Enable ACK timeout estimation algorithm (dynack) using mac80211
set_coverage_class API. Dynack is activated passing coverage class equals to -1
to lower drivers and it is automatically disabled setting valid value for
coverage class.
Define NL80211_ATTR_WIPHY_DYN_ACK flag attribute to enable dynack from
userspace. In order to activate dynack NL80211_FEATURE_ACKTO_ESTIMATION feature
flag must be set by lower drivers to indicate dynack capability.
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi83@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

If the remaining time in the current roc is not long
enough, mac80211 adds the new roc right after it
(if they have similar params).

However, in case of multiple rocs, the "next roc"
is not considered, resulting in multiple rocs,
each one with its own duration.

Refactor the code a bit and consider the next roc,
so a single max roc will be used instead of
multiple rocs (which might last much longer).
Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
Reviewed-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

The new duration (remaining duration after the current
ROC ends) was calculated but not used, making the
optimization worthless.
Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
Reviewed-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

The regdom intersection code simply tries intersecting
each rule of the source with each rule of the target.

Since the resulting intersections are not observed
as a whole, this can result in multiple overlapping/duplicate
entries.

Make the rule addition a bit more smarter, by looking
for rules that can be contained within other rules,
and adding only extended ones.
Signed-off-by: Eliad Peller <eliad@wizery.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

In case of a RRM-supporting connection, in the association request
frame: set the RRM capability flag, and add the required IEs.
Signed-off-by: Assaf Krauss <assaf.krauss@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Add a flag attribute to use in associations, for tagging the target
connection as supporting RRM. It is the responsibility of upper
layers to set this flag only if both the underlying device, and the
target network indeed support RRM.
To be used in ASSOCIATE and CONNECT commands.
Signed-off-by: Assaf Krauss <assaf.krauss@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Radio Resource Measurement (RRM) is a bundle of features which will
require the entire stack to participate.
In this patch, the driver is given the opportunity to advertise the
device's support for these RRM-related features, using feature flags:
1. Support for Quiet IEs.
2. Support for adding DS Parameter Set IE to probe requests.
3. Support for adding WFA TPC Report IE to probe requests.
4. Support for inserting tx power value to tx-ed packets at a fixed
   offset. This is used in action frames, such as RRM's Link
   Measurement Report, where the actual tx power should be reported
   in the frame.
Signed-off-by: Assaf Krauss <assaf.krauss@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

The function description claimed that on error the skb isn't
freed even though it is, and stated return values that are
different than what really happens in the code.
Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Our legal structure changed at some point (see wikipedia), but
we forgot to immediately switch over to the new copyright
notice.

For files that we have modified in the time since the change,
add the proper copyright notice now.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Our legal structure changed at some point (see wikipedia), but
we forgot to immediately switch over to the new copyright
notice.

For files that we have modified in the time since the change,
add the proper copyright notice now.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Not sure how the declaration of ieee80211_tdls_peer_del_work
landed after the double inclusion protection end.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

When starting an offloaded BA session it is
unknown what starting sequence number should be
used. Using last_seq worked in most cases except
after hw restart.

When hw restart is requested last_seq is
(rightfully so) kept unmodified. This ended up
with BA sessions being restarted with an aribtrary
BA window values resulting in dropped frames until
sequence numbers caught up.

Instead of last_seq pick seqno of a first Rxed
frame of a given BA session.

This fixes stalled traffic after hw restart with
offloaded BA sessions (currently only ath10k).
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

This adds one more ACPI ID of a Broadcom bluetooth chip.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

When a device driver is unloaded local->interfaces
list is cleared. If there was more than 1
interface running and connected (bound to a
chanctx) then chantype recalc was called and it
ended up with compat being NULL causing a call
trace warning.

Warn if compat becomes NULL as a result of
incompatible bss_conf.chandef of interfaces bound
to a given channel context only.

