Merge branch 'master' of git://git./linux/kernel/git/linville/wireless-next into for-davem

Conflicts:
drivers/net/wireless/iwlwifi/dvm/testmode.c
drivers/net/wireless/iwlwifi/pcie/trans.c

iwlwifi: unlock on error path

We introduced a lock here in ff1ffb850b ("iwlwifi: fix dynamic
loading").  But we missed an error path which needs an unlock.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Merge branch 'master' of git://git./linux/kernel/git/linville/wireless

Conflicts:
drivers/net/wireless/ath/ath9k/main.c
net/bluetooth/hci_event.c

Merge branch 'for-upstream' of git://git./linux/kernel/git/bluetooth/bluetooth-next

ath9k_hw: remove MCI_STATE_SET_BT_SLEEP

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: remove MCI_STATE_CONT_* state

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: remove MCI_STATE_BT

remove MCI_STATE_BT and use bt_state instead.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: remove p_data argument from ar9003_mci_state

As p_data is unuse, lets remove it from ar9003_mci_state.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: remove MCI_STATE_NEED_FLUSH_BT_INFO

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: remove MCI_STATE_SEND_WLAN_CHANNELS

Add a MCI util function to send wlan channel info to BT.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: add utility function to set BT version

Add a utility function to set bluetooth version and remove
MCI_STATE_SET_BT_COEX_VERSION.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: cleanup MCI gpm offset state

Add utility functions to get and test GPM offset and
remove MCI_STATE*_GPM_OFFSET states.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Fix softlockup in AR9485

steps to recreate:
load latest ath9k driver with AR9485
stop the network-manager and wpa_supplicant
bring the interface up

Call Trace:
[<ffffffffa0517490>] ? ath_hw_check+0xe0/0xe0 [ath9k]
[<ffffffff812cd1e8>] __const_udelay+0x28/0x30
[<ffffffffa03bae7a>] ar9003_get_pll_sqsum_dvc+0x4a/0x80 [ath9k_hw]
[<ffffffffa05174eb>] ath_hw_pll_work+0x5b/0xe0 [ath9k]
[<ffffffff810744fe>] process_one_work+0x11e/0x470
[<ffffffff8107530f>] worker_thread+0x15f/0x360
[<ffffffff810751b0>] ? manage_workers+0x230/0x230
[<ffffffff81079af3>] kthread+0x93/0xa0
[<ffffffff815fd3a4>] kernel_thread_helper+0x4/0x10
[<ffffffff81079a60>] ? kthread_freezable_should_stop+0x70/0x70
[<ffffffff815fd3a0>] ? gs_change+0x13/0x13

ensure that the PLL-WAR for AR9485/AR9340 is executed only if the STA is
associated (or) IBSS/AP mode had started beaconing. Ideally this WAR
is needed to recover from some rare beacon stuck during stress testing.
Before the STA is associated/IBSS had started beaconing, PLL4(0x1618c)
always seem to have zero even though we had configured PLL3(0x16188) to
query about PLL's locking status. When we keep on polling infinitely PLL4's
8th bit(ie check for PLL locking measurements is done), machine hangs
due to softlockup.

fixes https://bugzilla.redhat.com/show_bug.cgi?id=811142

Reported-by: Rolf Offermanns <rolf.offermanns@gmx.net>
Cc: stable@vger.kernel.org
Tested-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: fix btcoex duty cycle

* Reset duty cycle before updating btcoex scheme. Otherwise duty cycle
  reaches max limit and never be reduced again
* Adjust duty cycle with proper BDR profile value

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: defer btcoex scheme update

As btcoex scheme updation might sleep, remove the function call
from tasklet context and queue it up as a separate work.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: keep btcoex period in milliseconds

btcoex periord is converted into micro seconds during initialization
and converted back to milli seconds while starting timer. As MCI code
handles btcoex period in msec, lets keep the btcoex timer in msec and
convert them into other form whenever needed.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: simplify btcoex profile management

This patch simplifies profile management utility functions.

* Separate find_profile from add/del functions
* Return correct values when the profile list is empty or
  profile is ot found
* flush the profiles when there are entries in the list

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: process MCI interrupts only when btcoex is enabled

let us process MCI interrupts only when BTCOEX is enabled to avoid
processing bogus interrupts.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: program OBS register only when MCI is disabled

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: fix incorrect LNA register settings

After a full reset, mci_reset will put LNA update to the setting
for 2G mode. Those registers need to be forced to update when
the channel is in 5G.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: check GPM HW write pointer before chip reset

Both "MAC Warm Reset" and "MCI Reset Rx" will reset GPM HW write_ptr.
We should check software cached write_ptr against HW write_ptr before
reset. Otherwise the pending DMA data will be lost.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: Fix AR9462 power consumption on idle associated

The HW statemachine is sometimes found stuck in the state
WL_LNA_CTRL_DISABLE when BT is in sleep, which will cause
TX_HOLD always asserted and resmgr stuck in PENDING_TX state

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: restore power state on set channel failure

Not doing so, could cause imbalance in powersave count.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: choose legacy rate as last rate of MRR series

Choose legacy rate as the last rate of Multi Rate Retry series
if and only if the last selected rate is MCS and having higher
PER rate. The current code fills a legacy rate as last one even
though the previous rates in the series are having good PER value.
This could limit the aggregation that affects the uplink performance.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: Update initvals for AR9462

Adjust quick_drop value in the baseband AGC register to
improve RX in HT20 mode.

Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

brcmfmac: expose sdio internal counters in debugfs

The structure brcmf_sdio contains a number of counters that are useful
for debugging. These were not available in user-space. This patch
exposes them in debugfs under the filename 'counters'.

Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com>
Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com>
Signed-off-by: Arend van Spriel <arend@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

brcmfmac: add debugfs helper functions

This patch adds debugfs support to brcmfmac. It provide helper functions
to setup the debugfs folder structure for the driver, which has following
hierarchy:

<debugfs_mount>/brcmfmac/<dev_name>/

ie.: /sys/kernel/debug/brcmfmac/mmc0:0001:2/

The new source file provides functions to create and remove the two
folders and a function to retrieve the device-specific folder so files
can be created in it.

Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com>
Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com>
Signed-off-by: Arend van Spriel <arend@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

brcmsmac: fix smatch warning found in ampdu.c

This patch fixes potential NULL pointer dereference in ampdu. This
was found running smatch static code checker. Smatch warning says:

drivers/net/wireless/brcm80211/brcmsmac/ampdu.c:741 brcms_c_sendampdu()
 warn: variable dereferenced before check 'p'

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com>
Signed-off-by: Arend van Spriel <arend@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

brcmsmac: remove brcms_set_hint() function

The function brcms_set_hint() does not add any functionality
so regulatory_hint() can be called directly. The error value
has been removed from the message when regulatory_hint() fails.

Reported-by: Seth Forshee <seth.forshee@canonical.com>
Signed-off-by: Arend van Spriel <arend@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Fix softlockup in AR9485

steps to recreate:
load latest ath9k driver with AR9485
stop the network-manager and wpa_supplicant
bring the interface up

Call Trace:
[<ffffffffa0517490>] ? ath_hw_check+0xe0/0xe0 [ath9k]
[<ffffffff812cd1e8>] __const_udelay+0x28/0x30
[<ffffffffa03bae7a>] ar9003_get_pll_sqsum_dvc+0x4a/0x80 [ath9k_hw]
[<ffffffffa05174eb>] ath_hw_pll_work+0x5b/0xe0 [ath9k]
[<ffffffff810744fe>] process_one_work+0x11e/0x470
[<ffffffff8107530f>] worker_thread+0x15f/0x360
[<ffffffff810751b0>] ? manage_workers+0x230/0x230
[<ffffffff81079af3>] kthread+0x93/0xa0
[<ffffffff815fd3a4>] kernel_thread_helper+0x4/0x10
[<ffffffff81079a60>] ? kthread_freezable_should_stop+0x70/0x70
[<ffffffff815fd3a0>] ? gs_change+0x13/0x13

ensure that the PLL-WAR for AR9485/AR9340 is executed only if the STA is
associated (or) IBSS/AP mode had started beaconing. Ideally this WAR
is needed to recover from some rare beacon stuck during stress testing.
Before the STA is associated/IBSS had started beaconing, PLL4(0x1618c)
always seem to have zero even though we had configured PLL3(0x16188) to
query about PLL's locking status. When we keep on polling infinitely PLL4's
8th bit(ie check for PLL locking measurements is done), machine hangs
due to softlockup.

fixes https://bugzilla.redhat.com/show_bug.cgi?id=811142

Reported-by: Rolf Offermanns <rolf.offermanns@gmx.net>
Cc: stable@vger.kernel.org [3.0+]
Tested-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Merge branch 'for-john' of git://git./linux/kernel/git/jberg/mac80211

mac80211: stop polling in disassociation

Stop connection monitor poll during disassociation.
This clears the polling flags and if a scan was
deferred it will be run.

Without this fix, if a scan was deferred due to
connection monitoring while disassociation happens,
this scan blocks further scan requests until interface
down/up which causes problems connecting to another AP.

Signed-off-by: David Spinadel <david.spinadel@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

mac80211: check sdata_running on ieee80211_set_bitrate_mask

Otherwise, we might call the driver callback before
the interface was uploaded.

