openwrt/staging/blogic.git
7 years agoi40e: remove unnecessary msleep() delay in i40e_free_vfs
Jacob Keller [Wed, 19 Apr 2017 13:25:53 +0000 (09:25 -0400)]
i40e: remove unnecessary msleep() delay in i40e_free_vfs

The delay was added because of a desire to ensure that the VF driver can
finish up removing. However, pci_disable_sriov already has its own
ssleep() call that will sleep for an entire second, so there is no
reason to add extra delay on top of this by using msleep here. In
practice, an msleep() won't have a huge impact on timing but there is no
real value in keeping it, so lets just simplify the code and remove it.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoi40e: amortize wait time when disabling lots of VFs
Jacob Keller [Wed, 19 Apr 2017 13:25:52 +0000 (09:25 -0400)]
i40e: amortize wait time when disabling lots of VFs

Just as we do in i40e_reset_all_vfs, save some time when freeing VFs by
amortizing the wait time for stopping queues. We can use
i40e_vsi_stop_rings_no_wait() to begin the process of stopping all the
VF rings at once. Then, once we've started the process on each VF we can
begin waiting for the VFs to stop. This helps reduce the total wait time
by a large factor.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoi40e: Reprogram port offloads after reset
Alexander Duyck [Wed, 19 Apr 2017 13:25:51 +0000 (09:25 -0400)]
i40e: Reprogram port offloads after reset

This patch corrects a major oversight in that we were not reprogramming the
ports after a reset.  As a result we completely lost all of the Rx tunnel
offloads on receive including Rx checksum, RSS on inner headers, and ATR.

The fix for this is pretty standard as all we needed to do is reset the
filter bits to pending for all active filters and schedule the sync event.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoi40e: rename index to port to avoid confusion
Jacob Keller [Wed, 19 Apr 2017 13:25:50 +0000 (09:25 -0400)]
i40e: rename index to port to avoid confusion

The .index field of i40e_udp_port_config represents the udp port number.
Rename this variable to port so that it is more obvious.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoi40e: make use of i40e_reset_all_vfs when initializing new VFs
Jacob Keller [Thu, 13 Apr 2017 08:45:55 +0000 (04:45 -0400)]
i40e: make use of i40e_reset_all_vfs when initializing new VFs

When allocating a large number of VFs, the driver previously used
i40e_reset_vf in a sequence. Just as when performing a normal reset,
this accumulates a large amount of delay for handling all of the VFs in
sequence. This delay is mainly due to a hardware requirement to wait
after initiating a reset on the VF.

We recently added a new function, i40e_reset_all_vfs() which can be used
to amortize the delay time, by first triggering all VF resets, then
waiting once, and finally cleaning up and allocating the VFs. This is
almost as good as truly running the resets in parallel.

In order to avoid sending a spurious reset message to a client
interface, we have a check to see whether we've assigned
pf->num_alloc_vfs yet. This was originally intended as a way to
distinguish the "initialization" case from the regular reset case.

Unfortunately, this means that we can't directly use i40e_reset_all_vfs
yet. Lets avoid this check of pf->num_alloc_vfs by replacing it with
a proper VSI state bit which we can use instead. This makes the
intention much clearer and allows us to re-use the i40e_reset_all_vfs
function directly.

Change-ID: I694279b37eb6b5a91b6670182d0c15d10244fd6e
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Mitch Williams <mitch.a.williams@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoi40e: properly spell I40E_VF_STATE_* flags
Jacob Keller [Thu, 13 Apr 2017 08:45:54 +0000 (04:45 -0400)]
i40e: properly spell I40E_VF_STATE_* flags

These flags represent the state of the VF at various times. Do not
spell them as _STAT_ which can be confusing to readers who may think
these refer to statistics.

Change-ID: I6bc092cd472e8276896a1fd7498aced2084312df
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoMerge branch '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next...
David S. Miller [Sun, 30 Apr 2017 03:16:20 +0000 (23:16 -0400)]
Merge branch '10GbE' of git://git./linux/kernel/git/jkirsher/next-queue

Jeff Kirsher says:

====================
10GbE Intel Wired LAN Driver Updates 2017-04-29

This series contains updates to ixgbe and ixgbevf only, most notable is
the addition of XDP support to our 10GbE drivers.

Paul fixes ixgbe to acquire the PHY semaphore before accessing PHY
registers when issuing a device reset.

John adds XDP support (yeah!) for ixgbe.

Emil fixes an issue by flushing the MACVLAN filters on VF reset to avoid
conflicts with other VFs that may end up using the same MAC address.  Also
fixed a bug where ethtool -S displayed some empty fields for ixgbevf
because it was using ixgbe_stats instead ixgbevf_stats for
IXGBEVF_QUEUE_STATS_LEN.

Tony adds the ability to specify a zero MAC address in order to clear the
VF's MAC address from the RAR table.  Also adds support for a new
1000Base-T device based on x550EM_X MAC type.  Fixed an issue where the
RSS key specified by the user would be over-written with a pre-existing
value, so change the rss_key to a pointer so we can check to see if the
key has a value set before attempting to set it.  Fixed the logic for
mailbox support for getting RETA and RSS values, which are only supported
by 82599 and x540 devices.

v2: fixed up patches #2 and #3 based on feedback from Jakub and to
    address build issues when page sizes are larger than 4k
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoixgbevf: Check for RSS key before setting value
Tony Nguyen [Thu, 13 Apr 2017 14:26:07 +0000 (07:26 -0700)]
ixgbevf: Check for RSS key before setting value

The RSS key is being repopulated every time the interface is brought up
regardless of whether there is an existing value. If the user sets the RSS
key and the interface is brought up (e.g. reset), the user specified RSS
key will be overwritten.

This patch changes the rss_key to a pointer so we can check to see if the
key has been populated and preserve it accordingly.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbevf: Fix errors in retrieving RETA and RSS from PF
Tony Nguyen [Thu, 13 Apr 2017 14:26:06 +0000 (07:26 -0700)]
ixgbevf: Fix errors in retrieving RETA and RSS from PF

Mailbox support for getting RETA and RSS is available for only 82599 and
x540; a previous patch reversed the logic and these adapters were
returning not supported.

Also, the NACK check in ixgbevf_get_rss_key_locked() was checking for the
command IXGBE_VF_GET_RETA instead of IXGBE_VF_GET_RSS_KEY.

This patch corrects both issues by correcting the logic and checking for
the right command.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: Check for RSS key before setting value
Tony Nguyen [Thu, 13 Apr 2017 14:26:05 +0000 (07:26 -0700)]
ixgbe: Check for RSS key before setting value

The RSS key is being repopulated every time the interface is brought up
regardless of whether there is an existing value. If the user sets the RSS
key and the interface is brought up (e.g. reset), the user specified RSS
key will be overwritten.

This patch changes the rss_key to a pointer so we can check to see if the
key has been populated and preserve it accordingly.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: Add 1000Base-T device based on X550EM_X MAC
Paul Greenwalt [Thu, 13 Apr 2017 12:07:07 +0000 (08:07 -0400)]
ixgbe: Add 1000Base-T device based on X550EM_X MAC

Add support for new 1000Base-T device based on X550EM_X MAC
type. All PHY operations are disabled as the PHY is controlled
by FW.

Signed-off-by: Paul Greenwalt <paul.greenwalt@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: Allow setting zero MAC address for VF
Tony Nguyen [Wed, 12 Apr 2017 20:35:22 +0000 (13:35 -0700)]
ixgbe: Allow setting zero MAC address for VF

Currently, there is no logic that allows a VF's MAC address to be removed
from the RAR table.

Allow the user to specify a zero MAC address in order to clear the VF's
MAC address from the RAR table.  This functionality is also utilized by
libvirt when removing VFs.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbevf: fix size of queue stats length
Emil Tantilov [Fri, 31 Mar 2017 03:49:02 +0000 (20:49 -0700)]
ixgbevf: fix size of queue stats length

IXGBEVF_QUEUE_STATS_LEN is based on ixgebvf_stats, not ixgbe_stats.

This change fixes a bug where ethtool -S displayed some empty fields.

Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: clean macvlan MAC filter table on VF reset
Emil Tantilov [Tue, 28 Mar 2017 18:27:54 +0000 (11:27 -0700)]
ixgbe: clean macvlan MAC filter table on VF reset

Flush the macvlan filters on VF reset to avoid conflict with other VFs that
may end up using the same MAC address.

The main change here is the call to ixgbe_set_vf_macvlan() with index 0.

Moved ixgbe_set_vf_macvlan() in front of ixgbe_vf_reset_event() to avoid
adding a prototype.