The call trace looked like this:

 WARNING: CPU: 2 PID: 2594 at /devel/src/linux/net/mac80211/chan.c:557 ieee80211_recalc_chanctx_chantype+0x2cd/0x2e0()
 Modules linked in: ath10k_pci(-) ath10k_core ath
 CPU: 2 PID: 2594 Comm: rmmod Tainted: G        W     3.16.0-rc1+ #150
 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
  0000000000000009 ffff88001ea279c0 ffffffff818dfa93 0000000000000000
  ffff88001ea279f8 ffffffff810514a8 ffff88001ce09cd0 ffff88001e03cc58
  0000000000000000 ffff88001ce08840 ffff88001ce09cd0 ffff88001ea27a08
 Call Trace:
  [<ffffffff818dfa93>] dump_stack+0x4d/0x66
  [<ffffffff810514a8>] warn_slowpath_common+0x78/0xa0
  [<ffffffff81051585>] warn_slowpath_null+0x15/0x20
  [<ffffffff818a407d>] ieee80211_recalc_chanctx_chantype+0x2cd/0x2e0
  [<ffffffff818a3dda>] ? ieee80211_recalc_chanctx_chantype+0x2a/0x2e0
  [<ffffffff818a4919>] ieee80211_assign_vif_chanctx+0x1a9/0x770
  [<ffffffff818a6220>] __ieee80211_vif_release_channel+0x70/0x130
  [<ffffffff818a6dd3>] ieee80211_vif_release_channel+0x43/0xb0
  [<ffffffff81885f4e>] ieee80211_stop_ap+0x21e/0x5a0
  [<ffffffff8184b9b5>] __cfg80211_stop_ap+0x85/0x520
  [<ffffffff8181c188>] __cfg80211_leave+0x68/0x120
  [<ffffffff8181c268>] cfg80211_leave+0x28/0x40
  [<ffffffff8181c5f3>] cfg80211_netdev_notifier_call+0x373/0x6b0
  [<ffffffff8107f965>] notifier_call_chain+0x55/0x110
  [<ffffffff8107fa41>] raw_notifier_call_chain+0x11/0x20
  [<ffffffff816a8dc0>] call_netdevice_notifiers_info+0x30/0x60
  [<ffffffff816a8eb9>] __dev_close_many+0x59/0xf0
  [<ffffffff816a9021>] dev_close_many+0x81/0x120
  [<ffffffff816aa1c5>] rollback_registered_many+0x115/0x2a0
  [<ffffffff816aa3a6>] unregister_netdevice_many+0x16/0xa0
  [<ffffffff8187d841>] ieee80211_remove_interfaces+0x121/0x1b0
  [<ffffffff8185e0e6>] ieee80211_unregister_hw+0x56/0x110
  [<ffffffffa0011ac4>] ath10k_mac_unregister+0x14/0x60 [ath10k_core]
  [<ffffffffa0014fe7>] ath10k_core_unregister+0x27/0x40 [ath10k_core]
  [<ffffffffa003b1f4>] ath10k_pci_remove+0x44/0xa0 [ath10k_pci]
  [<ffffffff81373138>] pci_device_remove+0x28/0x60
  [<ffffffff814cb534>] __device_release_driver+0x64/0xd0
  [<ffffffff814cbcc8>] driver_detach+0xb8/0xc0
  [<ffffffff814cb23a>] bus_remove_driver+0x4a/0xb0
  [<ffffffff814cc697>] driver_unregister+0x27/0x50
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

The "rcu_dereference()" calls are used directly in conditions.
Since their return values are never dereferenced it is recommended to
use "rcu_access_pointer()" instead of "rcu_dereference()".
Therefore, this patch makes the replacements.

The following Coccinelle semantic patch was used:
@@
@@

(
 if(
 (<+...
- rcu_dereference
+ rcu_access_pointer
  (...)
  ...+>)) {...}
|
 while(
 (<+...
- rcu_dereference
+ rcu_access_pointer
  (...)
  ...+>)) {...}
)
Signed-off-by: Andreea-Cristina Bernat <bernat.ada@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

The use of "rcu_assign_pointer()" is NULLing out the pointer.
According to RCU_INIT_POINTER()'s block comment:
"1.   This use of RCU_INIT_POINTER() is NULLing out the pointer"
it is better to use it instead of rcu_assign_pointer() because it has a
smaller overhead.