Solves the following warning:
WARNING: at net/mac80211/driver-ops.h:12 ieee80211_set_bitrate_mask+0xbc/0x18c [mac80211]()
wlan0:  Failed check-sdata-in-driver check, flags: 0x0
Modules linked in: wlcore_sdio wl12xx wl18xx wlcore mac80211 cfg80211 [last unloaded: cfg80211]
[<c001b964>] (unwind_backtrace+0x0/0x12c) from [<c0495550>] (dump_stack+0x20/0x24)
[<c0495550>] (dump_stack+0x20/0x24) from [<c003ee28>] (warn_slowpath_common+0x5c/0x74)
[<c003ee28>] (warn_slowpath_common+0x5c/0x74) from [<c003eefc>] (warn_slowpath_fmt+0x40/0x48)
[<c003eefc>] (warn_slowpath_fmt+0x40/0x48) from [<bf5c1ad0>] (ieee80211_set_bitrate_mask+0xbc/0x18c [mac80211])
[<bf5c1ad0>] (ieee80211_set_bitrate_mask+0xbc/0x18c [mac80211]) from [<bf575960>] (nl80211_set_tx_bitrate_mask+0x350/0x358 [cfg80211])
[<bf575960>] (nl80211_set_tx_bitrate_mask+0x350/0x358 [cfg80211]) from [<c03e9e94>] (genl_rcv_msg+0x1a8/0x1e8)
[<c03e9e94>] (genl_rcv_msg+0x1a8/0x1e8) from [<c03e9164>] (netlink_rcv_skb+0x5c/0xc0)
[<c03e9164>] (netlink_rcv_skb+0x5c/0xc0) from [<c03e9ce0>] (genl_rcv+0x28/0x34)
[<c03e9ce0>] (genl_rcv+0x28/0x34) from [<c03e8e74>] (netlink_unicast+0x158/0x234)
[<c03e8e74>] (netlink_unicast+0x158/0x234) from [<c03e93e0>] (netlink_sendmsg+0x218/0x298)
[<c03e93e0>] (netlink_sendmsg+0x218/0x298) from [<c03b4e5c>] (sock_sendmsg+0xa4/0xc0)
[<c03b4e5c>] (sock_sendmsg+0xa4/0xc0) from [<c03b5af4>] (__sys_sendmsg+0x1d8/0x254)
[<c03b5af4>] (__sys_sendmsg+0x1d8/0x254) from [<c03b5ca8>] (sys_sendmsg+0x4c/0x70)
[<c03b5ca8>] (sys_sendmsg+0x4c/0x70) from [<c0013980>] (ret_fast_syscall+0x0/0x3c)

Note that calling the driver can also result
in undefined behaviour since it doesn't have
to deal with calls while down.

Signed-off-by: Eliad Peller <eliad@wizery.com>
[removed timestamps, added note - Johannes]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

cfg80211: fix potential deadlock in regulatory

reg_timeout_work() calls restore_regulatory_settings() which
takes cfg80211_mutex.

reg_set_request_processed() already holds cfg80211_mutex
before calling cancel_delayed_work_sync(reg_timeout),
so it might deadlock.

Call the async cancel_delayed_work instead, in order
to avoid the potential deadlock.

This is the relevant lockdep warning:

cfg80211: Calling CRDA for country: XX

======================================================
[ INFO: possible circular locking dependency detected ]
3.4.0-rc5-wl+ #26 Not tainted
-------------------------------------------------------
kworker/0:2/1391 is trying to acquire lock:
 (cfg80211_mutex){+.+.+.}, at: [<bf28ae00>] restore_regulatory_settings+0x34/0x418 [cfg80211]

but task is already holding lock:
 ((reg_timeout).work){+.+...}, at: [<c0059e94>] process_one_work+0x1f0/0x480

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 ((reg_timeout).work){+.+...}:
       [<c008fd44>] validate_chain+0xb94/0x10f0
       [<c0090b68>] __lock_acquire+0x8c8/0x9b0
       [<c0090d40>] lock_acquire+0xf0/0x114
       [<c005b600>] wait_on_work+0x4c/0x154
       [<c005c000>] __cancel_work_timer+0xd4/0x11c
       [<c005c064>] cancel_delayed_work_sync+0x1c/0x20
       [<bf28b274>] reg_set_request_processed+0x50/0x78 [cfg80211]
       [<bf28bd84>] set_regdom+0x550/0x600 [cfg80211]
       [<bf294cd8>] nl80211_set_reg+0x218/0x258 [cfg80211]
       [<c03c7738>] genl_rcv_msg+0x1a8/0x1e8
       [<c03c6a00>] netlink_rcv_skb+0x5c/0xc0
       [<c03c7584>] genl_rcv+0x28/0x34
       [<c03c6720>] netlink_unicast+0x15c/0x228
       [<c03c6c7c>] netlink_sendmsg+0x218/0x298
       [<c03933c8>] sock_sendmsg+0xa4/0xc0
       [<c039406c>] __sys_sendmsg+0x1e4/0x268
       [<c0394228>] sys_sendmsg+0x4c/0x70
       [<c0013840>] ret_fast_syscall+0x0/0x3c

-> #1 (reg_mutex){+.+.+.}:
       [<c008fd44>] validate_chain+0xb94/0x10f0
       [<c0090b68>] __lock_acquire+0x8c8/0x9b0
       [<c0090d40>] lock_acquire+0xf0/0x114
       [<c04734dc>] mutex_lock_nested+0x48/0x320
       [<bf28b2cc>] reg_todo+0x30/0x538 [cfg80211]
       [<c0059f44>] process_one_work+0x2a0/0x480
       [<c005a4b4>] worker_thread+0x1bc/0x2bc
       [<c0061148>] kthread+0x98/0xa4
       [<c0014af4>] kernel_thread_exit+0x0/0x8

-> #0 (cfg80211_mutex){+.+.+.}:
       [<c008ed58>] print_circular_bug+0x68/0x2cc
       [<c008fb28>] validate_chain+0x978/0x10f0
       [<c0090b68>] __lock_acquire+0x8c8/0x9b0
       [<c0090d40>] lock_acquire+0xf0/0x114
       [<c04734dc>] mutex_lock_nested+0x48/0x320
       [<bf28ae00>] restore_regulatory_settings+0x34/0x418 [cfg80211]
       [<bf28b200>] reg_timeout_work+0x1c/0x20 [cfg80211]
       [<c0059f44>] process_one_work+0x2a0/0x480
       [<c005a4b4>] worker_thread+0x1bc/0x2bc
       [<c0061148>] kthread+0x98/0xa4
       [<c0014af4>] kernel_thread_exit+0x0/0x8

other info that might help us debug this:

Chain exists of:
  cfg80211_mutex --> reg_mutex --> (reg_timeout).work

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock((reg_timeout).work);
                               lock(reg_mutex);
                               lock((reg_timeout).work);
  lock(cfg80211_mutex);

 *** DEADLOCK ***

2 locks held by kworker/0:2/1391:
 #0:  (events){.+.+.+}, at: [<c0059e94>] process_one_work+0x1f0/0x480
 #1:  ((reg_timeout).work){+.+...}, at: [<c0059e94>] process_one_work+0x1f0/0x480

stack backtrace:
[<c001b928>] (unwind_backtrace+0x0/0x12c) from [<c0471d3c>] (dump_stack+0x20/0x24)
[<c0471d3c>] (dump_stack+0x20/0x24) from [<c008ef70>] (print_circular_bug+0x280/0x2cc)
[<c008ef70>] (print_circular_bug+0x280/0x2cc) from [<c008fb28>] (validate_chain+0x978/0x10f0)
[<c008fb28>] (validate_chain+0x978/0x10f0) from [<c0090b68>] (__lock_acquire+0x8c8/0x9b0)
[<c0090b68>] (__lock_acquire+0x8c8/0x9b0) from [<c0090d40>] (lock_acquire+0xf0/0x114)
[<c0090d40>] (lock_acquire+0xf0/0x114) from [<c04734dc>] (mutex_lock_nested+0x48/0x320)
[<c04734dc>] (mutex_lock_nested+0x48/0x320) from [<bf28ae00>] (restore_regulatory_settings+0x34/0x418 [cfg80211])
[<bf28ae00>] (restore_regulatory_settings+0x34/0x418 [cfg80211]) from [<bf28b200>] (reg_timeout_work+0x1c/0x20 [cfg80211])
[<bf28b200>] (reg_timeout_work+0x1c/0x20 [cfg80211]) from [<c0059f44>] (process_one_work+0x2a0/0x480)
[<c0059f44>] (process_one_work+0x2a0/0x480) from [<c005a4b4>] (worker_thread+0x1bc/0x2bc)
[<c005a4b4>] (worker_thread+0x1bc/0x2bc) from [<c0061148>] (kthread+0x98/0xa4)
[<c0061148>] (kthread+0x98/0xa4) from [<c0014af4>] (kernel_thread_exit+0x0/0x8)
cfg80211: Calling CRDA to update world regulatory domain
cfg80211: World regulatory domain updated:
cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
cfg80211:   (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
cfg80211:   (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
cfg80211:   (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
cfg80211:   (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
cfg80211:   (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)

Cc: stable@kernel.org
Signed-off-by: Eliad Peller <eliad@wizery.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: WARN only once when we have trouble in reclaim

This flow can actually happen due to a corner case in
mac80211: the station is deleted before we get a chance
to reclaim all the packets in flight in AGG queue.
The tid_data for this station is zeroed, and we lose
the match with the Tx queue.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: use request_module instead of _nowait

Since request_module_nowait() can't be backported
use request_module() instead -- we don't need the
asynchronous behaviour of request_module_nowait()
here since we're running in the firmware request
work struct.

Tested-by: Donald H Fry <donald.h.fry@intel.com>
Reviewed-by: Donald H Fry <donald.h.fry@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: fix 6035 device parameters

Due to commit 26a7ca9a71a ("iwlwifi: refactor EEPROM
reading/parsing") adding a new parameter, while commit
d2c8b15d0cb ("iwlwifi: use correct supported firmware
for 6035 and 6000g2") added a new device structure we
need to add the parameter to the new device structure
to make 6035 device work.

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

Merge remote-tracking branch 'wireless-next/master' into iwlwifi-next

iwlwifi: warn if TFD index and WiFi Seq don't match

For AGG queues, we must match between the WiFi sequence
number and the TFD number. This is a HW (SCD) requirement.

This is a take two of my

    iwlwifi: add debug in Tx path in AGG flow

    This will allow us to catch bad cases in which the packets aren't in
    the right place on the ring.

which disappeared during code move.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: don't modify the timer if we don't Tx

In fragmentation we don't update the write pointer of the
HW immediately. So we shouldn't modify the timer in that
case.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: turn on a lockdep assertion

CMD_SYNC is zero so the if (cmd->flags & CMD_SYNC) is never true and we
never check the assertion.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: print even more info when a queue is stuck

Since the queue gets stuck from time to time, we are trying
to get as much information as we can when this occurs.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: Fix Makefile build order for built-in driver

When the driver is built into the kernel instead of a module
when the system boots it results in a panic. The order things are built in
results in their initialization order when built into the kernel. Wifi
has to be initialized before mvm or dvm.

Reviewed-by: Donald H Fry <donald.h.fry@intel.com>
Tested-by: Donald H Fry <donald.h.fry@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Brandon Misemer <brandon.misemer@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: comment that setting driver_data overrides info->control

Using the driver_data area in ieee80211_tx_info which
resides in the CB overrides the info->control field.
Add a comment to prevent mistakes.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: remove lock around txq_enable

This locking isn't needed. The only locking we need is when
we access prph registers but there is already a separate
lock for that.
Since we haven't returned from the mac80211's
IEEE80211_AMPDU_TX_OPERATIONAL ampdu_action, we cannot
receive any Tx frame for that sta / tid while enabling the
queue.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: don't configure a txq that is being disabled

This is not needed, we just need to tell the SCD not to use
that queue. We will reconfigure that queue when we will use
it again.

