ath.git patches for 4.8. Major changes:

ath9k

* implement temperature compensation support for AR9003+

ath10k

* disable wake_tx_queue() mac80211 op for older devices to workaround
  throughput regression

For rtl8188ee, we pass -Idrivers/net/wireless/rtlwifi/ to gcc,
however that directy no longer exists, so evidently this option
is no longer required here and can be removed to avoid a warning
when building with 'make W=1' or 'gcc -Wmissing-include-dirs'
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

We dropped strict MIPS dependency for bcm47xxsflash driver in:
commit 5651d6aa ("mtd: bcm47xxsflash: use ioremap_cache() instead of
KSEG0ADDR()") but using ioremap_cache still limits building it to few
selected architectures only.

A recent commit 57d8f7dd ("bcma: allow enabling serial flash support
on non-MIPS SoCs") automatically dropped MIPS dependency for
MTD_BCM47XXSFLASH which broke building e.g. on powerpc and cris.

The bcma change is alright as it doesn't break building bcma code in any
way. MTD_BCM47XXSFLASH on the other hand should be limited to archs
which need it and can build it (by providing ioremap_cache).

Fixes: 57d8f7dd ("bcma: allow enabling serial flash support on non-MIPS SoCs")
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Cc: Brian Norris <computersforpeace@gmail.com>
Acked-by: NBrian Norris <computersforpeace@gmail.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

Add support for using with both wl12xx and wl18xx.

- all wilink family needs special init command for entering wspi mode.
  extra clock cycles should be sent after the spi init command while the
  cs pin is high.
- Use inverted chip select for sending a dummy 4 bytes command that
  completes the init stage.
Signed-off-by: NEyal Reizer <eyalr@ti.com>
Acked-by: NRob Herring <robh@kernel.org>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

wlc_phy_txpower_get_current() does a logical OR of power->flags, which
presumes that power.flags was initiliazed earlier by the caller,
unfortunately, this is not the case, so make sure we zero out the struct
tx_power before calling into wlc_phy_txpower_get_current().

Reported-by: coverity (CID 146011)
Fixes: 5b435de0 ("net: wireless: add brcm80211 drivers")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Acked-by: NArend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

In case dma_mapping_error() returns an error in dma_rxfill, we would be
leaking a packet that we allocated with brcmu_pkt_buf_get_skb().

Reported-by: coverity (CID 1081819)
Fixes: 67d0cf50 ("brcmsmac: Fix WARNING caused by lack of calls to dma_mapping_error()")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Acked-by: NArend van Spriel <arend@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

In case brcmf_sdiod_recv_chain() cannot complete a succeful call to
brcmf_sdiod_buffrw, we would be leaking glom_skb and not free it as we
should, fix this.

Reported-by: coverity (CID 1164856)
Fixes: a413e39a ("brcmfmac: fix brcmf_sdcard_recv_chain() for host without sg support")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Acked-by: NArend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

We don't have access to datasheets to document all the bits but we can
name these registers at least.
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

Signed-off-by: NFelix Fietkau <nbd@nbd.name>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Remove a misleading debug message as well
Signed-off-by: NFelix Fietkau <nbd@nbd.name>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Reduce indentation, use a variable to save a few pointer dereferences
Signed-off-by: NFelix Fietkau <nbd@nbd.name>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

AR_PHY_65NM_CH0_THERM and AR_CH0_THERM were supposed to refer to the
same register, however they had different SREV checks.

Remove the duplicate and use the checks. Since there were other SREV
checks present in the only place that uses this, this will probaby not
affect runtime behavior.
Signed-off-by: NFelix Fietkau <nbd@nbd.name>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

The register layout of AR_PHY_SPECTRAL_SCAN has changed, only AR9280
uses the old layout
Signed-off-by: NFelix Fietkau <nbd@nbd.name>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

There were two paths in the code for "external" eeprom sources. The code
in eeprom.c only handled the cases where the eeprom data was loaded via
request_firmware. ahb.c and pci.c on the other hand had some duplicate
code which was only used when the eeprom data was passed via
ath9k_platform_data.
With this change all eeprom data handling is now unified in eeprom.c.
Signed-off-by: NMartin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

When an NL80211_DISCONNECT is sent to cfg80211, the driver's cfg80211
disconnect function sets the sme_state to SME_DISCONNECTED before receiving
a WMI_DISCONNECT_EVENT from the firmware.  This caused cfg80211 to not know
that the connection is disconnected because the driver fails to notify
cfg80211 upon receiving WMI_DISCONNECT_EVENT from the firmware believing
it is already disconnected.
Signed-off-by: NDan Kephart <dan.kephart@lairdtech.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

When disabling tethering after having enabled it, the ath6kl driver does
not set the IFF_LOWER_UP flag to 0 (carrier off) on the wlan interface.

