to clean
  and rinse the patch stack a few times.  Now they are however looking
  good.

   - some dead defines dropped from the Samsung driver, was targeted for
     -rc2 but got delayed
   - drop the strict mode from abx500, this was too strict
   - fix the R-Car sparse IRQs code to work as intended
   - fix the IRQ code for the pinctrl-single GPIO backend to not enforce
     threaded IRQs
   - clear the latched events/IRQs for the Broadcom BCM2835 driver
   - fix up debugfs for the Freescale imx1 driver
   - fix a typo bug in the Schmitt Trigger setup in the LPC18xx driver"

* tag 'pinctrl-v4.2-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
  pinctrl: lpc18xx: fix schmitt trigger setup
  Subject: pinctrl: imx1-core: Fix debug output in .pin_config_set callback
  pinctrl: bcm2835: Clear the event latch register when disabling interrupts
  pinctrl: single: ensure pcs irq will not be forced threaded
  sh-pfc: fix sparse GPIOs for R-Car SoCs
  pinctrl: abx500: remove strict mode
  pinctrl: samsung: Remove old unused defines

9 years agoMerge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs
Linus Torvalds [Tue, 21 Jul 2015 22:18:06 +0000 (15:18 -0700)]
Merge branch 'for_linus' of git://git./linux/kernel/git/jack/linux-fs

Pull UDF fix from Jan Kara:
 "A fix for UDF corruption when certain disk-format feature is enabled"

* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  udf: Don't corrupt unalloc spacetable when writing it

9 years agoMerge tag 'trace-v4.2-rc2-fix2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 21 Jul 2015 21:42:40 +0000 (14:42 -0700)]
Merge tag 'trace-v4.2-rc2-fix2' of git://git./linux/kernel/git/rostedt/linux-trace

Pull tracing sample code fix from Steven Rostedt:
 "He Kuang noticed that the sample code using the trace_event helper
  function __get_dynamic_array_len() is broken.

  This only changes the sample code, and I'm pushing this now instead of
  later because I don't want others using the broken code as an example
  when using it for real"

* tag 'trace-v4.2-rc2-fix2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
  tracing: Fix sample output of dynamic arrays

9 years agodrivers: net: cpsw: remove tx event processing in rx napi poll
Mugunthan V N [Tue, 21 Jul 2015 10:30:42 +0000 (16:00 +0530)]
drivers: net: cpsw: remove tx event processing in rx napi poll

With commit c03abd84634d ("net: ethernet: cpsw: don't requests IRQs
we don't use") common isr and napi are separated into separate tx isr
and rx isr/napi, but still in rx napi tx events are handled. So removing
the tx event handling in rx napi.

Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'lwtunnel'
David S. Miller [Tue, 21 Jul 2015 17:39:07 +0000 (10:39 -0700)]
Merge branch 'lwtunnel'

Thomas Graf says:

====================
Lightweight & flow based encapsulation

This series combines the work previously posted by Roopa, Robert and
myself. It's according to what we discussed at NFWS. The motivation
of this series is to:

 * Consolidate code between OVS and the rest of the kernel and get
   rid of OVS vports and instead represent them as pure net_devices.
 * Introduce a lightweight tunneling mechanism which enables flow
   based encapsulation to improve scalability on both RX and TX.
 * Do the above in an encapsulation unspecific way so that the
   encapsulation type is eventually abstracted away from the user.
 * Use the same forwarding decision for both native forwarding and
   encapsulation thus allowing to switch between native IPv6 and
   UDP encapsulation based on endpoint without requiring additional
   logic

The fundamental changes introduces in this series are:
 * A new RTA_ENCAP Netlink attribute for routes carrying encapsulation
   instructions. Depending on the specified type, the instructions
   apply to UDP encapsulations, MPLS and possible other in the future.
 * Depending on the encapsulation type, the output function of the
   dst is directly overwritten or the dst merely attaches metadata and
   relies on a subsequent net_device to apply it to the packet. The
   latter is typically used if an inner and outer IP header exist which
   require two subsequent routing lookups to be performed.
 * A new metadata_dst structure which can be attached to skbs to
   carry metadata in between subsystems. This new metadata transport
   is used to provide a single interface for VXLAN, routing and OVS
   to communicate through metadata.