Reported-by: Sritej Kanakadandi Sritej Rama <skanakad@cisco.com>
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: delay tail write to every 'n' packets
John Fastabend [Tue, 28 Mar 2017 16:47:03 +0000 (09:47 -0700)]
ixgbe: delay tail write to every 'n' packets

Current XDP implementation hits the tail on every XDP_TX return
code. This patch changes driver behavior to only hit the tail after
packet processing is complete.

With this patch I can run XDP drop programs @ 14+Mpps and XDP_TX
programs are at ~13.5Mpps.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: add support for XDP_TX action
John Fastabend [Mon, 24 Apr 2017 10:30:18 +0000 (03:30 -0700)]
ixgbe: add support for XDP_TX action

A couple design choices were made here. First I use a new ring
pointer structure xdp_ring[] in the adapter struct instead of
pushing the newly allocated XDP TX rings into the tx_ring[]
structure. This means we have to duplicate loops around rings
in places we want to initialize both TX rings and XDP rings.
But by making it explicit it is obvious when we are using XDP
rings and when we are using TX rings. Further we don't have
to do ring arithmatic which is error prone. As a proof point
for doing this my first patches used only a single ring structure
and introduced bugs in FCoE code and macvlan code paths.

Second I am aware this is not the most optimized version of
this code possible. I want to get baseline support in using
the most readable format possible and then once this series
is included I will optimize the TX path in another series
of patches.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: add XDP support for pass and drop actions
John Fastabend [Mon, 24 Apr 2017 10:30:17 +0000 (03:30 -0700)]
ixgbe: add XDP support for pass and drop actions

Basic XDP drop support for ixgbe. Uses READ_ONCE/xchg semantics on XDP
programs instead of RCU primitives as suggested by Daniel Borkmann and
Alex Duyck.

v2: fix the build issues seen w/ XDP when page sizes are larger than 4K
    and made minor fixes based on feedback from Jakub Kicinski

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoixgbe: Acquire PHY semaphore before device reset
Paul Greenwalt [Mon, 13 Mar 2017 09:47:56 +0000 (05:47 -0400)]
ixgbe: Acquire PHY semaphore before device reset

A recent firmware change fixed an issue to acquire the PHY semaphore before
accessing PHY registers. This led to a case where  SW can issue a device
reset clearing the MDIO registers. This patch makes SW acquire the PHY
semaphore before issuing a device reset.

Signed-off-by: Paul Greenwalt <paul.greenwalt@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
7 years agoMerge branch 'bpf-misc-next'
David S. Miller [Fri, 28 Apr 2017 19:48:16 +0000 (15:48 -0400)]
Merge branch 'bpf-misc-next'

Daniel Borkmann says:

====================
Misc BPF updates

This set cleans up ldimm64 leftovers from early eBPF days and
adds couple of test cases related to this to the verifier test
suite. It also cleans up the kallsyms spinlock (had same patch
also in queue) by relaxing it through switching to _bh variant.
It fixes up test_progs in relation to htons/ntohs and adds
accessor macros for the percpu tests in test_maps.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: provide a generic macro for percpu values for selftests
Daniel Borkmann [Wed, 26 Apr 2017 23:39:35 +0000 (01:39 +0200)]
bpf: provide a generic macro for percpu values for selftests

To overcome bugs as described and fixed in 89087c456fb5 ("bpf: Fix
values type used in test_maps"), provide a generic BPF_DECLARE_PERCPU()
and bpf_percpu() accessor macro for all percpu map values used in
tests.

Declaring variables works as follows (also works for structs):

  BPF_DECLARE_PERCPU(uint32_t, my_value);

They can then be accessed normally as uint32_t type through:

  bpf_percpu(my_value, <cpu_nr>)

For example:

  bpf_percpu(my_value, 0)++;

Implicitly, we make sure that the passed type is allocated and aligned
by gcc at least on a 8-byte boundary, so that it works together with
the map lookup/update syscall for percpu maps. We use it as a usage
example in test_maps, so that others are free to adapt this into their
code when necessary.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: fix _htons occurences in test_progs
Daniel Borkmann [Wed, 26 Apr 2017 23:39:34 +0000 (01:39 +0200)]
bpf: fix _htons occurences in test_progs

Dave reported that on sparc test_progs generates buggy swapped
eth->h_proto protocol comparisons:

  10: (15) if r3 == 0xdd86 goto pc+9
   R0=imm2,min_value=2,max_value=2 R1=pkt(id=0,off=0,r=14) R2=pkt_end R3=inv
   R4=pkt(id=0,off=14,r=14) R5=inv56 R10=fp

This is due to the unconditional ...

  #define htons __builtin_bswap16
  #define ntohs __builtin_bswap16

... in test_progs that causes this. Make use of asm/byteorder.h
and use __constant_htons() where possible and only perform the
bswap16 when on little endian in non-constant case.

Fixes: 6882804c916b ("selftests/bpf: add a test for overlapping packet range checks")
Fixes: 37821613626e ("selftests/bpf: add l4 load balancer test based on sched_cls")
Reported-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: bpf_lock on kallsysms doesn't need to be irqsave
Hannes Frederic Sowa [Wed, 26 Apr 2017 23:39:33 +0000 (01:39 +0200)]
bpf: bpf_lock on kallsysms doesn't need to be irqsave

Hannes rightfully spotted that the bpf_lock doesn't need to be
irqsave variant. We never perform any such updates where this
would be necessary (neither right now nor in future), therefore
relax this further.

Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf: add various test cases to verifier selftests
Daniel Borkmann [Wed, 26 Apr 2017 23:39:32 +0000 (01:39 +0200)]
bpf: add various test cases to verifier selftests

Add several test cases around ldimm64, fp arithmetic and direct
packet access.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobpf, x86_64/arm64: remove old ldimm64 artifacts from jits
Daniel Borkmann [Wed, 26 Apr 2017 23:39:31 +0000 (01:39 +0200)]
bpf, x86_64/arm64: remove old ldimm64 artifacts from jits

For both cases, the verifier is already rejecting such invalid
formed instructions. Thus, remove these artifacts from old times
and align it with ppc64, sparc64 and s390x JITs that don't have
them in the first place.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec...
David S. Miller [Fri, 28 Apr 2017 19:43:24 +0000 (15:43 -0400)]
Merge branch 'master' of git://git./linux/kernel/git/klassert/ipsec-next

Steffen Klassert says:

====================
pull request (net-next): ipsec-next 2017-04-28

Just one patch to fix a misplaced spin_unlock_bh in an error path.

Please pull or let me know if there are problems.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge tag 'mac80211-next-for-davem-2017-04-28' of git://git.kernel.org/pub/scm/linux...
David S. Miller [Fri, 28 Apr 2017 18:41:15 +0000 (14:41 -0400)]
Merge tag 'mac80211-next-for-davem-2017-04-28' of git://git./linux/kernel/git/jberg/mac80211-next

Johannes Berg says:

====================
Another set of patches for -next:
 * API support for concurrent scheduled scan requests
 * API changes for roaming reporting
 * BSS max idle support in mac80211
 * API changes for TX status reporting in mac80211
 * API changes for RX rate reporting in mac80211
 * rewrite monitor logic to prepare for BPF filters
 * bugfix for rare devices without 2.4 GHz support
 * a bugfix for recent DFS changes
 * some further cleanups

The API changes are actually at a nice time, since it's
typically quiet just before the merge window, and trees
can be synchronized easily during it.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agorhashtable: Do not lower max_elems when max_size is zero
Herbert Xu [Fri, 28 Apr 2017 06:10:48 +0000 (14:10 +0800)]
rhashtable: Do not lower max_elems when max_size is zero

The commit 6d684e54690c ("rhashtable: Cap total number of entries
to 2^31") breaks rhashtable users that do not set max_size.  This
is because when max_size is zero max_elems is also incorrectly set
to zero instead of 2^31.

This patch fixes it by only lowering max_elems when max_size is not
zero.

Fixes: 6d684e54690c ("rhashtable: Cap total number of entries to 2^31")
Reported-by: Florian Fainelli <f.fainelli@gmail.com>
Reported-by: kernel test robot <fengguang.wu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agocfg80211: add request id to cfg80211_sched_scan_*() api
Arend Van Spriel [Fri, 28 Apr 2017 12:40:28 +0000 (13:40 +0100)]
cfg80211: add request id to cfg80211_sched_scan_*() api

Have proper request id filled in the SCHED_SCAN_RESULTS and
SCHED_SCAN_STOPPED notifications toward user-space by having the
driver provide it through the api.

Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: Add support for BSS max idle period element
Avraham Stern [Wed, 26 Apr 2017 07:58:47 +0000 (10:58 +0300)]
mac80211: Add support for BSS max idle period element

Parse the BSS max idle period element and set the BSS configuration
accordingly so the driver can use this information to configure the
max idle period and to use protected management frames for keep alive
when required.