The following Coccinelle semantic patch was used:
@@
@@

- rcu_assign_pointer
+ RCU_INIT_POINTER
  (..., NULL)
Signed-off-by: Andreea-Cristina Bernat <bernat.ada@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

When using the cfg80211_inform_bss[_width]() functions drivers
cannot currently indicate whether the data was received in a
beacon or probe response. Fix that by passing a new enum that
indicates such (or unknown).

For good measure, use it in ath6kl.

Acked-by: Kalle Valo <kvalo@qca.qualcomm.com> [ath6kl]
Acked-by: Arend van Spriel <arend@broadcom.com> [brcmfmac]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

There are a few possible cases of where BSS data came from:
 1) only a beacon has been received
 2) only a probe response has been received
 3) the driver didn't report what it received (this happens when
    using cfg80211_inform_bss[_width]())
 4) both probe response and beacon data has been received

Unfortunately, in the userspace API, a few things weren't there:
 a) there was no way to differentiate cases 1) and 4) above
    without comparing the data of the IEs
 b) the TSF was always from the last frame, instead of being
    exposed for beacon/probe response separately like IEs

Fix this by
   i) exporting a new flag attribute that indicates whether or
      not probe response data has been received - this addresses (a)
  ii) exporting a BEACON_TSF attribute that holds the beacon's TSF
      if a beacon has been received
 iii) not exporting the beacon attributes in case (3) above as that
      would just lead userspace into thinking the data actually came
      from a beacon when that isn't clear

To implement this, track inside the IEs struct whether or not it
(definitely) came from a beacon.

Reported-by: William Seto
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

This reverts commit dda444d5.

Channel switching code has been reworked and
improved significantly since the time original
locking issues were found.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Header-less cloned skbs with sufficient headroom need not be cloned
unless the tailroom is going to be modified.

Fix ieee80211_skb_resize so it would only resize cloned skbs if either
the header isn't released or the tailroom is going to be modified.

Some drivers might have assumed that skbs are never cloned, so add a HW
flag that explicitly permits cloned TX skbs. Drivers which do not modify
TX skbs should set this flag to avoid copying skbs.
Signed-off-by: Ido Yariv <idox.yariv@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

When hw acceleration is enabled, the GENERATE_IV or PUT_IV_SPACE flags
will only require headroom space. Consequently, the tailroom-needed
counter can safely be decremented.
Signed-off-by: Ido Yariv <idox.yariv@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

TPC report element is contained in spectrum management's tpc report
action frames and in radio measurement's link measurement report
action frames. Add a function which checks whether an action frame
contains this element. This may be needed by the drivers in order
to set the correct tx power value in these frames.
Signed-off-by: Andrei Otcheretianski <andrei.otcheretianski@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

In the cfg80211_rx_mgmt(), parameter @gfp was used for the memory allocation.
But, memory get allocated under spin_lock_bh(), this implies atomic context.
So, one can't use GFP_KERNEL, only variants with no __GFP_WAIT. Actually, in all
occurrences GFP_ATOMIC is used (wil6210 use GFP_KERNEL by mistake),
and it should be this way or warning triggered in the memory allocation code.

Remove @gfp parameter as no actual choice exist, and use hard coded
GFP_ATOMIC for memory allocation.
Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Reorder some code to make ieee80211_vif_use_reserved_switch()
static, no other changes.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Use get_unaligned_le16 to access llid/plid.
Signed-off-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

If the AP only advertises support for 20MHz (in the
ht operation ie), disable 40MHz and VHT.

This can improve interoperability with APs that
don't like stations exceeding their own
advertised capabilities.
Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

We currently track the QoS capability twice: for all peer stations
in the WLAN_STA_WME flag, and for any clients associated to an AP
interface separately for drivers in the sta->sta.wme field.