The OVS interfaces remain as-is but will transparently create a real
VXLAN net_device in the background. iproute2 is extended with a new
use cases:

  VXLAN:
  ip route add 40.1.1.1/32 encap vxlan id 10 dst 50.1.1.2 dev vxlan0

  MPLS:
  ip route add 10.1.1.0/30 encap mpls 200 via inet 10.1.1.1 dev swp1

Performance implications:
  The additional memory allocation in the receive path should have
  performance implications although it is not observable in standard
  throughput tests if GRO is properly done. The correct net_device
  model outweights the additional cost of the allocation. Furthermore,
  this implication can be relaxed by reintroducing a direct unqueued
  path from a software device to a consumer like bridge or OVS if
  needed.

    $ netperf  -t TCP_STREAM -H 15.1.1.201
    MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
    15.1.1.201 (15.1.1.201) port 0 AF_INET : demo
    Recv   Send    Send
    Socket Socket  Message  Elapsed
    Size   Size    Size     Time     Throughput
    bytes  bytes   bytes    secs.    10^6bits/sec

     87380  16384  16384    10.00    9118.17

Changes since v1:
 * Properly initialize tun_id as reported by Julian
 * Drop dupliate netif_keep_dst() as reported by Alexei
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoopenvswitch: Use regular VXLAN net_device device
Thomas Graf [Tue, 21 Jul 2015 08:44:06 +0000 (10:44 +0200)]
openvswitch: Use regular VXLAN net_device device

This gets rid of all OVS specific VXLAN code in the receive and
transmit path by using a VXLAN net_device to represent the vport.
Only a small shim layer remains which takes care of handling the
VXLAN specific OVS Netlink configuration.

Unexports vxlan_sock_add(), vxlan_sock_release(), vxlan_xmit_skb()
since they are no longer needed.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoopenvswitch: Abstract vport name through ovs_vport_name()
Thomas Graf [Tue, 21 Jul 2015 08:44:05 +0000 (10:44 +0200)]
openvswitch: Abstract vport name through ovs_vport_name()

This allows to get rid of the get_name() vport ops later on.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoopenvswitch: Move dev pointer into vport itself
Thomas Graf [Tue, 21 Jul 2015 08:44:04 +0000 (10:44 +0200)]
openvswitch: Move dev pointer into vport itself

This is the first step in representing all OVS vports as regular
struct net_devices. Move the net_device pointer into the vport
structure itself to get rid of struct vport_netdev.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoopenvswitch: Make tunnel set action attach a metadata dst
Thomas Graf [Tue, 21 Jul 2015 08:44:03 +0000 (10:44 +0200)]
openvswitch: Make tunnel set action attach a metadata dst

Utilize the new metadata dst to attach encapsulation instructions to
the skb. The existing egress_tun_info via the OVS_CB() is left in
place until all tunnel vports have been converted to the new method.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agovxlan: Factor out device configuration
Thomas Graf [Tue, 21 Jul 2015 08:44:02 +0000 (10:44 +0200)]
vxlan: Factor out device configuration

This factors out the device configuration out of the RTNL newlink
API which allows for in-kernel creation of VXLAN net_devices.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agofib: Add fib rule match on tunnel id
Thomas Graf [Tue, 21 Jul 2015 08:44:01 +0000 (10:44 +0200)]
fib: Add fib rule match on tunnel id

This add the ability to select a routing table based on the tunnel
id which allows to maintain separate routing tables for each virtual
tunnel network.

ip rule add from all tunnel-id 100 lookup 100
ip rule add from all tunnel-id 200 lookup 200

A new static key controls the collection of metadata at tunnel level
upon demand.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoroute: Per route IP tunnel metadata via lightweight tunnel
Thomas Graf [Tue, 21 Jul 2015 08:44:00 +0000 (10:44 +0200)]
route: Per route IP tunnel metadata via lightweight tunnel

This introduces a new IP tunnel lightweight tunnel type which allows
to specify IP tunnel instructions per route. Only IPv4 is supported
at this point.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoroute: Extend flow representation with tunnel key
Thomas Graf [Tue, 21 Jul 2015 08:43:59 +0000 (10:43 +0200)]
route: Extend flow representation with tunnel key

Add a new flowi_tunnel structure which is a subset of ip_tunnel_key to
allow routes to match on tunnel metadata. For now, the tunnel id is
added to flowi_tunnel which allows for routes to be bound to specific
virtual tunnels.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agovxlan: Flow based tunneling
Thomas Graf [Tue, 21 Jul 2015 08:43:58 +0000 (10:43 +0200)]
vxlan: Flow based tunneling

Allows putting a VXLAN device into a new flow-based mode in which
skbs with a ip_tunnel_info dst metadata attached will be encapsulated
according to the instructions stored in there with the VXLAN device
defaults taken into consideration.