Clean up a bit the code on the way.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: print more info when a queue is stuck

Print some more info from the SCD's SRAM and dump the TRB
from the FH.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: iwl_{read,write}_targ_mem_words takes dwords

Change its name to better reflect this.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: s/iwl_read_targ_mem_words/iwl_read_targ_mem_bytes

This macro gets the bufsize in bytes.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: fix dynamic loading

Add locking to the dynamic loading code to prevent
corrupting the list if multiple device ever init at
the same time (which cannot happen for multiple PCI
devices, but could happen when different busses init
concurrently.)

Also remove a device from the list when it stops so
the list isn't left corrupted, including a fix from
Don to not crash when it was never added.

Reviewed-by: Donald H Fry <donald.h.fry@intel.com>
Tested-by: Donald H Fry <donald.h.fry@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Don Fry <donald.h.fry@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Merge git://git./linux/kernel/git/davem/net

Conflicts:
MAINTAINERS
drivers/net/wireless/iwlwifi/pcie/trans.c

The iwlwifi conflict was resolved by keeping the code added
in 'net' that turns off the buggy chip feature.

The MAINTAINERS conflict was merely overlapping changes, one
change updated all the wireless web site URLs and the other
changed some GIT trees to be Johannes's instead of John's.

Signed-off-by: David S. Miller <davem@davemloft.net>

ethtool: Make more commands available to unprivileged processes

'Get' commands should generally not require CAP_NET_ADMIN, with
the exception of those that expose internal state.

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net-next: add dev_loopback_xmit() to avoid duplicate code

Add dev_loopback_xmit() in order to deduplicate functions
ip_dev_loopback_xmit() (in net/ipv4/ip_output.c) and
ip6_dev_loopback_xmit() (in net/ipv6/ip6_output.c).

I was about to reinvent the wheel when I noticed that
ip_dev_loopback_xmit() and ip6_dev_loopback_xmit() do exactly what I
need and are not IP-only functions, but they were not available to reuse
elsewhere.

ip6_dev_loopback_xmit() does not have line "skb_dst_force(skb);", but I
understand that this is harmless, and should be in dev_loopback_xmit().

Signed-off-by: Michel Machado <michel@digirati.com.br>
CC: "David S. Miller" <davem@davemloft.net>
CC: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
CC: James Morris <jmorris@namei.org>
CC: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
CC: Patrick McHardy <kaber@trash.net>
CC: Eric Dumazet <edumazet@google.com>
CC: Jiri Pirko <jpirko@redhat.com>
CC: "Michał Mirosław" <mirq-linux@rere.qmqm.pl>
CC: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding: remove packet cloning in recv_probe()

Cloning all packets in input path have a significant cost.

Use skb_header_pointer()/skb_copy_bits() instead of pskb_may_pull() so
that recv_probe handlers (bond_3ad_lacpdu_recv / bond_arp_rcv /
rlb_arp_recv ) dont touch input skb.

bond_handle_frame() can avoid the skb_clone()/dev_kfree_skb()

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jay Vosburgh <fubar@us.ibm.com>
Cc: Andy Gospodarek <andy@greyhouse.net>
Cc: Jiri Bohac <jbohac@suse.cz>
Cc: Nicolas de Pesloüan <nicolas.2p.debian@free.fr>
Cc: Maciej Żenczykowski <maze@google.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

usbnet: don't initialize transfer buffer before submit status URB

The line below in intr_complete isn't needed,

memset(urb->transfer_buffer, 0, urb->transfer_buffer_length);

so just remove it.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

usbnet: remove declaration for intr_complete

Remove declaration for intr_complete so that ctags may be happy to
decrease duplicated symbols, also decrease one line code.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

usbnet: remove flag of EVENT_DEV_WAKING

The flag of EVENT_DEV_WAKING is not used any more, so just remove it.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

usbnet:cdc-phonet: remove usb_get/put_dev in .probe and .disconnect

usb_device is parent device of usb_interface in the view of driver
model, so its reference count is always held during .probe/.disconnect
of usb_interface instance.

This patch just removes the unnecessay usb_get/put_dev.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

usbnet:pegasus: remove usb_get/put_dev in .probe and .disconnect

usb_device is parent device of usb_interface in the view of driver
model, so its reference count is always held during .probe/.disconnect
of usb_interface instance.

This patch just removes the unnecessay usb_get/put_dev.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

usbnet: remove usb_get/put_dev in .probe and .disconnect

usb_device is parent device of usb_interface in the view of driver
model, so its reference count is always held during .probe/.disconnect
of usb_interface instance.

This patch just removes the unnecessay usb_get/put_dev.

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding: Fix corrupted queue_mapping

In the transmit path of the bonding driver, skb->cb is used to
stash the skb->queue_mapping so that the bonding device can set its
own queue mapping.  This value becomes corrupted since the skb->cb is
also used in __dev_xmit_skb.

When transmitting through bonding driver, bond_select_queue is
called from dev_queue_xmit.  In bond_select_queue the original
skb->queue_mapping is copied into skb->cb (via bond_queue_mapping)
and skb->queue_mapping is overwritten with the bond driver queue.

