linux
6 years agowlcore: sdio: allow pm to handle sdio power
Eyal Reizer [Thu, 26 Apr 2018 05:47:11 +0000 (08:47 +0300)]
wlcore: sdio: allow pm to handle sdio power

pm_runtime handles sdio power on and power off transitions.
An old workaround for trying to control the power explicitly from the
driver was in fact causing failures on suspend/resume as the mmc layer
already power the module on resume.

In case of resume pm_runtime_get sync returns a positive device's usage
count causing the driver to try an re-initialize an already initialized
device. This was causing sdio bus failure on resume.

Remove this manual power on/off sequence as it is in-fact not needed.

Signed-off-by: Eyal Reizer <eyalr@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoqtnfmac: fix qtnf_netdev_hard_start_xmit()'s return type
Luc Van Oostenryck [Tue, 24 Apr 2018 13:18:04 +0000 (15:18 +0200)]
qtnfmac: fix qtnf_netdev_hard_start_xmit()'s return type

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Fix this by returning 'netdev_tx_t' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Reviewed-by: Sergey Matyukevich <sergey.matyukevich.os@quantenna.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agonet: wireless: b43legacy: Replace GFP_ATOMIC with GFP_KERNEL in dma_tx_fragment
Jia-Ju Bai [Tue, 10 Apr 2018 13:54:19 +0000 (21:54 +0800)]
net: wireless: b43legacy: Replace GFP_ATOMIC with GFP_KERNEL in dma_tx_fragment

dma_tx_fragment() is never called in atomic context.

dma_tx_fragment() is only called by b43legacy_dma_tx(), which is
only called by b43legacy_tx_work().
b43legacy_tx_work() is only set a parameter of INIT_WORK() in
b43legacy_wireless_init().

Despite never getting called from atomic context,
dma_tx_fragment() calls alloc_skb() with GFP_ATOMIC,
which does not sleep for allocation.
GFP_ATOMIC is not necessary and can be replaced with GFP_KERNEL,
which can sleep and improve the possibility of sucessful allocation.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: keep user configured mac address during changing virtual interface
Xinming Hu [Wed, 25 Apr 2018 09:38:14 +0000 (17:38 +0800)]
mwifiex: keep user configured mac address during changing virtual interface

During changing virtual interface, keep using previous net device
mac address.

Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: always configure firmware mac address during changing virtual interface
Xinming Hu [Wed, 25 Apr 2018 09:38:13 +0000 (17:38 +0800)]
mwifiex: always configure firmware mac address during changing virtual interface

When interface type changed, firmware using a new connction pointer.
We need Re-configure the mac address.

Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: make firmware mac address consistent with host configuration
Xinming Hu [Wed, 25 Apr 2018 09:38:12 +0000 (17:38 +0800)]
mwifiex: make firmware mac address consistent with host configuration

For user configurated mac address, directly set to firmware with no change.

Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: fix mwifiex_hard_start_xmit()'s return type
Luc Van Oostenryck [Tue, 24 Apr 2018 13:18:02 +0000 (15:18 +0200)]
mwifiex: fix mwifiex_hard_start_xmit()'s return type

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Fix this by returning 'netdev_tx_t' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: check for pending reset before attempting to schedule tx
Felix Fietkau [Wed, 25 Apr 2018 09:11:24 +0000 (11:11 +0200)]
mt76: check for pending reset before attempting to schedule tx

The check within mt76_txq_send_burst is not enough, as it happens after
a first frame has already been queued up

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: add rcu locking around tx scheduling
Felix Fietkau [Wed, 25 Apr 2018 09:11:23 +0000 (11:11 +0200)]
mt76: add rcu locking around tx scheduling

Fixes a reported lockdep error in mac80211:

[  179.867321] =============================
[  179.871510] WARNING: suspicious RCU usage
[  179.875528] 4.14.32 #0 Not tainted
[  179.878924] -----------------------------
[  179.882981] backports-2017-11-01/net/mac80211/tx.c:594 suspicious rcu_dereference_check() usage!
[  179.891785]
[  179.891785] other info that might help us debug this:
[  179.891785]
[  179.899824]
[  179.899824] rcu_scheduler_active = 2, debug_locks = 1
[  179.906343] 2 locks held by ksoftirqd/0/7:
[  179.910479]  #0:  (&(&q->lock)->rlock){+.-.}, at: [<86b207a4>] mt76_dma_tx_cleanup+0x64/0x354 [mt76]
[  179.919734]  #1:  (&(&fq->lock)->rlock){+.-.}, at: [<87238410>] ieee80211_tx_dequeue+0x54/0xc3c [mac80211]
[  179.929890]
[  179.929890] stack backtrace:
[  179.934257] CPU: 0 PID: 7 Comm: ksoftirqd/0 Not tainted 4.14.32 #0
[  179.940421] Stack : 00000000 00000000 00000000 00000000 80e0fce2 00000036 00000000 00000000
[  179.948864]         87c3d24c 80696377 8061039c 00000000 00000007 00000001 87c5db78 6534689d
[  179.957306]         00000000 00000000 80e10000 87c5da74 00000001 0000015a 00000007 00000000
[  179.965748]         00000000 806a0000 000e4171 00000000 00000000 00000000 ffffffff 00000001
[  179.974189]         806c0000 8692b240 86b000d0 87316fe4 00000001 802c9a68 00000000 80700000
[  179.982632]         ...
[  179.985104] Call Trace:
[  179.987582] [<80010a48>] show_stack+0x58/0x100
[  179.992040] [<804c2c58>] dump_stack+0xe8/0x170
[  179.996868] [<87234a04>] ieee80211_tx_h_select_key+0xa8/0x5b8 [mac80211]
[  180.004299] [<87238d44>] ieee80211_tx_dequeue+0x988/0xc3c [mac80211]
[  180.011048] [<86b230dc>] mt76_txq_schedule+0x110/0x3a4 [mt76]
[  180.016821] [<86b209d0>] mt76_dma_tx_cleanup+0x290/0x354 [mt76]
[  180.022777] [<86be2e60>] mt7603_tx_tasklet+0x40/0x6c [mt7603e]
[  180.028637] [<80037058>] tasklet_action+0x110/0x1ec
[  180.033532] [<804e1dac>] __do_softirq+0x164/0x35c
[  180.038235] [<80037174>] run_ksoftirqd+0x40/0x84
[  180.042870] [<800580c8>] smpboot_thread_fn+0x1a8/0x1d8
[  180.048023] [<800542e8>] kthread+0x130/0x144
[  180.052297] [<8000b1f8>] ret_from_kernel_thread+0x14/0x1c

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: add rcu locking in tid reorder function
Felix Fietkau [Wed, 25 Apr 2018 09:11:22 +0000 (11:11 +0200)]
mt76: add rcu locking in tid reorder function

Avoids having the tid or station entry disappear prematurely.
Also cancel the reorder work earlier to avoid further processing delayed
by waiting for the lock to be released

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: fix concurrent rx calls on A-MPDU release
Felix Fietkau [Wed, 25 Apr 2018 09:11:21 +0000 (11:11 +0200)]
mt76: fix concurrent rx calls on A-MPDU release

Add a spinlock in mt76_rx_complete. Without this, multiple stats updates
could happen in parallel, which can lead to deadlocks. There are
probably more corner cases fixed by this change.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: fix is_mt7612 routine
Lorenzo Bianconi [Tue, 17 Apr 2018 22:27:18 +0000 (00:27 +0200)]
mt76x2: fix is_mt7612 routine

Fix is_mt7612 routine since asic version is set in mt76_dev revision
field and not in mt76x2_dev one.
Moreover remove mt76x2_dev rev field since it is never used in the
driver

Fixes: 7bc04215a66b ('mt76: add driver code for MT76x2e')
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoMerge tag 'iwlwifi-next-for-kalle-2018-04-26' of git://git.kernel.org/pub/scm/linux...
Kalle Valo [Mon, 30 Apr 2018 10:02:30 +0000 (13:02 +0300)]
Merge tag 'iwlwifi-next-for-kalle-2018-04-26' of git://git./linux/kernel/git/iwlwifi/iwlwifi-next

Second batch of iwlwifi patches for 4.18

* Some preparations for new hardware;
* A workaround to rescan the bus for the rare situation when older
  devices become irresponsive,
* Hardening of the firmware loading code to avoid issues with
  corrupted files;
* A few clean-ups and bugfixes.