The BSS max idle period element is defined in IEEE802.11-2016,
section 9.4.2.79

Signed-off-by: Avraham Stern <avraham.stern@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agocfg80211: unify cfg80211_roamed() and cfg80211_roamed_bss()
Avraham Stern [Wed, 26 Apr 2017 07:58:49 +0000 (10:58 +0300)]
cfg80211: unify cfg80211_roamed() and cfg80211_roamed_bss()

cfg80211_roamed() and cfg80211_roamed_bss() take the same arguments
except that cfg80211_roamed() requires the BSSID and
cfg80211_roamed_bss() requires the bss entry.

Unify the two functions by using a struct for driver initiated
roaming information so that either the BSSID or the bss entry can be
passed as an argument to the unified function.

Signed-off-by: Avraham Stern <avraham.stern@intel.com>
[modified the ath6k, brcm80211, rndis and wlan-ng drivers accordingly]
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
[modify brcmfmac to remove the useless cast, spotted by Arend]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: Fix possible sband related NULL pointer de-reference
Mohammed Shafi Shajakhan [Thu, 27 Apr 2017 07:15:38 +0000 (12:45 +0530)]
mac80211: Fix possible sband related NULL pointer de-reference

Existing API 'ieee80211_get_sdata_band' returns default 2 GHz band even
if the channel context configuration is NULL. This crashes for chipsets
which support 5 Ghz alone when it tries to access members of 'sband'.
Channel context configuration can be NULL in multivif case and when
channel switch is in progress (or) when it fails. Fix this by replacing
the API 'ieee80211_get_sdata_band' with  'ieee80211_get_sband' which
returns a NULL pointer for sband when the channel configuration is NULL.

An example scenario is as below:

In multivif mode (AP + STA) with drivers like ath10k, when we do a
channel switch in the AP vif (which has a number of clients connected)
and a STA vif which is connected to some other AP, when the channel
switch in AP vif fails, while the STA vifs tries to connect to the
other AP, there is a window where the channel context is NULL/invalid
and this results in a crash  while the clients connected to the AP vif
tries to reconnect and this race is very similar to the one investigated
by Michal in https://patchwork.kernel.org/patch/3788161/ and this does
happens with hardware that supports 5Ghz alone after long hours of
testing with continuous channel switch on the AP vif

ieee80211 phy0: channel context reservation cannot be finalized because
some interfaces aren't switching
wlan0: failed to finalize CSA, disconnecting
wlan0-1: deauthenticating from 8c:fd:f0:01:54:9c by local choice
(Reason: 3=DEAUTH_LEAVING)

WARNING: CPU: 1 PID: 19032 at net/mac80211/ieee80211_i.h:1013 sta_info_alloc+0x374/0x3fc [mac80211]
[<bf77272c>] (sta_info_alloc [mac80211])
[<bf78776c>] (ieee80211_add_station [mac80211]))
[<bf73cc50>] (nl80211_new_station [cfg80211])

Unable to handle kernel NULL pointer dereference at virtual
address 00000014
pgd = d5f4c000
Internal error: Oops: 17 [#1] PREEMPT SMP ARM
PC is at sta_info_alloc+0x380/0x3fc [mac80211]
LR is at sta_info_alloc+0x37c/0x3fc [mac80211]
[<bf772738>] (sta_info_alloc [mac80211])
[<bf78776c>] (ieee80211_add_station [mac80211])
[<bf73cc50>] (nl80211_new_station [cfg80211]))

Cc: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: add ieee80211_tx_status_ext
Felix Fietkau [Wed, 26 Apr 2017 15:11:37 +0000 (17:11 +0200)]
mac80211: add ieee80211_tx_status_ext

This allows the driver to pass in struct ieee80211_tx_status directly.
Make ieee80211_tx_status_noskb a wrapper around it.

As with ieee80211_tx_status_noskb, there is no _ni variant of this call,
because it probably won't be needed.

Even if the driver won't provide any extra status info other than what's
in struct ieee80211_tx_info already, it can optimize status reporting
this way by passing in the station pointer.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
[use C99 initializers]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: move ieee80211_tx_status_noskb below ieee80211_tx_status
Felix Fietkau [Wed, 26 Apr 2017 15:11:36 +0000 (17:11 +0200)]
mac80211: move ieee80211_tx_status_noskb below ieee80211_tx_status

Makes further cleanups more readable

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: make rate control tx status API more extensible
Felix Fietkau [Wed, 26 Apr 2017 15:11:35 +0000 (17:11 +0200)]
mac80211: make rate control tx status API more extensible

Rename .tx_status_noskb to .tx_status_ext and pass a new on-stack
struct ieee80211_tx_status instead of struct ieee80211_tx_info.

This struct can be used to pass extra information, e.g. for dynamic tx
power control

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: use bitfield macros for encoded rate
Johannes Berg [Wed, 26 Apr 2017 12:51:20 +0000 (14:51 +0200)]
mac80211: use bitfield macros for encoded rate

Instead of hand-coding the bit manipulations, use the bitfield
macros to generate the code for the encoded bitrate.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: rename ieee80211_rx_status::vht_nss to just nss
Johannes Berg [Wed, 26 Apr 2017 11:51:41 +0000 (13:51 +0200)]
mac80211: rename ieee80211_rx_status::vht_nss to just nss

This field will need to be used again for HE, so rename it now.

Again, mostly done with this spatch:

@@
expression status;
@@
-status->vht_nss
+status->nss
@@
expression status;
@@
-status.vht_nss
+status.nss

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: separate encoding/bandwidth from flags
Johannes Berg [Wed, 26 Apr 2017 10:14:59 +0000 (12:14 +0200)]
mac80211: separate encoding/bandwidth from flags

We currently use a lot of flags that are mutually incompatible,
separate this out into actual encoding and bandwidth enum values.

Much of this again done with spatch, with manual post-editing,
mostly to add the switch statements and get rid of the conversions.

@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_80MHZ
+status->bw = RATE_INFO_BW_80
@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_40MHZ
+status->bw = RATE_INFO_BW_40
@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_20MHZ
+status->bw = RATE_INFO_BW_20
@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_160MHZ
+status->bw = RATE_INFO_BW_160
@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_5MHZ
+status->bw = RATE_INFO_BW_5
@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_10MHZ
+status->bw = RATE_INFO_BW_10

@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_VHT
+status->encoding = RX_ENC_VHT
@@
expression status;
@@
-status->enc_flags |= RX_ENC_FLAG_HT
+status->encoding = RX_ENC_HT
@@
expression status;
@@
-status.enc_flags |= RX_ENC_FLAG_VHT
+status.encoding = RX_ENC_VHT
@@
expression status;
@@
-status.enc_flags |= RX_ENC_FLAG_HT
+status.encoding = RX_ENC_HT

@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_HT)
+(status->encoding == RX_ENC_HT)
@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_VHT)
+(status->encoding == RX_ENC_VHT)

@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_5MHZ)
+(status->bw == RATE_INFO_BW_5)
@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_10MHZ)
+(status->bw == RATE_INFO_BW_10)
@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_40MHZ)
+(status->bw == RATE_INFO_BW_40)
@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_80MHZ)
+(status->bw == RATE_INFO_BW_80)
@@
expression status;
@@
-(status->enc_flags & RX_ENC_FLAG_160MHZ)
+(status->bw == RATE_INFO_BW_160)

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: clean up rate encoding bits in RX status
Johannes Berg [Wed, 26 Apr 2017 09:13:00 +0000 (11:13 +0200)]
mac80211: clean up rate encoding bits in RX status

In preparation for adding support for HE rates, clean up
the driver report encoding for rate/bandwidth reporting
on RX frames.