Remove the WLAN_STA_WME flag and track the capability only in the
driver-visible field, getting rid of the limitation that the field
is only valid in AP mode.
Reviewed-by: Arik Nemtsov <arik@wizery.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Pull more ACPI and power management updates from Rafael Wysocki:
 "These are a couple of regression fixes, cpuidle menu governor
  optimizations, fixes for ACPI proccessor and battery drivers,
  hibernation fix to avoid problems related to the e820 memory map,
  fixes for a few cpufreq drivers and a new version of the suspend
  profiling tool analyze_suspend.py.

  Specifics:

   - Fix for an ACPI-based device hotplug regression introduced in 3.14
     that causes a kernel panic to trigger when memory hot-remove is
     attempted with CONFIG_ACPI_HOTPLUG_MEMORY unset from Tang Chen

   - Fix for a cpufreq regression introduced in 3.16 that triggers a
     "sleeping function called from invalid context" bug in
     dev_pm_opp_init_cpufreq_table() from Stephen Boyd

   - ACPI battery driver fix for a warning message added in 3.16 that
     prints silly stuff sometimes from Mariusz Ceier

   - Hibernation fix for safer handling of mismatches in the 820 memory
     map between the configurations during image creation and during the
     subsequent restore from Chun-Yi Lee

   - ACPI processor driver fix to handle CPU hotplug notifications
     correctly during system suspend/resume from Lan Tianyu

   - Series of four cpuidle menu governor cleanups that also should
     speed it up a bit from Mel Gorman

   - Fixes for the speedstep-smi, integrator, cpu0 and arm_big_little
     cpufreq drivers from Hans Wennborg, Himangi Saraogi, Markus
     Pargmann and Uwe Kleine-König

   - Version 3.0 of the analyze_suspend.py suspend profiling tool from
     Todd E Brandt"

* tag 'pm+acpi-3.17-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI / battery: Fix warning message in acpi_battery_get_state()
  PM / tools: analyze_suspend.py: update to v3.0
  cpufreq: arm_big_little: fix module license spec
  cpufreq: speedstep-smi: fix decimal printf specifiers
  ACPI / hotplug: Check scan handlers in acpi_scan_hot_remove()
  cpufreq: OPP: Avoid sleeping while atomic
  cpufreq: cpu0: Do not print error message when deferring
  cpufreq: integrator: Use set_cpus_allowed_ptr
  PM / hibernate: avoid unsafe pages in e820 reserved regions
  ACPI / processor: Make acpi_cpu_soft_notify() process CPU FROZEN events
  cpuidle: menu: Lookup CPU runqueues less
  cpuidle: menu: Call nr_iowait_cpu less times
  cpuidle: menu: Use ktime_to_us instead of reinventing the wheel
  cpuidle: menu: Use shifts when calculating averages where possible

Pull DEFINE_PCI_DEVICE_TABLE removal from Bjorn Helgaas:
 "Part two of the PCI changes for v3.17:

    - Remove DEFINE_PCI_DEVICE_TABLE macro use (Benoit Taine)

  It's a mechanical change that removes uses of the
  DEFINE_PCI_DEVICE_TABLE macro.  I waited until later in the merge
  window to reduce conflicts, but it's possible you'll still see a few"

* tag 'pci-v3.17-changes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  PCI: Remove DEFINE_PCI_DEVICE_TABLE macro use

Pull Sparc fixes from David Miller:
 "Hook up the memfd syscall, and properly claim all PCI resources
  discovered when building the PCI device tree"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc: Hook up memfd_create system call.
  sparc64: Properly claim resources as each PCI bus is probed.
  sparc64: Skip bogus PCI bridge ranges.
  sparc64: Expand PCI bridge probing debug logging.

Pull networking fixes from David Miller:
 "I'm sending this out, in particular, to get the iwlwifi fix
  propagated:

   1) Fix build due to missing include in i40e driver, from Lucas
      Tanure.