The upper layers (eg. connman) are not notified of the tethering mode
status change. So, tethering can not be activated anymore.

This patch adds a netif_carrier_off() call when stopping AP mode to fix
the problem.
Signed-off-by: NPierre Le Magourou <plemagourou@aldebaran.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

One minute after a successful connection, the kernel checks if the
frequency and the channel width are well configured for the country we
are in. (regulatory domain)

ath6kl driver was setting the NL80211_CHAN_HT20 channel without checking
for the HT capabilities. (we should have NL80211_CHAN_NO_HT in our
case because the firmware did not support HT)

This patch adds a check on ht_cap.ht_supported in order to create the
channel corresponding to the firmware capabilities.
Signed-off-by: NPierre Le Magourou <plemagourou@aldebaran.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

When enabling WLAN tethering, a new AP is visible and a STA could
connect to it. When the STA tries to authenticate to the newly created
AP, the WPA authentication mechanism is stuck in the 1/4 msg of 4-Way
Handshake.

In ath6kl_rx(), the ath6kl_find_sta() function is looking for the
h_source field of the Ethernet frame header received by the STA. The
datap pointer that points to the Ethernet frame header is incorrect,
and was pointing at the wrong offset in the buffer.

This commit adds a pad_before_data_start offset to set the datap pointer
to the Ethernet frame header. datap->h_source parameter is now really
pointing to the source ethernet address and the authentication process
can continue.
Signed-off-by: NPierre Le Magourou <plemagourou@aldebaran.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Ideally wake_tx_queue should be used regardless as
it is a requirement for reducing bufferbloat and
implementing airtime fairness in the future.

However some setups (typically low-end platforms
hosting QCA988X) suffer performance regressions
with the current wake_tx_queue implementation.
Therefore disable it unless it is really
beneficial with current codebase (which is when
firmware supports smart pull-push tx scheduling).
Signed-off-by: NMichal Kazior <michal.kazior@tieto.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

ath.git patches for 4.8. Major changes:

ath10k

* enable support for QCA9888

Merge tag 'iwlwifi-next-for-kalle-2016-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/iwlwifi-next

* bump firmware version to load newer firmwares

So far we had only MIPS devices with serial flash connected to the SoC's
ChipCommon. ARM devices got a separated SPI controller and weere using
standard SPI drivers.
This has changed with the wireless SoC BCM47189B0. It's ARM based but
has serial flash attached just like older devices. This allows using
existing driver with these devices.
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

Same skb is being reused for storing command response from firmware
in PCIe chipsets. There was a bug while updating the skb length.
This patch ensures skb length correctly gets updated based on rx_len.
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

In corner case, we may end up processing same interrupt twice.
We have a logic to read pending interrupts at the end of interrupt
processing routine. It has a race with interrupts read in interrupt
handler. This patch solves the problem by ORing the interrupt bitmap
in this case.

The symptom for this bug is below messages in dmesg log.

[   11.522123] mwifiex_pcie 0000:01:00.0: CMD_RESP: invalid cmd resp
[   11.680412] mwifiex_pcie 0000:01:00.0: There is no command but got cmdrsp

Link: https://bugzilla.kernel.org/show_bug.cgi?id=109681Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NCathy Luo <cluo@marvell.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

It's been observed that request_irq() failure leads to a system crash
due to a bug in mwifiex driver.
When this failure happens, mwifiex_add_card() already takes care of
clearing and freeing adapter->card pointer. This patch removes the
redundant cleanup code causing crash.
Signed-off-by: NKarthik D A <karthida@marvell.com>
Signed-off-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

The sdio dpc starts processing when in SDIOD_STATE_DATA. This state was
entered right after firmware download. This patch moves that transition
just before enabling sdio interrupt handling thus avoiding watchdog
expiry which would put the bus to sleep while probing.
Reviewed-by: NHante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: NPieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: NFranky Lin <franky.lin@broadcom.com>
Signed-off-by: NArend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

When the host-interface bus has hard time handling transmit packets
it informs higher layer about this and it would stop the netdev
queue when needed. However, since commit 9cd18359 ("brcmfmac:
Make FWS queueing configurable.") this was broken. With this patch
the behaviour is restored.