Similar on the receive side, if the VXLAN_F_COLLECT_METADATA flag is
set, the packet processing will populate a ip_tunnel_info struct for
each packet received and attach it to the skb using the new metadata
dst.  The metadata structure will contain the outer header and tunnel
header fields which have been stripped off. Layers further up in the
stack such as routing, tc or netfitler can later match on these fields
and perform forwarding. It is the responsibility of upper layers to
ensure that the flag is set if the metadata is needed. The flag limits
the additional cost of metadata collecting based on demand.

This prepares the VXLAN device to be steered by the routing and other
subsystems which allows to support encapsulation for a large number
of tunnel endpoints and tunnel ids through a single net_device which
improves the scalability.

It also allows for OVS to leverage this mode which in turn allows for
the removal of the OVS specific VXLAN code.

Because the skb is currently scrubed in vxlan_rcv(), the attachment of
the new dst metadata is postponed until after scrubing which requires
the temporary addition of a new member to vxlan_metadata. This member
is removed again in a later commit after the indirect VXLAN receive API
has been removed.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoarp: Inherit metadata dst when creating ARP requests
Thomas Graf [Tue, 21 Jul 2015 08:43:57 +0000 (10:43 +0200)]
arp: Inherit metadata dst when creating ARP requests

If output device wants to see the dst, inherit the dst of the
original skb and pass it on to generate the ARP request.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agodst: Metadata destinations
Thomas Graf [Tue, 21 Jul 2015 08:43:56 +0000 (10:43 +0200)]
dst: Metadata destinations

Introduces a new dst_metadata which enables to carry per packet metadata
between forwarding and processing elements via the skb->dst pointer.

The structure is set up to be a union. Thus, each separate type of
metadata requires its own dst instance. If demand arises to carry
multiple types of metadata concurrently, metadata dst entries can be
made stackable.

The metadata dst entry is refcnt'ed as expected for now but a non
reference counted use is possible if the reference is forced before
queueing the skb.

In order to allow allocating dsts with variable length, the existing
dst_alloc() is split into a dst_alloc() and dst_init() function. The
existing dst_init() function to initialize the subsystem is being
renamed to dst_subsys_init() to make it clear what is what.

The check before ip_route_input() is changed to ignore metadata dsts
and drop the dst inside the routing function thus allowing to interpret
metadata in a later commit.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoicmp: Don't leak original dst into ip_route_input()
Thomas Graf [Tue, 21 Jul 2015 08:43:55 +0000 (10:43 +0200)]
icmp: Don't leak original dst into ip_route_input()

ip_route_input() unconditionally overwrites the dst. Hide the original
dst attached to the skb by calling skb_dst_set(skb, NULL) prior to
ip_route_input().

Reported-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoip_tunnel: Make ovs_tunnel_info and ovs_key_ipv4_tunnel generic
Thomas Graf [Tue, 21 Jul 2015 08:43:54 +0000 (10:43 +0200)]
ip_tunnel: Make ovs_tunnel_info and ovs_key_ipv4_tunnel generic

Rename the tunnel metadata data structures currently internal to
OVS and make them generic for use by all IP tunnels.

Both structures are kernel internal and will stay that way. Their
members are exposed to user space through individual Netlink
attributes by OVS. It will therefore be possible to extend/modify
these structures without affecting user ABI.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agompls: ip tunnel support
Roopa Prabhu [Tue, 21 Jul 2015 08:43:53 +0000 (10:43 +0200)]
mpls: ip tunnel support

This implementation uses lwtunnel infrastructure to register
hooks for mpls tunnel encaps.

It picks cues from iptunnel_encaps infrastructure and previous
mpls iptunnel RFC patches from Eric W. Biederman and Robert Shearman

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agompls: export mpls functions for use by mpls iptunnels
Roopa Prabhu [Tue, 21 Jul 2015 08:43:52 +0000 (10:43 +0200)]
mpls: export mpls functions for use by mpls iptunnels

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoipv6: rt6_info output redirect to tunnel output
Roopa Prabhu [Tue, 21 Jul 2015 08:43:51 +0000 (10:43 +0200)]
ipv6: rt6_info output redirect to tunnel output

This is similar to ipv4 redirect of dst output to lwtunnel
output function for encapsulation and xmit.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoipv4: redirect dst output to lwtunnel output
Roopa Prabhu [Tue, 21 Jul 2015 08:43:50 +0000 (10:43 +0200)]
ipv4: redirect dst output to lwtunnel output

For input routes with tunnel encap state this patch redirects
dst output functions to lwtunnel_output which later resolves to
the corresponding lwtunnel output function.