Much of this patch was done with the following spatch:

@@
expression status;
@@
-status->flag & (RX_FLAG_HT | RX_FLAG_VHT)
+status->enc_flags & (RX_ENC_FLAG_HT | RX_ENC_FLAG_VHT)

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_SHORTPRE
+status->enc_flags op RX_ENC_FLAG_SHORTPRE
@@
expression status;
@@
-status->flag & RX_FLAG_SHORTPRE
+status->enc_flags & RX_ENC_FLAG_SHORTPRE

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_HT
+status->enc_flags op RX_ENC_FLAG_HT
@@
expression status;
@@
-status->flag & RX_FLAG_HT
+status->enc_flags & RX_ENC_FLAG_HT

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_40MHZ
+status->enc_flags op RX_ENC_FLAG_40MHZ
@@
expression status;
@@
-status->flag & RX_FLAG_40MHZ
+status->enc_flags & RX_ENC_FLAG_40MHZ

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_SHORT_GI
+status->enc_flags op RX_ENC_FLAG_SHORT_GI
@@
expression status;
@@
-status->flag & RX_FLAG_SHORT_GI
+status->enc_flags & RX_ENC_FLAG_SHORT_GI

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_HT_GF
+status->enc_flags op RX_ENC_FLAG_HT_GF
@@
expression status;
@@
-status->flag & RX_FLAG_HT_GF
+status->enc_flags & RX_ENC_FLAG_HT_GF

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_VHT
+status->enc_flags op RX_ENC_FLAG_VHT
@@
expression status;
@@
-status->flag & RX_FLAG_VHT
+status->enc_flags & RX_ENC_FLAG_VHT

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_STBC_MASK
+status->enc_flags op RX_ENC_FLAG_STBC_MASK
@@
expression status;
@@
-status->flag & RX_FLAG_STBC_MASK
+status->enc_flags & RX_ENC_FLAG_STBC_MASK

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_LDPC
+status->enc_flags op RX_ENC_FLAG_LDPC
@@
expression status;
@@
-status->flag & RX_FLAG_LDPC
+status->enc_flags & RX_ENC_FLAG_LDPC

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_10MHZ
+status->enc_flags op RX_ENC_FLAG_10MHZ
@@
expression status;
@@
-status->flag & RX_FLAG_10MHZ
+status->enc_flags & RX_ENC_FLAG_10MHZ

@@
assignment operator op;
expression status;
@@
-status->flag op RX_FLAG_5MHZ
+status->enc_flags op RX_ENC_FLAG_5MHZ
@@
expression status;
@@
-status->flag & RX_FLAG_5MHZ
+status->enc_flags & RX_ENC_FLAG_5MHZ

@@
assignment operator op;
expression status;
@@
-status->vht_flag op RX_VHT_FLAG_80MHZ
+status->enc_flags op RX_ENC_FLAG_80MHZ
@@
expression status;
@@
-status->vht_flag & RX_VHT_FLAG_80MHZ
+status->enc_flags & RX_ENC_FLAG_80MHZ

@@
assignment operator op;
expression status;
@@
-status->vht_flag op RX_VHT_FLAG_160MHZ
+status->enc_flags op RX_ENC_FLAG_160MHZ
@@
expression status;
@@
-status->vht_flag & RX_VHT_FLAG_160MHZ
+status->enc_flags & RX_ENC_FLAG_160MHZ

@@
assignment operator op;
expression status;
@@
-status->vht_flag op RX_VHT_FLAG_BF
+status->enc_flags op RX_ENC_FLAG_BF
@@
expression status;
@@
-status->vht_flag & RX_VHT_FLAG_BF
+status->enc_flags & RX_ENC_FLAG_BF

@@
assignment operator op;
expression status, STBC;
@@
-status->flag op STBC << RX_FLAG_STBC_SHIFT
+status->enc_flags op STBC << RX_ENC_FLAG_STBC_SHIFT

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_SHORTPRE
+status.enc_flags op RX_ENC_FLAG_SHORTPRE
@@
expression status;
@@
-status.flag & RX_FLAG_SHORTPRE
+status.enc_flags & RX_ENC_FLAG_SHORTPRE

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_HT
+status.enc_flags op RX_ENC_FLAG_HT
@@
expression status;
@@
-status.flag & RX_FLAG_HT
+status.enc_flags & RX_ENC_FLAG_HT

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_40MHZ
+status.enc_flags op RX_ENC_FLAG_40MHZ
@@
expression status;
@@
-status.flag & RX_FLAG_40MHZ
+status.enc_flags & RX_ENC_FLAG_40MHZ

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_SHORT_GI
+status.enc_flags op RX_ENC_FLAG_SHORT_GI
@@
expression status;
@@
-status.flag & RX_FLAG_SHORT_GI
+status.enc_flags & RX_ENC_FLAG_SHORT_GI

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_HT_GF
+status.enc_flags op RX_ENC_FLAG_HT_GF
@@
expression status;
@@
-status.flag & RX_FLAG_HT_GF
+status.enc_flags & RX_ENC_FLAG_HT_GF

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_VHT
+status.enc_flags op RX_ENC_FLAG_VHT
@@
expression status;
@@
-status.flag & RX_FLAG_VHT
+status.enc_flags & RX_ENC_FLAG_VHT

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_STBC_MASK
+status.enc_flags op RX_ENC_FLAG_STBC_MASK
@@
expression status;
@@
-status.flag & RX_FLAG_STBC_MASK
+status.enc_flags & RX_ENC_FLAG_STBC_MASK

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_LDPC
+status.enc_flags op RX_ENC_FLAG_LDPC
@@
expression status;
@@
-status.flag & RX_FLAG_LDPC
+status.enc_flags & RX_ENC_FLAG_LDPC

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_10MHZ
+status.enc_flags op RX_ENC_FLAG_10MHZ
@@
expression status;
@@
-status.flag & RX_FLAG_10MHZ
+status.enc_flags & RX_ENC_FLAG_10MHZ

@@
assignment operator op;
expression status;
@@
-status.flag op RX_FLAG_5MHZ
+status.enc_flags op RX_ENC_FLAG_5MHZ
@@
expression status;
@@
-status.flag & RX_FLAG_5MHZ
+status.enc_flags & RX_ENC_FLAG_5MHZ

@@
assignment operator op;
expression status;
@@
-status.vht_flag op RX_VHT_FLAG_80MHZ
+status.enc_flags op RX_ENC_FLAG_80MHZ
@@
expression status;
@@
-status.vht_flag & RX_VHT_FLAG_80MHZ
+status.enc_flags & RX_ENC_FLAG_80MHZ

@@
assignment operator op;
expression status;
@@
-status.vht_flag op RX_VHT_FLAG_160MHZ
+status.enc_flags op RX_ENC_FLAG_160MHZ
@@
expression status;
@@
-status.vht_flag & RX_VHT_FLAG_160MHZ
+status.enc_flags & RX_ENC_FLAG_160MHZ

@@
assignment operator op;
expression status;
@@
-status.vht_flag op RX_VHT_FLAG_BF
+status.enc_flags op RX_ENC_FLAG_BF
@@
expression status;
@@
-status.vht_flag & RX_VHT_FLAG_BF
+status.enc_flags & RX_ENC_FLAG_BF

@@
assignment operator op;
expression status, STBC;
@@
-status.flag op STBC << RX_FLAG_STBC_SHIFT
+status.enc_flags op STBC << RX_ENC_FLAG_STBC_SHIFT

@@
@@
-RX_FLAG_STBC_SHIFT
+RX_ENC_FLAG_STBC_SHIFT

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agoMerge tag 'wireless-drivers-next-for-davem-2017-04-27' of git://git.kernel.org/pub...
David S. Miller [Thu, 27 Apr 2017 21:15:54 +0000 (17:15 -0400)]
Merge tag 'wireless-drivers-next-for-davem-2017-04-27' of git://git./linux/kernel/git/kvalo/wireless-drivers-next

Kalle Valo says:

====================
wireless-drivers-next patches for 4.12

Few remaining patches for 4.12 submitted during the last week.

Major changes:

iwlwifi

* the firmware for 7265D and 3168 NICs is frozen at version 29

* more support for the upcoming A000 series
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'qed-ptp-enhancements'
David S. Miller [Thu, 27 Apr 2017 20:51:35 +0000 (16:51 -0400)]
Merge branch 'qed-ptp-enhancements'

Sudarsana Reddy Kalluru says:

====================
qed*: PTP enhancements.

The patch series contains set of enhancements for qed/qede ptp
implementation.
Patches (1)-(3) adds resource locking implementation to allow
PTP functionality only on the first detected ethernet PF of the port.
The change is required as the adapter currently supports only one
instance of the PTP client on a given port.
Patch (4) removes the un-needed header file.
Patch (5) moves the ptt-lock get/release logic to the ptp specific
code.

Please consider applying this series to "net-next" branch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: Acquire/release ptt_ptp lock when enabling/disabling PTP.
sudarsana.kalluru@cavium.com [Wed, 26 Apr 2017 16:00:53 +0000 (09:00 -0700)]
qed: Acquire/release ptt_ptp lock when enabling/disabling PTP.

Move the code for acquiring/releasing ptt_ptp lock to ptp specific
implementations i.e., ptp_enable()/disable() respectively.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: Remove the un-needed ptp header file.
sudarsana.kalluru@cavium.com [Wed, 26 Apr 2017 16:00:52 +0000 (09:00 -0700)]
qed: Remove the un-needed ptp header file.

The patch deletes the qed_ptp.h file which is not required.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqede: Add support for PTP resource locking.
sudarsana.kalluru@cavium.com [Wed, 26 Apr 2017 16:00:51 +0000 (09:00 -0700)]
qede: Add support for PTP resource locking.