6 years agoiwlwifi: mvm: set wakeup filters for wowlan "any" configuration
Eliad Peller [Thu, 22 Feb 2018 16:45:49 +0000 (18:45 +0200)]
iwlwifi: mvm: set wakeup filters for wowlan "any" configuration

In case of "any" wowlan trigger is configured, no valid wakeup filter
was configured.

Moreover, the fw assumes there's no connection when there are no configured
wakeup filters.
This leads to the station info not being updated on D3 command, causing
rate_n_flags to be 0 when the offloading code sends tx frame (triggering
SYSASSERT_102C due to invalid antenna param)

Note: "any" trigger is currently assumed to only be used when entering
d0i3 (which has a different flow). However, we still reach this code
when using d3_test.

Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: mvm: remove check for non low latency TIDs
Sara Sharon [Sun, 4 Feb 2018 13:01:37 +0000 (15:01 +0200)]
iwlwifi: mvm: remove check for non low latency TIDs

Firmware will only send non low-latency TIDs in the
bitmap, so the check is now redundant.

Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: make bitfield a u32 instead of u16
Luca Coelho [Thu, 22 Feb 2018 12:50:06 +0000 (14:50 +0200)]
iwlwifi: make bitfield a u32 instead of u16

The bitfield we use in struct iwl_cfg contains 17 bits, but we declare
it as u16.  The compiler doesn't seem to have any problems with that
(it probably automatically makes it u32), but it's cleaner to use a
type that is long enough for all the flags.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: pcie: remove non-responsive device
Luca Coelho [Tue, 12 Dec 2017 06:58:41 +0000 (08:58 +0200)]
iwlwifi: pcie: remove non-responsive device

If we fail to to grab NIC access because the device is not responding
(i.e. CSR_GP_CNTRL returns 0xFFFFFFFF), remove the device from the PCI
bus, to avoid any further damage, and to let the user space rescan.

In order to inform the userspace that a rescan is needed, we send a
kobject uevent with "INACCESSIBLE".

This functionality is disabled by default, but can be enabled via a
new module parameter called "remove_when_gone".  In the future we may
change this module parameter to include 3 modes instead: do nothing;
auto-rescan or; send uevent.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Rajat Jain <rajatja@google.com>
6 years agoiwlwifi: fw: combine loading of last page block into main copy loop
Luca Coelho [Wed, 14 Feb 2018 06:54:40 +0000 (08:54 +0200)]
iwlwifi: fw: combine loading of last page block into main copy loop

Now that we check and copy only the actual size of the page block,
there is no need to treat the last block separately.  Remove the
mostly duplicate code and make the main copy loop handle also the last
block.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: fw: harden page loading code
Luca Coelho [Tue, 13 Feb 2018 09:09:40 +0000 (11:09 +0200)]
iwlwifi: fw: harden page loading code

The page loading code trusts the data provided in the firmware images
a bit too much and may cause a buffer overflow or copy unknown data if
the block sizes don't match what we expect.

To prevent potential problems, harden the code by checking if the
sizes we are copying are what we expect.

Cc: stable@vger.kernel.org
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: mvm: move skb padding reservation earlier
Johannes Berg [Tue, 20 Feb 2018 15:32:36 +0000 (16:32 +0100)]
iwlwifi: mvm: move skb padding reservation earlier

Future changes will require moving the HE radiotap data into
the SKB head, but this means we need to have the alignment
reservation done before that. To prepare, move the alignment
reservation earlier here.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: support new csr addresses for hw address
Golan Ben Ami [Wed, 22 Nov 2017 09:40:15 +0000 (11:40 +0200)]
iwlwifi: support new csr addresses for hw address

In future devices we use different csr addresses for hw addresses.
Update csr addresses to support new and legacy devices.

Signed-off-by: Golan Ben Ami <golan.ben.ami@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: pcie: allow sending pre-built A-MSDUs
Eliad Peller [Tue, 13 Feb 2018 16:41:46 +0000 (18:41 +0200)]
iwlwifi: pcie: allow sending pre-built A-MSDUs

In case of A-MSDUs, the trans layer is taking care of building
the subframes (out of the given skb), according to the given gso_size.