   2) Memory leak in openvswitch port allocation, from Chirstoph Jaeger.

   3) Check DMA mapping errors in myri10ge, from Stanislaw Gruszka.

   4) Fix various deadlock scenerios in sunvnet driver, from Sowmini
      Varadhan.

   5) Fix cxgb4i build failures with incompatible Kconfig settings of
      the driver vs ipv6, from Anish Bhatt.

   6) Fix generation of ACK packet timestamps in the presence of TSO
      which will be split up, from Willem de Bruijn.

   7) Don't enable sched scan in iwlwifi driver, it causes firmware
      crashes in some revisions.  From Emmanuel Grumbach.

   8) Revert a macvlan simplification that causes crashes.

   9) Handle RTT calculations properly in the presence of repair'd SKBs,
      from Andrey Vagin.

  10) SIT tunnel lookup uses wrong device index in compares, from
      Shmulik Ladkani.

  11) Handle MTU reductions in TCP properly for ipv4 mapped ipv6
      sockets, from Neal Cardwell.

  12) Add missing annotations in rhashtable code, from Thomas Graf.

  13) Fix false interpretation of two RTOs as being from the same TCP
      loss event in the FRTO code, from Neal Cardwell"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (42 commits)
  netlink: Annotate RCU locking for seq_file walker
  rhashtable: fix annotations for rht_for_each_entry_rcu()
  rhashtable: unexport and make rht_obj() static
  rhashtable: RCU annotations for next pointers
  tcp: fix ssthresh and undo for consecutive short FRTO episodes
  tcp: don't allow syn packets without timestamps to pass tcp_tw_recycle logic
  tcp: fix tcp_release_cb() to dispatch via address family for mtu_reduced()
  sit: Fix ipip6_tunnel_lookup device matching criteria
  net: ethernet: ibm: ehea: Remove duplicate object from Makefile
  net: xgene: Check negative return value of xgene_enet_get_ring_size()
  tcp: don't use timestamp from repaired skb-s to calculate RTT (v2)
  net: xilinx: Remove .owner field for driver
  Revert "macvlan: simplify the structure port"
  iwlwifi: mvm: disable scheduled scan to prevent firmware crash
  xen-netback: remove loop waiting function
  xen-netback: don't stop dealloc kthread too early
  xen-netback: move NAPI add/remove calls
  xen-netback: fix debugfs entry creation
  xen-netback: fix debugfs write length check
  net-timestamp: fix missing tcp fragmentation cases
  ...

John W. Linville says:

====================
pull request: wireless 2014-08-14

Please pull this batch of fixes intended for the 3.17 stream...

Arend van Spriel brings two brcmfmac fixes, one which fixes a memory
leak and one which corrects some merge damage.

Emmanuel Grumbach fixes Linus's iwlwifi firmware-related log spam.

Rickard Strandqvist does some proper NULL termination after a call
to strncpy.

Ronald Wahl corrects a carl9170 problem with sending URBs with the
wrong endpoint type (resulting in log spam).
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Silences the following sparse warnings:
net/netlink/af_netlink.c:2926:21: warning: context imbalance in 'netlink_seq_start' - wrong count at exit
net/netlink/af_netlink.c:2972:13: warning: context imbalance in 'netlink_seq_stop' - unexpected unlock
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Call rcu_deference_raw() directly from within rht_for_each_entry_rcu()
as list_for_each_entry_rcu() does.

Fixes the following sparse warnings:
net/netlink/af_netlink.c:2906:25:    expected struct rhash_head const *__mptr
net/netlink/af_netlink.c:2906:25:    got struct rhash_head [noderef] <asn:4>*<noident>

Fixes: e341694e ("netlink: Convert netlink_lookup() to use RCU protected hash table")
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

No need to export rht_obj(), all inner to outer object translations
occur internally. It was intended to be used with rht_for_each() which
now primarily serves as the iterator for rhashtable_remove_pprev() to
effectively flush and free the full table.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Properly annotate next pointers as access is RCU protected in
the lookup path.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix TCP FRTO logic so that it always notices when snd_una advances,
indicating that any RTO after that point will be a new and distinct
loss episode.