The patch adds necessary changes to the driver to use qed resource
locking functionality. Currently the ptp initialization is spread
between driver probe/open implementations, associated APIs are
qede_ptp_register_phc()/qede_ptp_start(). Clubbed this functionality
into single API qed_ptp_enable() to simplify the usage of qed resource
locking implementation. The new API will be invoked in the probe path.
Similarly the ptp clean-up code is moved to qede_ptp_disable() which
gets invoked in the driver unload path.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: Add support for PTP resource locking.
sudarsana.kalluru@cavium.com [Wed, 26 Apr 2017 16:00:50 +0000 (09:00 -0700)]
qed: Add support for PTP resource locking.

The patch adds support for per-port resource lock in favour of PTP.
PTP module acquires/releases the MFW resource lock while enabling/
disabling the PTP on the interface. The PF instance which has the
ownership of this resource lock will get the exclusive access to the
PTP clock functionality on the port.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoqed: Add support for MFW resource locking.
sudarsana.kalluru@cavium.com [Wed, 26 Apr 2017 16:00:49 +0000 (09:00 -0700)]
qed: Add support for MFW resource locking.

The patch adds API for default initialization of the MFW resource
locking.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: vrf: Do not allow looback to be moved to a VRF
David Ahern [Wed, 26 Apr 2017 14:58:22 +0000 (07:58 -0700)]
net: vrf: Do not allow looback to be moved to a VRF

Moving the loopback into a VRF breaks networking for the default VRF.
Since the VRF device is the loopback for VRF domains, there is no
reason to move the loopback. Given the repercussions, block attempts
to set lo into a VRF.

Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Reviewed-by: Greg Rose <gvrose8192@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agofib_rules: fix error return code
Wei Yongjun [Wed, 26 Apr 2017 14:03:50 +0000 (14:03 +0000)]
fib_rules: fix error return code

Fix to return error code -EINVAL from the error handling
case instead of 0, as done elsewhere in this function.

Fixes: 622ec2c9d524 ("net: core: add UID to flows, rules, and routes")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agobridge: add per-port broadcast flood flag
Mike Manning [Wed, 26 Apr 2017 13:48:09 +0000 (14:48 +0100)]
bridge: add per-port broadcast flood flag

Support for l2 multicast flood control was added in commit b6cb5ac8331b
("net: bridge: add per-port multicast flood flag"). It allows broadcast
as it was introduced specifically for unknown multicast flood control.
But as broadcast is a special case of multicast, this may also need to
be disabled. For this purpose, introduce a flag to disable the flooding
of received l2 broadcasts. This approach is backwards compatible and
provides flexibility in filtering for the desired packet types.

Cc: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: Mike Manning <mmanning@brocade.com>
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: fib: Decrease one unnecessary rt cache flush in fib_disable_ip
Gao Feng [Wed, 26 Apr 2017 11:04:04 +0000 (19:04 +0800)]
net: fib: Decrease one unnecessary rt cache flush in fib_disable_ip

The func fib_flush already flushes the rt cache if necessary, so it
is not necessary to invoke rt_cache_flush again in fib_disable_ip.

Signed-off-by: Gao Feng <fgao@ikuai8.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agol2tp: remove useless device duplication test in l2tp_eth_create()
Guillaume Nault [Wed, 26 Apr 2017 09:54:47 +0000 (11:54 +0200)]
l2tp: remove useless device duplication test in l2tp_eth_create()

There's no need to verify that cfg->ifname is unique at this point.
register_netdev() will return -EEXIST if asked to create a device with
a name that's alrealy in use.

Signed-off-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: remove unnecessary carrier status check
Zhang Shengju [Wed, 26 Apr 2017 09:49:38 +0000 (17:49 +0800)]
net: remove unnecessary carrier status check

Since netif_carrier_on() will do nothing if device's carrier is already
on, so it's unnecessary to do carrier status check.

It's the same for netif_carrier_off().

Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: update comment for netif_dormant() function
Zhang Shengju [Wed, 26 Apr 2017 03:05:12 +0000 (11:05 +0800)]
net: update comment for netif_dormant() function

This patch updates the comment for netif_dormant() function to reflect
the intended usage.

Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agosamples/bpf: Add support for SKB_MODE to xdp1 and xdp_tx_iptunnel
David Ahern [Thu, 27 Apr 2017 16:11:13 +0000 (09:11 -0700)]
samples/bpf: Add support for SKB_MODE to xdp1 and xdp_tx_iptunnel

Add option to xdp1 and xdp_tx_iptunnel to insert xdp program in
SKB_MODE:
 - update set_link_xdp_fd to take a flags argument that is added to the
   RTM_SETLINK message

 - Add -S option to xdp1 and xdp_tx_iptunnel user code. When passed in
   XDP_FLAGS_SKB_MODE is set in the flags arg passed to set_link_xdp_fd

Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agorhashtable: Cap total number of entries to 2^31
Herbert Xu [Thu, 27 Apr 2017 05:44:51 +0000 (13:44 +0800)]
rhashtable: Cap total number of entries to 2^31

When max_size is not set or if it set to a sufficiently large
value, the nelems counter can overflow.  This would cause havoc
with the automatic shrinking as it would then attempt to fit a
huge number of entries into a tiny hash table.

This patch fixes this by adding max_elems to struct rhashtable
to cap the number of elements.  This is set to 2^31 as nelems is
not a precise count.  This is sufficiently smaller than UINT_MAX
that it should be safe.

When max_size is set max_elems will be lowered to at most twice
max_size as is the status quo.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: tcp_rack_reo_timeout() must update tp->tcp_mstamp
Eric Dumazet [Thu, 27 Apr 2017 04:10:55 +0000 (21:10 -0700)]
tcp: tcp_rack_reo_timeout() must update tp->tcp_mstamp

I wrongly assumed tp->tcp_mstamp was up to date at the time
tcp_rack_reo_timeout() was called.

It is not true, since we only update tcp->tcp_mstamp when receiving
a packet (as initially done in commit 69e996c58a35 ("tcp: add
tp->tcp_mstamp field")

tcp_rack_reo_timeout() being called by a timer and not an incoming
packet, we need to refresh tp->tcp_mstamp

Fixes: 7c1c7308592f ("tcp: do not pass timestamp to tcp_rack_detect_loss()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
David S. Miller [Thu, 27 Apr 2017 02:39:08 +0000 (22:39 -0400)]
Merge git://git./linux/kernel/git/davem/net

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc
Linus Torvalds [Wed, 26 Apr 2017 22:10:45 +0000 (15:10 -0700)]
Merge git://git./linux/kernel/git/davem/sparc

Pull sparc fixes from David Miller:
 "I didn't want the release to go out without the statx system call
  properly hooked up"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc: Update syscall tables.
  sparc64: Fill in rest of HAVE_REGS_AND_STACK_ACCESS_API

7 years agostatx: Kill fd-with-NULL-path support in favour of AT_EMPTY_PATH
David Howells [Wed, 26 Apr 2017 21:15:55 +0000 (22:15 +0100)]
statx: Kill fd-with-NULL-path support in favour of AT_EMPTY_PATH

With the new statx() syscall, the following both allow the attributes of
the file attached to a file descriptor to be retrieved:

statx(dfd, NULL, 0, ...);

and:

statx(dfd, "", AT_EMPTY_PATH, ...);

Change the code to reject the first option, though this means copying
the path and engaging pathwalk for the fstat() equivalent.  dfd can be a
non-directory provided path is "".

[ The timing of this isn't wonderful, but applying this now before we
  have statx() in any released kernel, before anybody starts using the
  NULL special case.    - Linus ]

Fixes: a528d35e8bfc ("statx: Add a system call to make enhanced file info available")
Reported-by: Michael Kerrisk <mtk.manpages@gmail.com>
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Eric Sandeen <sandeen@sandeen.net>
cc: fstests@vger.kernel.org
cc: linux-api@vger.kernel.org
cc: linux-man@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
7 years agomac80211: make multicast variable a bool in ieee80211_accept_frame()
Luca Coelho [Wed, 26 Apr 2017 07:58:54 +0000 (10:58 +0300)]
mac80211: make multicast variable a bool in ieee80211_accept_frame()

The multicast variable in the ieee80211_accept_frame() function is
treated as a boolean, but defined as int.  Fix that.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: disentangle iflist_mtx and chanctx_mtx
Johannes Berg [Wed, 26 Apr 2017 07:58:53 +0000 (10:58 +0300)]
mac80211: disentangle iflist_mtx and chanctx_mtx

At least on iwlwifi, sometimes lockdep complains that we can
lock
 chanctx_mtx -> mvm.mutex -> iflist_mtx
 (due to iterate_interfaces)
and
 iflist_mtx -> chanctx_mtx

Remove the latter dependency in mac80211 by using the RTNL
that we already hold in one case, and can relatively easily
achieve in the other case.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agomac80211: don't parse encrypted management frames in ieee80211_frame_acked
Emmanuel Grumbach [Wed, 26 Apr 2017 07:58:51 +0000 (10:58 +0300)]
mac80211: don't parse encrypted management frames in ieee80211_frame_acked

ieee80211_frame_acked is called when a frame is acked by
the peer. In case this is a management frame, we check
if this an SMPS frame, in which case we can update our
antenna configuration.