Subsequently in dev_queue_xmit, __dev_xmit_skb is called which writes
the packet length into skb->cb, thereby overwriting the stashed
queue mappping.  In bond_dev_queue_xmit (called from hard_start_xmit),
the queue mapping for the skb is set to the stashed value which is now
the skb length and hence is an invalid queue for the slave device.

If we want to save skb->queue_mapping into skb->cb[], best place is to
add a field in struct qdisc_skb_cb, to make sure it wont conflict with
other layers (eg : Qdiscc, Infiniband...)

This patchs also makes sure (struct qdisc_skb_cb)->data is aligned on 8
bytes :

netem qdisc for example assumes it can store an u64 in it, without
misalignment penalty.

Note : we only have 20 bytes left in (struct qdisc_skb_cb)->data[].
The largest user is CHOKe and it fills it.

Based on a previous patch from Tom Herbert.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Tom Herbert <therbert@google.com>
Cc: John Fastabend <john.r.fastabend@intel.com>
Cc: Roland Dreier <roland@kernel.org>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv4: Add interface option to enable routing of 127.0.0.0/8

Routing of 127/8 is tradtionally forbidden, we consider
packets from that address block martian when routing and do
not process corresponding ARP requests.

This is a sane default but renders a huge address space
practically unuseable.

The RFC states that no address within the 127/8 block should
ever appear on any network anywhere but it does not forbid
the use of such addresses outside of the loopback device in
particular. For example to address a pool of virtual guests
behind a load balancer.

This patch adds a new interface option 'route_localnet'
enabling routing of the 127/8 address block and processing
of ARP requests on a specific interface.

Note that for the feature to work, the default local route
covering 127/8 dev lo needs to be removed.

Example:
  $ sysctl -w net.ipv4.conf.eth0.route_localnet=1
  $ ip route del 127.0.0.0/8 dev lo table local
  $ ip addr add 127.1.0.1/16 dev eth0
  $ ip route flush cache

V2: Fix invalid check to auto flush cache (thanks davem)

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding:record primary when modify it via sysfs

If we modify primary via sysfs and it is not a valid slave,
we should record it for future use, and this behavior is the same with
bond_check_params().

Signed-off-by: Weiping Pan <wpan@redhat.com>
Acked-by: Nicolas de Pesloüan <nicolas.2p.debian@free.fr>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'master' of git://1984.lsi.us.es/net

mwifiex: fix incorrect privacy setting in beacon and probe response

Test procedure:
1. Start AP with security setting (e.g. WPA2)
2. Stop AP
3. Start AP with open security

Here it's observed that privacy is enabled in beacons and
probe responses.

This patch fixes it by checking the privacy parameter from
cfg80211_ap_settings. If privacy is not set in cfg80211_ap_settings,
set open authentication and no encryption in FW.

Signed-off-by: Avinash Patil <patila@marvell.com>
Signed-off-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Merge branch 'master' of git://git./linux/kernel/git/linville/wireless-next into for-davem

Merge branch 'master' of git://git./linux/kernel/git/linville/wireless

Conflicts:
MAINTAINERS
drivers/net/wireless/iwlwifi/pcie/trans.c

mac80211: add missing kernel-doc

Add a few kernel-doc descriptions that were missed
during mesh development.

Reported-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Ashok Nagarajan <ashok@cozybit.com>
Acked-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

tilegx network driver: initial support

This change adds support for the tilegx network driver based on the
GXIO IORPC support in the tilegx software stack, using the on-chip
mPIPE packet processing engine.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

phy: Use pr_<level>

Use a more current logging style.

Add pr_fmt and missing newlines.
Remove embedded prefixes.
Neaten phy_print_status to avoid using KERN_CONT.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

tg3: Apply short DMA frag workaround to 5906

5906 devices also need the short DMA fragment workaround.  This patch
makes the necessary change.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Tested-by: Christian Kujau <lists@nerdbynature.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

af_packet: use sizeof instead of constant in spkt_device

This small patch removes access to the last element of the spkt_device
array through a constant. Instead, it is accessed by sizeof() to respect
possible changes in if_packet.h.

Signed-off-by: Daniel Borkmann <daniel.borkmann@tik.ee.ethz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: stmmac: Fix clock en-/disable calls

clk_{un}prepare is mandatory for platforms using common clock framework.
Since these drivers are used by SPEAr platform, which supports common
clock framework, add clk_{un}prepare() support for them. Otherwise
the clocks are not correctly en-/disabled and ethernet support doesn't
work.

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Viresh Kumar <viresh.linux@gmail.com>
Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

inet: Fix BUG triggered by __rt{,6}_get_peer().

If no peer actually gets attached (either because create is zero or
the peer allocation fails) we'll trigger a BUG because we
unconditionally do an rt{,6}_peer_ptr() afterwards.

Fix this by guarding it with the proper check.