This has been tested to work with mpls ip tunnels.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agolwtunnel: support dst output redirect function
Roopa Prabhu [Tue, 21 Jul 2015 08:43:49 +0000 (10:43 +0200)]
lwtunnel: support dst output redirect function

This patch introduces lwtunnel_output function to call corresponding
lwtunnels output function to xmit the packet.

It adds two variants lwtunnel_output and lwtunnel_output6 for ipv4 and
ipv6 respectively today. But this is subject to change when lwtstate will
reside in dst or dst_metadata (as per upstream discussions).

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoipv6: support for fib route lwtunnel encap attributes
Roopa Prabhu [Tue, 21 Jul 2015 08:43:48 +0000 (10:43 +0200)]
ipv6: support for fib route lwtunnel encap attributes

This patch adds support in ipv6 fib functions to parse Netlink
RTA encap attributes and attach encap state data to rt6_info.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoipv4: support for fib route lwtunnel encap attributes
Roopa Prabhu [Tue, 21 Jul 2015 08:43:47 +0000 (10:43 +0200)]
ipv4: support for fib route lwtunnel encap attributes

This patch adds support in ipv4 fib functions to parse user
provided encap attributes and attach encap state data to fib_nh
and rtable.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agolwtunnel: infrastructure for handling light weight tunnels like mpls
Roopa Prabhu [Tue, 21 Jul 2015 08:43:46 +0000 (10:43 +0200)]
lwtunnel: infrastructure for handling light weight tunnels like mpls

Provides infrastructure to parse/dump/store encap information for
light weight tunnels like mpls. Encap information for such tunnels
is associated with fib routes.

This infrastructure is based on previous suggestions from
Eric Biederman to follow the xfrm infrastructure.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agortnetlink: introduce new RTA_ENCAP_TYPE and RTA_ENCAP attributes
Roopa Prabhu [Tue, 21 Jul 2015 08:43:45 +0000 (10:43 +0200)]
rtnetlink: introduce new RTA_ENCAP_TYPE and RTA_ENCAP attributes

This patch introduces two new RTA attributes to attach encap
data to fib routes.

Example iproute2 command to attach mpls encap data to ipv4 routes

$ip route add 10.1.1.0/30 encap mpls 200 via inet 10.1.1.1 dev swp1

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Suggested-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoinet: frags: fix defragmented packet's IP header for af_packet
Edward Hyunkoo Jee [Tue, 21 Jul 2015 07:43:59 +0000 (09:43 +0200)]
inet: frags: fix defragmented packet's IP header for af_packet

When ip_frag_queue() computes positions, it assumes that the passed
sk_buff does not contain L2 headers.

However, when PACKET_FANOUT_FLAG_DEFRAG is used, IP reassembly
functions can be called on outgoing packets that contain L2 headers.

Also, IPv4 checksum is not corrected after reassembly.

Fixes: 7736d33f4262 ("packet: Add pre-defragmentation support for ipv4 fanouts.")
Signed-off-by: Edward Hyunkoo Jee <edjee@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Jerry Chu <hkchu@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoqmi_wwan: Add support for Dell Wireless 5809e 4G Modem
Pieter Hollants [Mon, 20 Jul 2015 08:14:13 +0000 (10:14 +0200)]
qmi_wwan: Add support for Dell Wireless 5809e 4G Modem

Added the USB IDs 0x413c:0x81b1 for the "Dell Wireless 5809e Gobi(TM) 4G
LTE Mobile Broadband Card", a Dell-branded Sierra Wireless EM7305 LTE
card in M.2 form factor, used eg. in Dell's Latitude E7540 Notebook
series.

Signed-off-by: Pieter Hollants <pieter@hollants.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet: mvneta: fix refilling for Rx DMA buffers
Simon Guinot [Sun, 19 Jul 2015 11:00:53 +0000 (13:00 +0200)]
net: mvneta: fix refilling for Rx DMA buffers

With the actual code, if a memory allocation error happens while
refilling a Rx descriptor, then the original Rx buffer is both passed
to the networking stack (in a SKB) and let in the Rx ring. This leads
to various kernel oops and crashes.

As a fix, this patch moves Rx descriptor refilling ahead of building
SKB with the associated Rx buffer. In case of a memory allocation
failure, data is dropped and the original DMA buffer is put back into
the Rx ring.