When we parse the management frame we look at the category
in case it is an action frame. That byte sits after the IV
in case the frame was encrypted. This means that if the
frame was encrypted, we basically look at the IV instead
of looking at the category. It is then theorically
possible that we think that an SMPS action frame was acked
where really we had another frame that was encrypted.

Since the only management frame whose ack needs to be
tracked is the SMPS action frame, and that frame is not
a robust management frame, it will never be encrypted.
The easiest way to fix this problem is then to not look
at frames that were encrypted.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agoieee80211: fix kernel-doc parsing errors
Johannes Berg [Wed, 26 Apr 2017 07:58:52 +0000 (10:58 +0300)]
ieee80211: fix kernel-doc parsing errors

Some of the enum definitions are unnamed but there's still
an attempt at documenting them - that doesn't work. Name
them to make that work.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agoieee80211: add FT-802.1X AKM suite selector
Luca Coelho [Wed, 26 Apr 2017 07:58:48 +0000 (10:58 +0300)]
ieee80211: add FT-802.1X AKM suite selector

Add the definition for FT-8021.1X AKM selector as defined in
IEEE Std 802.11-2016, table 9-133.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agoieee80211: add SUITE_B AKM selectors
Luca Coelho [Wed, 26 Apr 2017 07:58:46 +0000 (10:58 +0300)]
ieee80211: add SUITE_B AKM selectors

Add the definitions for SUITE_B and SUITE_B_192 AKM selectors as
defined in IEEE802.11REVmc_D5.0, table 9-132.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agocfg80211: add request id parameter to .sched_scan_stop() signature
Arend Van Spriel [Fri, 21 Apr 2017 12:05:02 +0000 (13:05 +0100)]
cfg80211: add request id parameter to .sched_scan_stop() signature

For multiple scheduled scan support the driver needs to know which
scheduled scan request is being stopped. Pass the request id in the
.sched_scan_stop() callback.

Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agonl80211: add support for BSSIDs in scheduled scan matchsets
Arend Van Spriel [Fri, 21 Apr 2017 12:05:01 +0000 (13:05 +0100)]
nl80211: add support for BSSIDs in scheduled scan matchsets

This patch allows for the scheduled scan request to specify matchsets
for specific BSSIDs.

Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
[docs, netlink policy fix]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agonl80211: allow multiple active scheduled scan requests
Arend Van Spriel [Fri, 21 Apr 2017 12:05:00 +0000 (13:05 +0100)]
nl80211: allow multiple active scheduled scan requests

This patch implements the idea to have multiple scheduled scan requests
running concurrently. It mainly illustrates how to deal with the incoming
request from user-space in terms of backward compatibility. In order to
use multiple scheduled scans user-space needs to provide a flag attribute
NL80211_ATTR_SCHED_SCAN_MULTI to indicate support. If not the request is
treated as a legacy scan.

Drivers currently supporting scheduled scan are now indicating they support
a single scheduled scan request. This obsoletes WIPHY_FLAG_SUPPORTS_SCHED_SCAN.

Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
[clean up netlink destroy path to avoid allocations, code cleanups]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agocfg80211: simplify netlink socket owner interface deletion
Johannes Berg [Wed, 26 Apr 2017 05:43:41 +0000 (07:43 +0200)]
cfg80211: simplify netlink socket owner interface deletion

There's no need to allocate a portid structure and then, for
each of those, walk the interfaces - we can just add a flag
to each interface and walk those directly. Due to padding in
the struct, we can even do it without any memory cost, and
it even simplifies the code.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
7 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Wed, 26 Apr 2017 20:42:32 +0000 (13:42 -0700)]
Merge git://git./linux/kernel/git/davem/net

Pull networking fixes from David Miller:

 1) MLX5 bug fixes from Saeed Mahameed et al:
     - released wrong resources when firmware timeout happens
     - fix wrong check for encapsulation size limits
     - UAR memory leak
     - ETHTOOL_GRXCLSRLALL failed to fill in info->data

 2) Don't cache l3mdev on mis-matches local route, causes net devices to
    leak refs. From Robert Shearman.

 3) Handle fragmented SKBs properly in macsec driver, the problem is
    that we were mis-sizing the sgvec table. From Jason A. Donenfeld.

 4) We cannot have checksum offload enabled for inner UDP tunneled
    packet during IPSEC, from Ansis Atteka.

 5) Fix double SKB free in ravb driver, from Dan Carpenter.

 6) Fix CPU port handling in b53 DSA driver, from Florian Dainelli.

 7) Don't use on-stack buffers for usb_control_msg() in CAN usb driver,
    from Maksim Salau.

 8) Fix device leak in macvlan driver, from Herbert Xu. We have to purge
    the broadcast queue properly on port destroy.

 9) Fix tx ring entry limit on EF10 devices in sfc driver. From Bert
    Kenward.

10) Fix memory leaks in team driver, from Pan Bian.

11) Don't setup ipv6_stub before it can be actually used, from Paolo
    Abeni.

12) Fix tipc socket flow control accounting, from Parthasarathy
    Bhuvaragan.

13) Fix crash on module unload in hso driver, from Andreas Kemnade.

14) Fix purging of bridge multicast entries, the problem is that if we
    don't defer it to ndo_uninit it's possible for new entries to get
    added after we purge. Fix from Xin Long.

15) Don't return garbage for PACKET_HDRLEN getsockopt, from Alexander
    Potapenko.

16) Fix autoneg stall properly in PHY layer, and revert micrel driver
    change that was papering over it. From Alexander Kochetkov.

17) Don't dereference an ipv4 route as an ipv6 one in the ip6_tunnnel
    code, from Cong Wang.

18) Clear out the congestion control private of the TCP socket in all of
    the right places, from Wei Wang.

19) rawv6_ioctl measures SKB length incorrectly, fix from Jamie
    Bainbridge.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (41 commits)
  ipv6: check raw payload size correctly in ioctl
  tcp: memset ca_priv data to 0 properly
  ipv6: check skb->protocol before lookup for nexthop
  net: core: Prevent from dereferencing null pointer when releasing SKB
  macsec: dynamically allocate space for sglist
  Revert "phy: micrel: Disable auto negotiation on startup"
  net: phy: fix auto-negotiation stall due to unavailable interrupt
  net/packet: check length in getsockopt() called with PACKET_HDRLEN
  net: ipv6: regenerate host route if moved to gc list
  bridge: move bridge multicast cleanup to ndo_uninit
  ipv6: fix source routing
  qed: Fix error in the dcbx app meta data initialization.
  netvsc: fix calculation of available send sections
  net: hso: fix module unloading
  tipc: fix socket flow control accounting error at tipc_recv_stream
  tipc: fix socket flow control accounting error at tipc_send_stream
  ipv6: move stub initialization after ipv6 setup completion
  team: fix memory leaks
  sfc: tx ring can only have 2048 entries for all EF10 NICs
  macvlan: Fix device ref leak when purging bc_queue
  ...

7 years agoipv6: check raw payload size correctly in ioctl
Jamie Bainbridge [Wed, 26 Apr 2017 00:43:27 +0000 (10:43 +1000)]
ipv6: check raw payload size correctly in ioctl

In situations where an skb is paged, the transport header pointer and
tail pointer can be the same because the skb contents are in frags.

This results in ioctl(SIOCINQ/FIONREAD) incorrectly returning a
length of 0 when the length to receive is actually greater than zero.

skb->len is already correctly set in ip6_input_finish() with
pskb_pull(), so use skb->len as it always returns the correct result
for both linear and paged data.

Signed-off-by: Jamie Bainbridge <jbainbri@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: memset ca_priv data to 0 properly
Wei Wang [Wed, 26 Apr 2017 00:38:02 +0000 (17:38 -0700)]
tcp: memset ca_priv data to 0 properly

Always zero out ca_priv data in tcp_assign_congestion_control() so that
ca_priv data is cleared out during socket creation.
Also always zero out ca_priv data in tcp_reinit_congestion_control() so
that when cc algorithm is changed, ca_priv data is cleared out as well.
We should still zero out ca_priv data even in TCP_CLOSE state because
user could call connect() on AF_UNSPEC to disconnect the socket and
leave it in TCP_CLOSE state and later call setsockopt() to switch cc
algorithm on this socket.