However, in some testing flows, we want to build the whole A-MSDU
frame in a different place (e.g. userspace), and ask the driver
to send it as-is.

In case of gso_size==0, simply treat the frame as normal-frame,
although the A-MSDU flag is set.

Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: cfg: remove unnecessary cfg data in non-dvm devices
Luca Coelho [Tue, 13 Feb 2018 11:52:43 +0000 (13:52 +0200)]
iwlwifi: cfg: remove unnecessary cfg data in non-dvm devices

The max_data_size and max_inst_size values are only needed for DVM
devices.  Remove the assignment to those fields in 7000 and newer
families so we can also remove the otherwise unnecessary inclusion of
iwl-agn.h headers.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: introduce Image Loader (IML) - new firmware image
Golan Ben-Ami [Mon, 17 Jul 2017 16:42:35 +0000 (19:42 +0300)]
iwlwifi: introduce Image Loader (IML) - new firmware image

In future devices a new image will be introduced - IML. The IML, image
loader, is loaded by the ROM, and as part of the new self-init flow,
loads the rest of the firmware images to the device.

Store the image, so the ROM can load it to the device.

Signed-off-by: Golan Ben Ami <golan.ben.ami@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoiwlwifi: allow different csr flags for different device families
Golan Ben Ami [Sun, 22 Oct 2017 12:58:26 +0000 (15:58 +0300)]
iwlwifi: allow different csr flags for different device families

Different device families may have different flag values
for passing a message to the fw (i.e. SW_RESET).
In order to keep the code readable, and avoid conditioning
upon the family, store a value for each flag, which indicates
the bit that needs to be enabled.

Signed-off-by: Golan Ben Ami <golan.ben.ami@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
6 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers.git
Kalle Valo [Thu, 26 Apr 2018 10:26:37 +0000 (13:26 +0300)]
Merge git://git./linux/kernel/git/kvalo/wireless-drivers.git

To fix a conflict reported by Stephen Rothwell <sfr@canb.auug.org.au>:

Today's linux-next merge of the wireless-drivers-next tree got a
conflict in:

  drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.c

between commit:

  77e30e10ee28 ("iwlwifi: mvm: query regdb for wmm rule if needed")

from the wireless-drivers tree and commits:

  9c4f7d512740 ("iwlwifi: move all NVM parsing code to the common files")
  4c625c564ba2 ("iwlwifi: get rid of fw/nvm.c")

from the wireless-drivers-next tree.

6 years agoMerge ath-next from git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git
Kalle Valo [Wed, 25 Apr 2018 08:30:54 +0000 (11:30 +0300)]
Merge ath-next from git://git./linux/kernel/git/kvalo/ath.git

ath.git patches for 4.18. Major changes:

ath10k

* enable temperature reads for QCA6174 and QCA9377

* add firmware memory dump support for QCA9984

* continue adding WCN3990 support via SNOC bus

6 years agoqtnfmac: pearl: pcie: fix memory leak in qtnf_fw_work_handler
Gustavo A. R. Silva [Thu, 5 Apr 2018 15:49:49 +0000 (10:49 -0500)]
qtnfmac: pearl: pcie: fix memory leak in qtnf_fw_work_handler

In case memory resources for fw were succesfully allocated, release
them before jumping to fw_load_fail.

Addresses-Coverity-ID: 1466092 ("Resource leak")
Fixes: c3b2f7ca4186 ("qtnfmac: implement asynchronous firmware loading")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Sergey Matyukevich <sergey.matyukevich.os@quantenna.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agort2x00: rt2800: add antenna diversity for RT5370G
Xose Vazquez Perez [Thu, 5 Apr 2018 12:22:32 +0000 (14:22 +0200)]
rt2x00: rt2800: add antenna diversity for RT5370G

RT5370G has hardware RX antenna diversity like RT5390R.
Based on DPO_RT5572_LinuxSTA_2.6.1.3_20121022.tar.bz2 manufacturer driver:
https://d86o2zu8ugzlg.cloudfront.net/mediatek-craft/drivers/DPO_RT5572_LinuxSTA_2.6.1.3_20121022.tar.bz2

Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Helmut Schaa <helmut.schaa@googlemail.com>
Cc: Gertjan van Wingerde <gwingerde@gmail.com>
Cc: Jakub Kicinski <jakub.kicinski@netronome.com>
Cc: wireless newbie <wnewbie72@gmail.com>
Cc: Kalle Valo <kvalo@codeaurora.org>
Cc: linux wireless ml <linux-wireless@vger.kernel.org>
Signed-off-by: Xose Vazquez Perez <xose.vazquez@gmail.com>
Acked-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agobrcmsmac: phy_lcn: remove duplicate code
Gustavo A. R. Silva [Thu, 5 Apr 2018 00:09:44 +0000 (19:09 -0500)]
brcmsmac: phy_lcn: remove duplicate code

Remove and refactor some code in order to avoid having identical code
for different branches.

Notice that this piece of code hasn't been modified since 2011.

Addresses-Coverity-ID: 1226756 ("Identical code for different branches")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agobrcmfmac: add support for BCM4366E chipset
Dan Haab [Tue, 3 Apr 2018 08:21:56 +0000 (10:21 +0200)]
brcmfmac: add support for BCM4366E chipset

BCM4366E is a wireless chipset with a BCM43664 ChipCommon. It's
supported by the same firmware as 4366c0.

Signed-off-by: Dan Haab <dan.haab@luxul.com>
[arend: rebase patch and remove unnecessary definition]
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: pcie: tighten a check in mwifiex_pcie_process_event_ready()
Dan Carpenter [Thu, 5 Apr 2018 11:17:19 +0000 (14:17 +0300)]
mwifiex: pcie: tighten a check in mwifiex_pcie_process_event_ready()

If "evt_len" is 1 then we try to memcpy() negative 3 bytes and it would
cause memory corruption.

Fixes: d930faee141b ("mwifiex: add support for Marvell pcie8766 chipset")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: uap: support cfg80211 ignore_broadcast_ssid=2
Xinming Hu [Mon, 2 Apr 2018 12:31:23 +0000 (20:31 +0800)]
mwifiex: uap: support cfg80211 ignore_broadcast_ssid=2

Firmware already support hidden ssid and keep ssid length,
Open the capability in driver.

Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomwifiex: uap: filter duplicate ERP IE
Xinming Hu [Mon, 2 Apr 2018 12:31:22 +0000 (20:31 +0800)]
mwifiex: uap: filter duplicate ERP IE

Firmware parse and attach ERP IE from bss configuration,
do not set again from tail IE.

Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: fix TXD_INFO bitmask definition
Lorenzo Bianconi [Mon, 16 Apr 2018 23:05:59 +0000 (01:05 +0200)]
mt76x2: fix TXD_INFO bitmask definition

Despite this issue is not harmful since it is not actually used in
mt76 driver, fix DMA txinfo bitmask definition

Fixes: 7bc04215a66b ('mt76: add driver code for MT76x2e')
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt7601u: run calibration works after finishing scanning
Stanislaw Gruszka [Mon, 16 Apr 2018 11:56:18 +0000 (13:56 +0200)]
mt7601u: run calibration works after finishing scanning

When finishing scanning we switch to operational channel sill with
SCANNING flag. This mean that we never perform calibration works after
scanning. To fix the problem queue calibration works on
.sw_scan_complete() routine.

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Jakub Kicinski <kubakici@wp.pl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt7601u: use EWMA to calculate avg_rssi
Stanislaw Gruszka [Mon, 16 Apr 2018 11:56:17 +0000 (13:56 +0200)]
mt7601u: use EWMA to calculate avg_rssi

avr_rssi is not calculated correctly as we do not divide result
by 256 (mt76 sum avg_rssi1 and avg_rssi2 and divide by 512).
However dividing by 256 will make avg_rssi almost the same as
last rssi value - not really an average. So use EWMA to calculate
avg_rssi. I've chosen weight_rcp=4 to convergence quicker on signal
strength changes.