Signed-off-by: Simon Guinot <simon.guinot@sequanux.org>
Fixes: c5aff18204da ("net: mvneta: driver for Marvell Armada 370/XP network unit")
Cc: <stable@vger.kernel.org> # v3.8+
Tested-by: Yoann Sculo <yoann@sculo.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoxfrm: Fix a typo
Jakub Wilk [Sat, 18 Jul 2015 12:41:51 +0000 (14:41 +0200)]
xfrm: Fix a typo

Signed-off-by: Jakub Wilk <jwilk@jwilk.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agostmmac: fix setting of driver data in stmmac_dvr_probe
Joachim Eastwood [Fri, 17 Jul 2015 21:48:17 +0000 (23:48 +0200)]
stmmac: fix setting of driver data in stmmac_dvr_probe

Commit 803f8fc46274b ("stmmac: move driver data setting into
stmmac_dvr_probe") mistakenly set priv and not priv->dev as
driver data. This meant that the remove, resume and suspend
callbacks that fetched and tried to use this data would most
likely explode. Fix the issue by using the correct variable.

Fixes: 803f8fc46274b ("stmmac: move driver data setting into stmmac_dvr_probe")
Signed-off-by: Joachim Eastwood <manabian@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'sch_panic'
David S. Miller [Tue, 21 Jul 2015 07:25:03 +0000 (00:25 -0700)]
Merge branch 'sch_panic'

Daniel Borkmann says:

====================
Couple of classifier fixes

This fixes a couple of panics in the form of (analogous for
cls_flow{,er}):

[  912.759276] BUG: unable to handle kernel NULL pointer dereference at (null)
[  912.759373] IP: [<ffffffffa09d4d6d>] cls_bpf_change+0x23d/0x268 [cls_bpf]
[  912.759441] PGD 8783c067 PUD 5f684067 PMD 0
[  912.759491] Oops: 0002 [#1] SMP DEBUG_PAGEALLOC
[  912.759543] Modules linked in: cls_bpf(E) act_gact [...]
[  912.772734] CPU: 3 PID: 10489 Comm: tc Tainted: G        W   E   4.2.0-rc2+ #73
[  912.775004] Hardware name: Apple Inc. MacBookAir5,1/Mac-66F35F19FE2A0D05, BIOS MBA51.88Z.00EF.B02.1211271028 11/27/2012
[  912.777327] task: ffff88025eaa8000 ti: ffff88005f734000 task.ti: ffff88005f734000
[  912.779662] RIP: 0010:[<ffffffffa09d4d6d>]  [<ffffffffa09d4d6d>] cls_bpf_change+0x23d/0x268 [cls_bpf]
[  912.781991] RSP: 0018:ffff88005f7379c8  EFLAGS: 00010286
[  912.784183] RAX: ffff880201d64e48 RBX: 0000000000000000 RCX: ffff880201d64e40
[  912.786402] RDX: 0000000000000000 RSI: ffffffffa09d51c0 RDI: ffffffffa09d51a6
[  912.788625] RBP: ffff88005f737a68 R08: 0000000000000000 R09: 0000000000000000
[  912.790854] R10: 0000000000000001 R11: 0000000000000001 R12: ffff880078ab5a80
[  912.793082] R13: ffff880232b31570 R14: ffff88005f737ae0 R15: ffff8801e215d1d0
[  912.795181] FS:  00007f3c0c80d740(0000) GS:ffff880265400000(0000) knlGS:0000000000000000
[  912.797281] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  912.799402] CR2: 0000000000000000 CR3: 000000005460f000 CR4: 00000000001407e0
[  912.799403] Stack:
[  912.799407]  ffffffff00000000 ffff88023ea18000 000000005f737a08 0000000000000000
[  912.799415]  ffffffff81f06140 ffff880201d64e40 0000000000000000 ffff88023ea1804c
[  912.799418]  0000000000000000 ffff88023ea18044 ffff88023ea18030 ffff88023ea18038
[  912.799418] Call Trace:
[  912.799437]  [<ffffffff816d5685>] tc_ctl_tfilter+0x335/0x910
[  912.799443]  [<ffffffff813622a8>] ? security_capable+0x48/0x60
[  912.799448]  [<ffffffff816b90e5>] rtnetlink_rcv_msg+0x95/0x240
[  912.799454]  [<ffffffff810f612d>] ? trace_hardirqs_on+0xd/0x10
[  912.799456]  [<ffffffff816b902f>] ? rtnetlink_rcv+0x1f/0x40
[  912.799459]  [<ffffffff816b902f>] ? rtnetlink_rcv+0x1f/0x40
[  912.799461]  [<ffffffff816b9050>] ? rtnetlink_rcv+0x40/0x40
[  912.799464]  [<ffffffff816df38f>] netlink_rcv_skb+0xaf/0xc0
[  912.799467]  [<ffffffff816b903e>] rtnetlink_rcv+0x2e/0x40
[  912.799469]  [<ffffffff816deaef>] netlink_unicast+0xef/0x1b0
[  912.799471]  [<ffffffff816defa0>] netlink_sendmsg+0x3f0/0x620
[  912.799476]  [<ffffffff81687028>] sock_sendmsg+0x38/0x50
[  912.799479]  [<ffffffff81687938>] ___sys_sendmsg+0x288/0x290
[  912.799482]  [<ffffffff810f7852>] ? __lock_acquire+0x572/0x2050
[  912.799488]  [<ffffffff810265db>] ? native_sched_clock+0x2b/0x90
[  912.799493]  [<ffffffff8116135f>] ? __audit_syscall_entry+0xaf/0x100
[  912.799497]  [<ffffffff8116135f>] ? __audit_syscall_entry+0xaf/0x100
[  912.799501]  [<ffffffff8112aa19>] ? current_kernel_time+0x69/0xd0
[  912.799505]  [<ffffffff81266f16>] ? __fget_light+0x66/0x90
[  912.799508]  [<ffffffff81688812>] __sys_sendmsg+0x42/0x80
[  912.799510]  [<ffffffff81688862>] SyS_sendmsg+0x12/0x20
[  912.799515]  [<ffffffff817f9a6e>] entry_SYSCALL_64_fastpath+0x12/0x76
[  912.799540] Code: 4d 88 49 8b 57 08 48 89 51 08 49 8b 57 10 48 89 c8 48 83 c0 08 48
                     89 51 10 48 8b 51 10 48 c7 c6 c0 51 9d a0 48 c7 c7 a6 51 9d a0 <48>
                     89 02 48 8b 51 08 48 89 42 08 48 b8 00 02 20 00 00 00 ad de
[  912.799544] RIP  [<ffffffffa09d4d6d>] cls_bpf_change+0x23d/0x268 [cls_bpf]
[  912.799544]  RSP <ffff88005f7379c8>
[  912.799545] CR2: 0000000000000000
[  912.807380] ---[ end trace a6440067cfdc7c29 ]---