Fixes: 2b0a8c9ee ("tcp: add CDG congestion control")
Reported-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Wei Wang <weiwan@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoipv6: check skb->protocol before lookup for nexthop
WANG Cong [Tue, 25 Apr 2017 21:37:15 +0000 (14:37 -0700)]
ipv6: check skb->protocol before lookup for nexthop

Andrey reported a out-of-bound access in ip6_tnl_xmit(), this
is because we use an ipv4 dst in ip6_tnl_xmit() and cast an IPv4
neigh key as an IPv6 address:

        neigh = dst_neigh_lookup(skb_dst(skb),
                                 &ipv6_hdr(skb)->daddr);
        if (!neigh)
                goto tx_err_link_failure;

        addr6 = (struct in6_addr *)&neigh->primary_key; // <=== HERE
        addr_type = ipv6_addr_type(addr6);

        if (addr_type == IPV6_ADDR_ANY)
                addr6 = &ipv6_hdr(skb)->daddr;

        memcpy(&fl6->daddr, addr6, sizeof(fl6->daddr));

Also the network header of the skb at this point should be still IPv4
for 4in6 tunnels, we shold not just use it as IPv6 header.

This patch fixes it by checking if skb->protocol is ETH_P_IPV6: if it
is, we are safe to do the nexthop lookup using skb_dst() and
ipv6_hdr(skb)->daddr; if not (aka IPv4), we have no clue about which
dest address we can pick here, we have to rely on callers to fill it
from tunnel config, so just fall to ip6_route_output() to make the
decision.

Fixes: ea3dc9601bda ("ip6_tunnel: Add support for wildcard tunnel endpoints.")
Reported-by: Andrey Konovalov <andreyknvl@google.com>
Tested-by: Andrey Konovalov <andreyknvl@google.com>
Cc: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agovirtio-net: on tx, only call napi_disable if tx napi is on
Willem de Bruijn [Tue, 25 Apr 2017 19:59:17 +0000 (15:59 -0400)]
virtio-net: on tx, only call napi_disable if tx napi is on

As of tx napi, device down (`ip link set dev $dev down`) hangs unless
tx napi is enabled. Else napi_enable is not called, so napi_disable
will spin on test_and_set_bit NAPI_STATE_SCHED.

Only call napi_disable if tx napi is enabled.

Fixes: 5a719c2552ca ("virtio-net: transmit napi")
Reported-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'ibmvnic-Move-sub-crq-init-out-of-interrupt-context'
David S. Miller [Wed, 26 Apr 2017 18:49:16 +0000 (14:49 -0400)]
Merge branch 'ibmvnic-Move-sub-crq-init-out-of-interrupt-context'

Nathan Fontenot says:

====================
ibmvnic: Move sub crq init out of interrupt context

The sub crqs are currently intialized in interrupt context when
handling a crq response fromn the vios server. There is no reason
they must be initialized there.

Moving the initialization of the sub crqs to the ibmvnic_init routine
allows us to do the initialization outside of interrupt context and
make all of the allocations with GFP_KERNEL instead of GFP_ATOMIC.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoibmvnic: Move initialization of sub crqs to ibmvnic_init
Nathan Fontenot [Tue, 25 Apr 2017 19:01:10 +0000 (15:01 -0400)]
ibmvnic: Move initialization of sub crqs to ibmvnic_init

The sub crq structures are initialized in interrupt context while
handling the response to crqs when negotiating capabilities for
the driver. The sub crqs do not need to be initialized at this point
and can be moved to being done from ibmvnic_init. Moving the init
of the sub crqs to ibmvnic_init also allows use to allocate the
memory with GFP_KERNEL instead of GFP_ATOMIC.

Signed-off-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoibmvnic: Split initialization of scrqs to its own routine
Nathan Fontenot [Tue, 25 Apr 2017 19:01:04 +0000 (15:01 -0400)]
ibmvnic: Split initialization of scrqs to its own routine

Split the sending of capability request crqs and the initialization
of sub crqs into their own routines. This is a first step to moving
the allocation of sub-crqs out of interrupt context.

Signed-off-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: core: Prevent from dereferencing null pointer when releasing SKB
Myungho Jung [Tue, 25 Apr 2017 18:58:15 +0000 (11:58 -0700)]
net: core: Prevent from dereferencing null pointer when releasing SKB

Added NULL check to make __dev_kfree_skb_irq consistent with kfree
family of functions.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=195289
Signed-off-by: Myungho Jung <mhjungk@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agodt-bindings: mdio: Clarify binding document
Florian Fainelli [Tue, 25 Apr 2017 18:33:03 +0000 (11:33 -0700)]
dt-bindings: mdio: Clarify binding document

The described GPIO reset property is applicable to *all* child PHYs. If
we have one reset line per PHY present on the MDIO bus, these
automatically become properties of the child PHY nodes.

Finally, indicate how the RESET pulse width must be defined, which is
the maximum value of all individual PHYs RESET pulse widths determined
by reading their datasheets.

Fixes: 69226896ad63 ("mdio_bus: Issue GPIO RESET to PHYs.")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge branch 'tcp-do-not-use-tcp_time_stamp-for-rcv-autotuning'
David S. Miller [Wed, 26 Apr 2017 18:44:39 +0000 (14:44 -0400)]
Merge branch 'tcp-do-not-use-tcp_time_stamp-for-rcv-autotuning'

Eric Dumazet says:

====================
tcp: do not use tcp_time_stamp for rcv autotuning

Some devices or linux distributions use HZ=100 or HZ=250

TCP receive buffer autotuning has poor behavior caused by this choice.
Since autotuning happens after 4 ms or 10 ms, short distance flows
get their receive buffer tuned to a very high value, but after an initial
period where it was frozen to (too small) initial value.

With BBR (or other CC allowing to increase BDP), we are willing to
increase tcp_rmem[2], but this receive autotuning defect is a blocker
for hosts dealing with gazillions of TCP flows in the data centers,
since many of them have inflated RCVBUF. Risk of OOM is too high.

Note that TSO autodefer, tcp cubic, and TCP TS options (RFC 7323)
also suffer from our dependency to jiffies (via tcp_time_stamp).

We have ongoing efforts to improve all that in the future.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: switch rcv_rtt_est and rcvq_space to high resolution timestamps
Eric Dumazet [Tue, 25 Apr 2017 17:15:41 +0000 (10:15 -0700)]
tcp: switch rcv_rtt_est and rcvq_space to high resolution timestamps

Some devices or distributions use HZ=100 or HZ=250

TCP receive buffer autotuning has poor behavior caused by this choice.
Since autotuning happens after 4 ms or 10 ms, short distance flows
get their receive buffer tuned to a very high value, but after an initial
period where it was frozen to (too small) initial value.

With tp->tcp_mstamp introduction, we can switch to high resolution
timestamps almost for free (at the expense of 8 additional bytes per
TCP structure)

Note that some TCP stacks use usec TCP timestamps where this
patch makes even more sense : Many TCP flows have < 500 usec RTT.
Hopefully this finer TS option can be standardized soon.

Tested:
 HZ=100 kernel
 ./netperf -H lpaa24 -t TCP_RR -l 1000 -- -r 10000,10000 &

 Peer without patch :
 lpaa24:~# ss -tmi dst lpaa23
 ...
 skmem:(r0,rb8388608,...)
 rcv_rtt:10 rcv_space:3210000 minrtt:0.017

 Peer with the patch :
 lpaa23:~# ss -tmi dst lpaa24
 ...
 skmem:(r0,rb428800,...)
 rcv_rtt:0.069 rcv_space:30000 minrtt:0.017

We can see saner RCVBUF, and more precise rcv_rtt information.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: remove ack_time from struct tcp_sacktag_state
Eric Dumazet [Tue, 25 Apr 2017 17:15:40 +0000 (10:15 -0700)]
tcp: remove ack_time from struct tcp_sacktag_state

It is no longer needed, everything uses tp->tcp_mstamp instead.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: use tp->tcp_mstamp in tcp_clean_rtx_queue()
Eric Dumazet [Tue, 25 Apr 2017 17:15:39 +0000 (10:15 -0700)]
tcp: use tp->tcp_mstamp in tcp_clean_rtx_queue()

Following patch will remove ack_time from struct tcp_sacktag_state

Same info is now found in tp->tcp_mstamp

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: do not pass timestamp to tcp_rack_advance()
Eric Dumazet [Tue, 25 Apr 2017 17:15:38 +0000 (10:15 -0700)]
tcp: do not pass timestamp to tcp_rack_advance()

No longer needed, since tp->tcp_mstamp holds the information.

This is needed to remove sack_state.ack_time in a following patch.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: do not pass timestamp to tcp_rate_gen()
Eric Dumazet [Tue, 25 Apr 2017 17:15:37 +0000 (10:15 -0700)]
tcp: do not pass timestamp to tcp_rate_gen()

No longer needed, since tp->tcp_mstamp holds the information.