Cc: stable@vger.kernel.org # v4.5, v4.6, v4.7
Fixes: 9cd18359 ("brcmfmac: Make FWS queueing configurable.")
Tested-by: NPer Förlin <per.forlin@gmail.com>
Reviewed-by: NHante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: NPieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: NFranky Lin <franky.lin@broadcom.com>
Signed-off-by: NArend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

memory is malloced in mwifiex_cfg80211_start_ap() and should be
freed before leaving from the error handling cases, otherwise it
will cause memory leak.
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

Fixes the following sparse warning:

drivers/net/wireless/marvell/libertas/cfg.c:2047:5: warning:
 symbol 'lbs_set_power_mgmt' was not declared. Should it be static?
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

After discovering there are 2 very different 14e4:4365 PCI devices we
made ID tables less generic. Back then we believed there are only 2 such
devices:
1) 14e4:4365 1028:0016 with SoftMAC BCM43142 chipset
2) 14e4:4365 14e4:4365 with FullMAC BCM4366 chipset

>From the recent report it appears there is also 14e4:4365 105b:e092
which should be claimed by bcma. Add back support for it.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=121881
Fixes: 515b399c ("bcma: claim only 14e4:4365 PCI Dell card with SoftMAC BCM43142")
Reported-by: NIgor Mammedov <imammedo@redhat.com>
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Cc: Stable <stable@vger.kernel.org> [4.6+]
Tested-by: NIgor Mammedov <imammedo@redhat.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

Adding this opcode, allows the TI wireless driver,
to report throughput directly from FW to mac80211.

This is used mainly for mesh metric calculation.
Signed-off-by: NMaxim Altshul <maxim.altshul@ti.com>
[kvalo@codeaurora.org: fix indentation]
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

FW will provide a TX rate per link for each FW status,
and wlcore will be able to store the information for
the use of the mesh hwmp module.

This is used mainly in mesh.
Rates are reported when a mesh interface is up.
Signed-off-by: NMaxim Altshul <maxim.altshul@ti.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

This was succesfully tested with 4366B1. A small workaround is needed
for the main interface otherwise it would stuck at the hidden state.
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

1. Added support for interface and role of mesh type.
2. Enabled enable/start of mesh-point role,
   and opening and closing a connection with a mesh peer.
3. Added multirole combination of mesh and ap
   under the same limits of dual ap mode.
4. Add support for 'sta_rc_update' opcode for mesh IF.
   The 'sta_rc_update' opcode is being used in mesh_plink.c.
Add support in wlcore to handle this opcode correctly for mesh
(as opposed to current implementation that handles STA only).
5. Bumped the firmware version to support new Mesh functionality
Signed-off-by: NMaital Hahn <maitalm@ti.com>
Signed-off-by: NYaniv Machani <yanivma@ti.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

For multiple REG Host Commands (e.g HostCmd_CMD_802_11_EEPROM_ACCESS,
HostCmd_CMD_MAC_REG_ACCESS etc.) "cpu_to_leX"-converted values are
saved to driver. So, "leX_to_cpu" conversion is required too many
times afterwards in driver.

This patch reduces the endian: conversion without saving "cpu_to_leX"
converted values in driver. This will convert endianness in prepare
command and command response path.
Signed-off-by: NPrasun Maiti <prasunmaiti87@gmail.com>
Acked-by: NAmitkumar Karwar <akarwar@marvell.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

There are two generics functions phy_ethtool_{get|set}_link_ksettings,
so we can use them instead of defining the same code in the driver.
Signed-off-by: NPhilippe Reynes <tremyfr@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The private structure contain a pointer to phydev, but the structure
net_device already contain such pointer. So we can remove the pointer
phydev in the private structure, and update the driver to use the
one contained in struct net_device.
Signed-off-by: NPhilippe Reynes <tremyfr@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There are two generics functions phy_ethtool_{get|set}_link_ksettings,
so we can use them instead of defining the same code in the driver.

There was a check on CAP_NET_ADMIN in bfin_mac_ethtool_setsettings,
but this check is already done in dev_ethtool, so no need to repeat
it before calling the generic function.
Signed-off-by: NPhilippe Reynes <tremyfr@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The private structure contain a pointer to phydev, but the structure
net_device already contain such pointer. So we can remove the pointer
phydev in the private structure, and update the driver to use the
one contained in struct net_device.
Signed-off-by: NPhilippe Reynes <tremyfr@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

alloc_workqueue replaces deprecated create_singlethread_workqueue().

A dedicated workqueue has been used since the workitem viz
lp->txtimeout_reinit is involved in reinitialization if a TX timeout
occurs, which is necessary to guarantee forward progress in packet
processing. As a network device can be used during memory reclaim, the
workqueue needs forward progress guarantee under memory pressure.
WQ_MEM_RECLAIM has been set to ensure this.

Since there is only a single work item, explicit concurrency limit is
unnecessary here.
Signed-off-by: NBhaktipriya Shridhar <bhaktipriya96@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>