I've split them into 3 patches, so they can be backported easier
when needed.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosched: cls_flow: fix panic on filter replace
Daniel Borkmann [Fri, 17 Jul 2015 20:38:45 +0000 (22:38 +0200)]
sched: cls_flow: fix panic on filter replace

The following test case causes a NULL pointer dereference in cls_flow:

  tc filter add dev foo parent 1: handle 0x1 flow hash keys dst action ok
  tc filter replace dev foo parent 1: pref 49152 handle 0x1 \
            flow hash keys mark action drop

To be more precise, actually two different panics are fixed, the first
occurs because tcf_exts_init() is not called on the newly allocated
filter when we do a replace. And the second panic uncovered after that
happens since the arguments of list_replace_rcu() are swapped, the old
element needs to be the first argument and the new element the second.

Fixes: 70da9f0bf999 ("net: sched: cls_flow use RCU")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosched: cls_flower: fix panic on filter replace
Daniel Borkmann [Fri, 17 Jul 2015 20:38:44 +0000 (22:38 +0200)]
sched: cls_flower: fix panic on filter replace

The following test case causes a NULL pointer dereference in cls_flower:

  tc filter add dev foo parent 1: flower eth_type ipv4 action ok flowid 1:1
  tc filter replace dev foo parent 1: pref 49152 handle 0x1 \
            flower eth_type ipv6 action ok flowid 1:1

The problem is that commit 77b9900ef53a ("tc: introduce Flower classifier")
accidentally swapped the arguments of list_replace_rcu(), the old
element needs to be the first argument and the new element the second.

Fixes: 77b9900ef53a ("tc: introduce Flower classifier")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosched: cls_bpf: fix panic on filter replace
Daniel Borkmann [Fri, 17 Jul 2015 20:38:43 +0000 (22:38 +0200)]
sched: cls_bpf: fix panic on filter replace

The following test case causes a NULL pointer dereference in cls_bpf:

  FOO="1,6 0 0 4294967295,"
  tc filter add dev foo parent 1: bpf bytecode "$FOO" flowid 1:1 action ok
  tc filter replace dev foo parent 1: pref 49152 handle 0x1 \
            bpf bytecode "$FOO" flowid 1:1 action drop

The problem is that commit 1f947bf151e9 ("net: sched: rcu'ify cls_bpf")
accidentally swapped the arguments of list_replace_rcu(), the old
element needs to be the first argument and the new element the second.