This is needed to remove sack_state.ack_time in a following patch.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: do not pass timestamp to tcp_fastretrans_alert()
Eric Dumazet [Tue, 25 Apr 2017 17:15:36 +0000 (10:15 -0700)]
tcp: do not pass timestamp to tcp_fastretrans_alert()

Not used anymore now tp->tcp_mstamp holds the information.

This is needed to remove sack_state.ack_time in a following patch.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: do not pass timestamp to tcp_rack_identify_loss()
Eric Dumazet [Tue, 25 Apr 2017 17:15:35 +0000 (10:15 -0700)]
tcp: do not pass timestamp to tcp_rack_identify_loss()

Not used anymore now tp->tcp_mstamp holds the information.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: do not pass timestamp to tcp_rack_mark_lost()
Eric Dumazet [Tue, 25 Apr 2017 17:15:34 +0000 (10:15 -0700)]
tcp: do not pass timestamp to tcp_rack_mark_lost()

This is no longer used, since tcp_rack_detect_loss() takes
the timestamp from tp->tcp_mstamp

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: do not pass timestamp to tcp_rack_detect_loss()
Eric Dumazet [Tue, 25 Apr 2017 17:15:33 +0000 (10:15 -0700)]
tcp: do not pass timestamp to tcp_rack_detect_loss()

We can use tp->tcp_mstamp as it contains a recent timestamp.

This removes a call to skb_mstamp_get() from tcp_rack_reo_timeout()

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agotcp: add tp->tcp_mstamp field
Eric Dumazet [Tue, 25 Apr 2017 17:15:32 +0000 (10:15 -0700)]
tcp: add tp->tcp_mstamp field

We want to use precise timestamps in TCP stack, but we do not
want to call possibly expensive kernel time services too often.

tp->tcp_mstamp is guaranteed to be updated once per incoming packet.

We will use it in the following patches, removing specific
skb_mstamp_get() calls, and removing ack_time from
struct tcp_sacktag_state.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agomacsec: dynamically allocate space for sglist
Jason A. Donenfeld [Tue, 25 Apr 2017 17:08:18 +0000 (19:08 +0200)]
macsec: dynamically allocate space for sglist

We call skb_cow_data, which is good anyway to ensure we can actually
modify the skb as such (another error from prior). Now that we have the
number of fragments required, we can safely allocate exactly that amount
of memory.

Fixes: c09440f7dcb3 ("macsec: introduce IEEE 802.1AE driver")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Acked-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agorhashtable: remove insecure_max_entries param
Florian Westphal [Tue, 25 Apr 2017 09:41:34 +0000 (11:41 +0200)]
rhashtable: remove insecure_max_entries param

no users in the tree, insecure_max_entries is always set to
ht->p.max_size * 2 in rhtashtable_init().

Replace only spot that uses it with a ht->p.max_size check.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoRevert "phy: micrel: Disable auto negotiation on startup"
David S. Miller [Wed, 26 Apr 2017 18:33:14 +0000 (14:33 -0400)]
Revert "phy: micrel: Disable auto negotiation on startup"

This reverts commit 99f81afc139c6edd14d77a91ee91685a414a1c66.

It was papering over the real problem, which is fixed by commit
f555f34fdc58 ("net: phy: fix auto-negotiation stall due to unavailable
interrupt")

Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agonet: phy: fix auto-negotiation stall due to unavailable interrupt
Alexander Kochetkov [Thu, 20 Apr 2017 11:00:04 +0000 (14:00 +0300)]
net: phy: fix auto-negotiation stall due to unavailable interrupt

The Ethernet link on an interrupt driven PHY was not coming up if the Ethernet
cable was plugged before the Ethernet interface was brought up.

The patch trigger PHY state machine to update link state if PHY was requested to
do auto-negotiation and auto-negotiation complete flag already set.

During power-up cycle the PHY do auto-negotiation, generate interrupt and set
auto-negotiation complete flag. Interrupt is handled by PHY state machine but
doesn't update link state because PHY is in PHY_READY state. After some time
MAC bring up, start and request PHY to do auto-negotiation. If there are no new
settings to advertise genphy_config_aneg() doesn't start PHY auto-negotiation.
PHY continue to stay in auto-negotiation complete state and doesn't fire
interrupt. At the same time PHY state machine expect that PHY started
auto-negotiation and is waiting for interrupt from PHY and it won't get it.

Fixes: 321beec5047a ("net: phy: Use interrupts when available in NOLINK state")
Signed-off-by: Alexander Kochetkov <al.kochet@gmail.com>
Cc: stable <stable@vger.kernel.org> # v4.9+
Tested-by: Roger Quadros <rogerq@ti.com>
Tested-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
7 years agoMerge tag 'sound-4.11' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Linus Torvalds [Wed, 26 Apr 2017 16:30:33 +0000 (09:30 -0700)]
Merge tag 'sound-4.11' of git://git./linux/kernel/git/tiwai/sound

Pull sound fixes from Takashi Iwai:
 "Since we got a bonus week, let me try to screw a few pending fixes.

  A slightly large fix is the locking fix in ASoC STI driver, but it's
  pretty board-specific, and the risk is fairly low.

  All the rest are small / trivial fixes, mostly marked as stable, for
  ALSA sequencer core, ASoC topology, ASoC Intel bytcr and Firewire
  drivers"

* tag 'sound-4.11' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ASoC: intel: Fix PM and non-atomic crash in bytcr drivers
  ALSA: firewire-lib: fix inappropriate assignment between signed/unsigned type
  ALSA: seq: Don't break snd_use_lock_sync() loop by timeout
  ASoC: topology: Fix to store enum text values
  ASoC: STI: Fix null ptr deference in IRQ handler
  ALSA: oxfw: fix regression to handle Stanton SCS.1m/1d

7 years agoMerge tag 'iwlwifi-next-for-kalle-2017-04-26' of git://git.kernel.org/pub/scm/linux...
Kalle Valo [Wed, 26 Apr 2017 11:21:00 +0000 (14:21 +0300)]
Merge tag 'iwlwifi-next-for-kalle-2017-04-26' of git://git./linux/kernel/git/iwlwifi/iwlwifi-next

One more pull-request intended for 4.12.  These are the changes:

 * The firmware for 7265D and 3168 NICs is frozen at version 29;
 * Sari continues working heavily on support for A000 series;
 * A bunch of fixes;
 * Some cleanups here and there;

7 years agobrcmfmac: Make skb header writable before use
James Hughes [Tue, 25 Apr 2017 09:15:06 +0000 (10:15 +0100)]
brcmfmac: Make skb header writable before use

The driver was making changes to the skb_header without
ensuring it was writable (i.e. uncloned).
This patch also removes some boiler plate header size
checking/adjustment code as that is also handled by the
skb_cow_header function used to make header writable.

Signed-off-by: James Hughes <james.hughes@raspberrypi.org>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
7 years agobrcmfmac: Ensure pointer correctly set if skb data location changes
James Hughes [Mon, 24 Apr 2017 11:40:50 +0000 (12:40 +0100)]
brcmfmac: Ensure pointer correctly set if skb data location changes

The incoming skb header may be resized if header space is
insufficient, which might change the data adddress in the skb.
Ensure that a cached pointer to that data is correctly set by
moving assignment to after any possible changes.

Signed-off-by: James Hughes <james.hughes@raspberrypi.org>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
7 years agorndis_wlan: add return value validation
Pan Bian [Mon, 24 Apr 2017 00:40:28 +0000 (08:40 +0800)]
rndis_wlan: add return value validation

Function create_singlethread_workqueue() will return a NULL pointer if
there is no enough memory, and its return value should be validated
before using. However, in function rndis_wlan_bind(), its return value
is not checked. This may cause NULL dereference bugs. This patch fixes
it.

Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
7 years agolibertas: check return value of alloc_workqueue
Pan Bian [Sun, 23 Apr 2017 13:19:38 +0000 (21:19 +0800)]
libertas: check return value of alloc_workqueue

Function alloc_workqueue() will return a NULL pointer if there is no
enough memory, and its return value should be validated before using.
However, in function if_spi_probe(), its return value is not checked.
This may result in a NULL dereference bug. This patch fixes the bug.

Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
7 years agomt7601u: check return value of alloc_skb
Pan Bian [Sun, 23 Apr 2017 07:00:23 +0000 (15:00 +0800)]
mt7601u: check return value of alloc_skb

Function alloc_skb() will return a NULL pointer if there is no enough
memory. However, in function mt7601u_mcu_msg_alloc(), its return value
is not validated before it is used. This patch fixes it.

Signed-off-by: Pan Bian <bianpan2016@163.com>
Acked-by: Jakub Kicinski <kubakici@wp.pl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>