Signed-off-by: David S. Miller <davem@davemloft.net>

netfilter: nf_ct_tcp, udp: fix compilation with sysctl disabled

This patch fixes the compilation of the TCP and UDP trackers with sysctl
compilation disabled:

net/netfilter/nf_conntrack_proto_udp.c: In function ‘udp_init_net_data’:
net/netfilter/nf_conntrack_proto_udp.c:279:13: error: ‘struct nf_proto_net’ has no member named
 ‘user’
net/netfilter/nf_conntrack_proto_tcp.c:1606:9: error: ‘struct nf_proto_net’ has no member named
 ‘user’
net/netfilter/nf_conntrack_proto_tcp.c:1643:9: error: ‘struct nf_proto_net’ has no member named
 ‘user’

Reported-by: Fengguang Wu <wfg@linux.intel.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

lpc_eth: fix tx completion

__lpc_handle_xmit() has two bugs :

1) It can leak skbs in case TXSTATUS_ERROR is set

2) It can wake up txqueue while no slot was freed.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Roland Stigge <stigge@antcom.de>
Tested-by: Roland Stigge <stigge@antcom.de>
Cc: Kevin Wells <kevin.wells@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

lpc_eth: add missing ndo_change_mtu()

lpc_eth does a copy of transmitted skbs to DMA area, without checking
skb lengths, so can trigger buffer overflows :

memcpy(pldat->tx_buff_v + txidx * ENET_MAXF_SIZE, skb->data, len);

One way to get bigger skbs is to allow MTU changes above the 1500 limit.

Calling eth_change_mtu() in ndo_change_mtu() makes sure this cannot
happen.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Roland Stigge <stigge@antcom.de>
Cc: Kevin Wells <kevin.wells@nxp.com>
Acked-by: Roland Stigge <stigge@antcom.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: keep name_hlist close to name

__dev_get_by_name() is slow because pm_qos_req has been inserted between
name[] and name_hlist, adding cache misses.

pm_qos_req has nothing to do at the beginning of struct net_device

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'master' of git://1984.lsi.us.es/net-next

ssb: add missing PCI ID for b/g/n single band BCM4322

14e4:432c is found on some bcm63xx devices. The device is working fine
with b43.

Reported-by: Álvaro Fernández Rojas <noltari@gmail.com>
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: Initvals update for AR9462

MSI is enabled by default for most of the 4th generation
chips. Add this for AR9462 - this fixes PowerSave operation,
the chip was not entering Network-Sleep mode earlier.
With proper powering down of the MAC now, power consumption
in associated state is reduced considerably.

Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mwifiex: scan less channels per scan command to improve Tx traffic

Currently 4 channels are scanned per scan command. if scan request
is issued by user during Tx traffic, radio will be out of channel
for "4 * per_chan_scan_time" for each scan command and will not be
able to receive Rx packets. This adds delay in data traffic. We can
minimize it by reducing number of channels scanned per scan command
in this scenario.

We can not always scan 1 channel per scan command due to limitation
of number of command buffers. So we add code to decide number of
channels scanned per scan command in associated state.

Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mwifiex: fix simultaneous scan and Tx traffic problem

If scan operation is started when Tx traffic is already running,
driver locks Tx queue until it gets completed. With this logic
there is a delay for Tx packets.

This patch implements new approach to give Tx path higher priority
in this case. Driver internally sends multiple synchronous scan
commands to firmware when scan is requested by user. Now we will
make sure that Tx queue is empty everytime before sending next scan
command. If Tx queue isn't empty scan command will be postponsed by
20msec. This rule will be followed until Tx queue becomes empty or
timeout of 1 second happens. In case of timeout scan operation will
be aborted.

Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mwifiex: shorten per channel scan time

Currently the scan time per channel for active scanning is set to
200ms. It takes quite a while to finsh scanning on all channels,
especially with a dual band configuration.

Change the per channel scan time settings to the following values:

passive scan: 110ms
active scan: 30ms
specific scan: 30ms

Above settings have been tested on x86 and arm platforms.

Signed-off-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Merge branch 'for-john' of git://git./linux/kernel/git/iwlwifi/iwlwifi-next

Conflicts:
drivers/net/wireless/iwlwifi/iwl-eeprom.c

Merge tag 'nfc-next-3.6-1' of git://git./linux/kernel/git/sameo/nfc-3.0

Merge branch 'master' of git://git./linux/kernel/git/linville/wireless into for-davem

Merge branch 'for-upstream' of git://git./linux/kernel/git/bluetooth/bluetooth

rndis_wlan: fix matching bssid check in rndis_check_bssid_list()

rndis_check_bssid_list() originally tried to check if bssid->mac and
match_bssid are equal using compare_ether_addr() when it should use
!compare_ether_addr(). This check was added by commit
b5257c952dda24df7078c74b7b811b44c6e49206 as part of workaround for
hardware issue.

Commit 2e42e4747ea72943c21551d8a206b51a9893b1e0 that replaced
compare_ether_addr with ether_addr_equal relieved that this compare
to be inverse of what it should be.

Compare was added as response to hardware bug, where bssid-list does
not contain BSSID and other information of currently connected AP
(spec insists that device must provide this information in the list
when connected). Lack bssid-data on current connection then causes
WARN_ON somewhere in cfg80211. Workaround was to check if bssid-list
returns current bssid and if it does not, manually construct bssid
information in other ways. And this workaround worked, with inverse
check. Which must mean that when hardware is experiencing the problem,
it's actually returning empty bssid-list and this check didn't make
any difference for workaround.

However inverse check causes workaround be activated when bssid-list
returns only entry, currently connected BSSID. That does not cause
problems in itself, just slightly more inaccurate information in
scan-list.