Fixes: 1f947bf151e9 ("net: sched: rcu'ify cls_bpf")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Acked-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'cxgb4-dcb'
David S. Miller [Tue, 21 Jul 2015 07:23:24 +0000 (00:23 -0700)]
Merge branch 'cxgb4-dcb'

Anish Bhatt says:

====================
cxgb4 DCB updates

The following patchset covers changes to work better with  the userspace
tools cgdcbxd and cgrulesengd and improves firmware support for
host-managed mode.

Also exports traffic class information that was previously not being
exported via dcbnl_ops and unfifies how app selector information is passed
to firmware.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agocxgb4 : Fill DCB priority in vlan control headers
Anish Bhatt [Fri, 17 Jul 2015 20:12:33 +0000 (13:12 -0700)]
cxgb4 : Fill DCB priority in vlan control headers

Signed-off-by: Anish Bhatt <anish@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agocxgb4 : Fill in number of DCB traffic classes supported
Anish Bhatt [Fri, 17 Jul 2015 20:12:32 +0000 (13:12 -0700)]
cxgb4 : Fill in number of DCB traffic classes supported

Signed-off-by: Anish Bhatt <anish@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agocxgb4 : Allow firmware DCB info to be queried in host state
Anish Bhatt [Fri, 17 Jul 2015 20:12:31 +0000 (13:12 -0700)]
cxgb4 : Allow firmware DCB info to be queried in host state

Since finally DCB traffic management is still handled by firmware,
allow firmware to be fully programmed and queried even in host
managed state for the cases where this was previously rejected.

Signed-off-by: Anish Bhatt <anish@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agocxgb4 : Only pass app selector of 0 or 3 to firmware
Anish Bhatt [Fri, 17 Jul 2015 20:12:30 +0000 (13:12 -0700)]
cxgb4 : Only pass app selector of 0 or 3 to firmware

This keeps app format passed to firmware the same irrespective
of DCBx version in use.

Signed-off-by: Anish Bhatt <anish@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosctp: fix cut and paste issue in comment
Marcelo Ricardo Leitner [Fri, 17 Jul 2015 16:50:21 +0000 (13:50 -0300)]
sctp: fix cut and paste issue in comment

Cookie ACK is always received by the association initiator, so fix the
comment to avoid confusion.

Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge branch 'sctp-src-addr'
David S. Miller [Tue, 21 Jul 2015 07:20:18 +0000 (00:20 -0700)]
Merge branch 'sctp-src-addr'

Marcelo Ricardo Leitner says:

====================
sctp: fix src address selection if using secondary address

This series improves the way SCTP chooses its src address so that the
choosen one will always belong to the interface being used for output.

v1->v2:
 - split out the refactoring from the fix itself
 - Doing a full reverse routing as in v1 is not necessary. Only looking
   for the interface that has the address and comparing its number is
   enough.
====================

Acked-by: Vlad Yasevich <vyasevich@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosctp: fix src address selection if using secondary addresses
Marcelo Ricardo Leitner [Fri, 17 Jul 2015 15:34:18 +0000 (12:34 -0300)]
sctp: fix src address selection if using secondary addresses

In short, sctp is likely to incorrectly choose src address if socket is
bound to secondary addresses. This patch fixes it by adding a new check
that checks if such src address belongs to the interface that routing
identified as output.

This is enough to avoid rp_filter drops on remote peer.

Details:

Currently, sctp will do a routing attempt without specifying the src
address and compare the returned value (preferred source) with the
addresses that the socket is bound to. When using secondary addresses,
this will not match.