Previously there was a very specific sequence that could cause FRTO to
fail to notice a new loss episode had started:

(1) RTO timer fires, enter FRTO and retransmit packet 1 in write queue
(2) receiver ACKs packet 1
(3) FRTO sends 2 more packets
(4) RTO timer fires again (should start a new loss episode)

The problem was in step (3) above, where tcp_process_loss() returned
early (in the spot marked "Step 2.b"), so that it never got to the
logic to clear icsk_retransmits. Thus icsk_retransmits stayed
non-zero. Thus in step (4) tcp_enter_loss() would see the non-zero
icsk_retransmits, decide that this RTO is not a new episode, and
decide not to cut ssthresh and remember the current cwnd and ssthresh
for undo.

There were two main consequences to the bug that we have
observed. First, ssthresh was not decreased in step (4). Second, when
there was a series of such FRTO (1-4) sequences that happened to be
followed by an FRTO undo, we would restore the cwnd and ssthresh from
before the entire series started (instead of the cwnd and ssthresh
from before the most recent RTO). This could result in cwnd and
ssthresh being restored to values much bigger than the proper values.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Fixes: e33099f9 ("tcp: implement RFC5682 F-RTO")
Signed-off-by: David S. Miller <davem@davemloft.net>

tcp_tw_recycle heavily relies on tcp timestamps to build a per-host
ordering of incoming connections and teardowns without the need to
hold state on a specific quadruple for TCP_TIMEWAIT_LEN, but only for
the last measured RTO. To do so, we keep the last seen timestamp in a
per-host indexed data structure and verify if the incoming timestamp
in a connection request is strictly greater than the saved one during
last connection teardown. Thus we can verify later on that no old data
packets will be accepted by the new connection.

During moving a socket to time-wait state we already verify if timestamps
where seen on a connection. Only if that was the case we let the
time-wait socket expire after the RTO, otherwise normal TCP_TIMEWAIT_LEN
will be used. But we don't verify this on incoming SYN packets. If a
connection teardown was less than TCP_PAWS_MSL seconds in the past we
cannot guarantee to not accept data packets from an old connection if
no timestamps are present. We should drop this SYN packet. This patch
closes this loophole.

Please note, this patch does not make tcp_tw_recycle in any way more
usable but only adds another safety check:
Sporadic drops of SYN packets because of reordering in the network or
in the socket backlog queues can happen. Users behing NAT trying to
connect to a tcp_tw_recycle enabled server can get caught in blackholes
and their connection requests may regullary get dropped because hosts
behind an address translator don't have synchronized tcp timestamp clocks.
tcp_tw_recycle cannot work if peers don't have tcp timestamps enabled.

In general, use of tcp_tw_recycle is disadvised.

Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Florian Westphal <fw@strlen.de>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Make sure we use the correct address-family-specific function for
handling MTU reductions from within tcp_release_cb().

Previously AF_INET6 sockets were incorrectly always using the IPv6
code path when sometimes they were handling IPv4 traffic and thus had
an IPv4 dst.
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Diagnosed-by: Willem de Bruijn <willemb@google.com>
Fixes: 563d34d0 ("tcp: dont drop MTU reduction indications")
Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

As of 4fddbf5d ("sit: strictly restrict incoming traffic to tunnel link device"),
when looking up a tunnel, tunnel's underlying interface (t->parms.link)
is verified to match incoming traffic's ingress device.

However the comparison was incorrectly based on skb->dev->iflink.

Instead, dev->ifindex should be used, which correctly represents the
interface from which the IP stack hands the ipip6 packets.

This allows setting up sit tunnels bound to vlan interfaces (otherwise
incoming ipip6 traffic on the vlan interface was dropped due to
ipip6_tunnel_lookup match failure).
Signed-off-by: Shmulik Ladkani <shmulik.ladkani@gmail.com>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>