Cc: Joe Perches <joe@perches.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: remove incompatible IBSS interface check in change_iface

'cfg80211: fix interface combinations' ensures that if an interface
type is not advertised by the driver in any of the interface combinations
(via ieee80211_iface_combination) then it shall be treated as a single
incompatible interface. if there are more than one interfaces present
and changing them to incompatible interface type is not possible.
These checks will be properly handled by cfg80211_change_iface ->
cfg80211_can_change_interface.

this patch is dependent on 'cfg80211: fix interface combinations'

Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Fix a WARNING on suspend/resume with IBSS

this patch is dependent on the patch "cfg80211: fix interface
combinations"

In ath9k currently we have ADHOC interface as a single incompatible
interface. when drv_add_interface is called during resume we got to
consider number of vifs already present in addition to checking the
drivers 'opmode' information about ADHOC.  we incorrectly assume
an ADHOC interface is already present. Then we may miss some driver
specific data for the ADHOC interface after resume.

The above mentioned checks can be removed from the driver,
as the patch 'cfg80211: fix interface combinations' ensures that
if an interface type is not advertised by the driver in any of the
interface combinations(via ieee80211_iface_combination) then it shall
be treated as a single incompatible interface. Fixes the following
warning on suspend/resume with ibss interface.

        ath: phy0: Cannot create ADHOC interface when other
        interfaces already exist.
        WARNING: at net/mac80211/driver-ops.h:12
        ieee80211_reconfig+0x1882/0x1ca0 [mac80211]()
        Hardware name: 2842RK1
        wlan2:  Failed check-sdata-in-driver check, flags: 0x0

        Call Trace:
        [<c01361b2>] warn_slowpath_common+0x72/0xa0
        [<f8aaa7c2>] ? ieee80211_reconfig+0x1882/0x1ca0
        [mac80211]
        [<f8aaa7c2>] ? ieee80211_reconfig+0x1882/0x1ca0
        [mac80211]
        [<c0136283>] warn_slowpath_fmt+0x33/0x40
        [<f8aaa7c2>] ieee80211_reconfig+0x1882/0x1ca0 [mac80211]
        [<c06c1d1a>] ? mutex_lock_nested+0x23a/0x2f0
        [<f8a95097>] ieee80211_resume+0x27/0x70 [mac80211]
        [<fd177edf>] wiphy_resume+0x8f/0xa0 [cfg80211]

Cc: stable@vger.kernel.org
Cc: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

inet: Avoid potential NULL peer dereference.

We handle NULL in rt{,6}_set_peer but then our caller will try to pass
that NULL pointer into inet_putpeer() which isn't ready for it.

Fix this by moving the NULL check one level up, and then remove the
now unnecessary NULL check from inetpeer_ptr_set_peer().

Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

iwlwifi: decouple testmode and iwl-test

The iwl-test flows were based on the cfg80211 testmode APIs.

To remove this coupling, the op mode (during the initialization
of the iwl_test object) is responsible to set the callbacks that
should be used by iwl-test to allocate skbs for events and replies
and to send events and replies.

The current op modes implement these callbacks based on the cfg80211
testmode APIs.

Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

iwlwifi: refactor testmode

Create an object that will enacpsulate the testmode functionality
that is common to all op modes.

 * Copy definitions from dvm/dev.h
 * Copy the testmode logic from dvm/testmode.c
 * Link iwl-test object into the iwlwifi module
 * Modify DVM to use iwl-test object

Reviewed-by: Amit Beka <amit.beka@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

inet: Use FIB table peer roots in routes.

Signed-off-by: David S. Miller <davem@davemloft.net>

inet: Add inetpeer tree roots to the FIB tables.

Signed-off-by: David S. Miller <davem@davemloft.net>

inet: Add family scope inetpeer flushes.

This implementation can deal with having many inetpeer roots, which is
a necessary prerequisite for per-FIB table rooted peer tables.

Each family (AF_INET, AF_INET6) has a sequence number which we bump
when we get a family invalidation request.

Each peer lookup cheaply checks whether the flush sequence of the
root we are using is out of date, and if so flushes it and updates
the sequence number.

Signed-off-by: David S. Miller <davem@davemloft.net>

ipv4: Kill ip_rt_frag_needed().

There is zero point to this function.

It's only real substance is to perform an extremely outdated BSD4.2
ICMP check, which we can safely remove.  If you really have a MTU
limited link being routed by a BSD4.2 derived system, here's a nickel
go buy yourself a real router.

The other actions of ip_rt_frag_needed(), checking and conditionally
updating the peer, are done by the per-protocol handlers of the ICMP
event.

TCP, UDP, et al. have a handler which will receive this event and
transmit it back into the associated route via dst_ops->update_pmtu().

This simplification is important, because it eliminates the one place
where we do not have a proper route context in which to make an
inetpeer lookup.

Signed-off-by: David S. Miller <davem@davemloft.net>

inet: Hide route peer accesses behind helpers.

We encode the pointer(s) into an unsigned long with one state bit.

The state bit is used so we can store the inetpeer tree root to use
when resolving the peer later.

Later the peer roots will be per-FIB table, and this change works to
facilitate that.

Signed-off-by: David S. Miller <davem@davemloft.net>

dummy: fix rcu_sched self-detected stalls

Trying to "modprobe dummy numdummies=30000" triggers :

INFO: rcu_sched self-detected stall on CPU { 8} (t=60000 jiffies)

After this splat, RTNL is locked and reboot is needed.

We must call cond_resched() to avoid this, even holding RTNL.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>