Then it will try specifying each of the addresses that the socket is
bound to and re-routing, checking if that address is valid as src for
that dst. Thing is, this check alone is weak:

# ip r l
192.168.100.0/24 dev eth1  proto kernel  scope link  src 192.168.100.149
192.168.122.0/24 dev eth0  proto kernel  scope link  src 192.168.122.147

# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 52:54:00:15:18:6a brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.147/24 brd 192.168.122.255 scope global dynamic eth0
       valid_lft 2160sec preferred_lft 2160sec
    inet 192.168.122.148/24 scope global secondary eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:fe15:186a/64 scope link
       valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 52:54:00:b3:91:46 brd ff:ff:ff:ff:ff:ff
    inet 192.168.100.149/24 brd 192.168.100.255 scope global dynamic eth1
       valid_lft 2162sec preferred_lft 2162sec
    inet 192.168.100.148/24 scope global secondary eth1
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:feb3:9146/64 scope link
       valid_lft forever preferred_lft forever
4: ens9: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 52:54:00:05:47:ee brd ff:ff:ff:ff:ff:ff
    inet6 fe80::5054:ff:fe05:47ee/64 scope link
       valid_lft forever preferred_lft forever

# ip r g 192.168.100.193 from 192.168.122.148
192.168.100.193 from 192.168.122.148 dev eth1
    cache

Even if you specify an interface:

# ip r g 192.168.100.193 from 192.168.122.148 oif eth1
192.168.100.193 from 192.168.122.148 dev eth1
    cache

Although this would be valid, peers using rp_filter will drop such
packets as their src doesn't match the routes for that interface.

Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agosctp: reduce indent level on sctp_v4_get_dst
Marcelo Ricardo Leitner [Fri, 17 Jul 2015 15:34:17 +0000 (12:34 -0300)]
sctp: reduce indent level on sctp_v4_get_dst

Paves the day for the next patch. Functionality stays untouched.

Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agoMerge tag 'mac80211-for-davem-2015-07-17' of git://git.kernel.org/pub/scm/linux/kerne...
David S. Miller [Tue, 21 Jul 2015 07:17:53 +0000 (00:17 -0700)]
Merge tag 'mac80211-for-davem-2015-07-17' of git://git./linux/kernel/git/jberg/mac80211

Johannes Berg says:

====================
Some fixes for the current cycle:

 1. Arik introduced an rtnl-locked regulatory API to be able
    to differentiate between place do/don't have the RTNL;
    this fixes missing locking in some of the code paths

 2. Two small mesh bugfixes from Bob, one to avoid treating
    a certain malformed over-the-air frame and one to avoid
    sending a garbage field over the air.

 3. A fix for powersave during WoWLAN suspend from Krishna Chaitanya.

 4. A fix for a powersave vs. aggregation teardown race, from Michal.

 5. Thomas reduced the loglevel of CRDA messages to avoid spamming
    the kernel log with mostly irrelevant information.

 6. Tom fixed a dangling debugfs directory pointer that could cause
    crashes if subsequent addition of the same interface to debugfs
    failed for some reason.

 7. A fix from myself for a list corruption issue in mac80211 during
    combined interface shutdown/removal - shut down interfaces first
    and only then remove them to avoid that.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet/mdio: fix mdio_bus_match for c45 PHY
Shaohui Xie [Fri, 17 Jul 2015 10:07:19 +0000 (18:07 +0800)]
net/mdio: fix mdio_bus_match for c45 PHY

We store c45 PHY's id information in c45_ids, so it should be used to
check the matching between PHY driver and PHY device for c45 PHY.

Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet/vxlan: Fix kernel unaligned access in __vxlan_find_mac
Sowmini Varadhan [Mon, 20 Jul 2015 07:54:50 +0000 (09:54 +0200)]
net/vxlan: Fix kernel unaligned access in __vxlan_find_mac

__vxlan_find_mac invokes ether_addr_equal on the eth_addr field,
which triggers unaligned access messages, so rearrange vxlan_fdb
to avoid this in the most non-intrusive way.

Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Reviewed-by: Jiri Pirko <jiri@resnulli.us>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agonet: ratelimit warnings about dst entry refcount underflow or overflow
Konstantin Khlebnikov [Fri, 17 Jul 2015 11:01:11 +0000 (14:01 +0300)]
net: ratelimit warnings about dst entry refcount underflow or overflow

Kernel generates a lot of warnings when dst entry reference counter
overflows and becomes negative. That bug was seen several times at
machines with outdated 3.10.y kernels. Most like it's already fixed
in upstream. Anyway that flood completely kills machine and makes
further debugging impossible.

Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
9 years agorhashtable: Allow other tasks to be scheduled in large lookup loops
Thomas Graf [Fri, 17 Jul 2015 08:52:48 +0000 (10:52 +0200)]
rhashtable: Allow other tasks to be scheduled in large lookup loops

Depending on system speed, the large lookup/insert/delete loops of the testsuite can
take a considerable amount of time to complete causing watchdog warnings to appear.
Allow other tasks to be scheduled throughout the loops.

Reported-by: Meelis Roos <mroos@linux.ee>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>