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Jakub Kicinski <kubakici@wp.pl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: set default values in TX_ALC_CFG_{1, 2} for tempetaure compensation
Lorenzo Bianconi [Wed, 4 Apr 2018 08:38:32 +0000 (10:38 +0200)]
mt76x2: set default values in TX_ALC_CFG_{1, 2} for tempetaure compensation

Initialize default values for temperature compensation in TX_ALC_CFG_{1,2}
if tssi has been enabled

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: fix tx_alc_enabled check
Lorenzo Bianconi [Wed, 4 Apr 2018 08:23:46 +0000 (10:23 +0200)]
mt76x2: fix tx_alc_enabled check

Fix mt76x2_temp_tx_alc_enabled routine since in order to enable tx_alc
temperature compensation it necessary to take into account BIT(15) of
MT_EE_TX_POWER_EXT_PA_5G eeprom info

Fixes: 7bc04215a66b ("mt76: add driver code for MT76x2e")
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: remove unnecessary MT_TX_ALC_CFG_4 configuration
Lorenzo Bianconi [Tue, 3 Apr 2018 20:30:19 +0000 (22:30 +0200)]
mt76x2: remove unnecessary MT_TX_ALC_CFG_4 configuration

Remove unnecessary MT_TX_ALC_CFG_4 configuration since the register value
will be properly set according to the usage of an external power amplifier

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: check qos ack policy before reordering packets
Felix Fietkau [Tue, 3 Apr 2018 19:52:54 +0000 (21:52 +0200)]
mt76: check qos ack policy before reordering packets

Do not attempt to reorder packets not part of a BA session

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: set RX_FLAG_DUP_VALIDATED for A-MPDU reordered packets
Felix Fietkau [Tue, 3 Apr 2018 19:52:53 +0000 (21:52 +0200)]
mt76: set RX_FLAG_DUP_VALIDATED for A-MPDU reordered packets

Required for fast-rx and allows mac80211 to skip an unnnecessary check.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: fix potential sleep in atomic context
Felix Fietkau [Tue, 3 Apr 2018 19:52:52 +0000 (21:52 +0200)]
mt76: fix potential sleep in atomic context

Use cancel_delayed_work instead of cancel_delayed_work_sync

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: rework tx power handling
Felix Fietkau [Tue, 3 Apr 2018 19:52:51 +0000 (21:52 +0200)]
mt76: rework tx power handling

There were a number of issues in the existing tx power handling code:

1. The EEPROM target power for chain 0 refers to the actual output power
after the channel and bandwidth delta have been added to the initial
channel gain. This means the delta values should not be added/subtracted
for the rate power calculation

2. When power is reduced significantly, the initial channel gain
underflows very quickly, while the per-rate power offsets are high.
This miscalculation causes effective tx power to be increased when it
should actually be lowered.

Fix this by trying to adjust the channel gain to the lowest power from
the rate table. In case of under- or overflow, compensate by adding an
appropriate delta to the rate power values.

This makes power configuration more accurate on a wider range of
configurable values

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: toggle driver station powersave bit before notifying mac80211
Felix Fietkau [Tue, 3 Apr 2018 19:52:50 +0000 (21:52 +0200)]
mt76: toggle driver station powersave bit before notifying mac80211

Avoids race conditions from mac80211 enqueueing tx packets before the
tx-drop bit is cleared

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: add missing VHT maximum A-MPDU length capability
Felix Fietkau [Tue, 3 Apr 2018 19:52:49 +0000 (21:52 +0200)]
mt76: add missing VHT maximum A-MPDU length capability

Signficantly improves throughput with some clients

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76: stop tx queues from the driver callback instead of common code
Felix Fietkau [Tue, 3 Apr 2018 19:52:48 +0000 (21:52 +0200)]
mt76: stop tx queues from the driver callback instead of common code

Allows the driver to control whether to send a BlockAckReq after waking
up. MT76x2 needs this set to true, for MT7603 (to be submitted later)
it needs to be false.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: make mt76x2_mac_reset routine static
Lorenzo Bianconi [Tue, 3 Apr 2018 15:01:43 +0000 (17:01 +0200)]
mt76x2: make mt76x2_mac_reset routine static

Add static qualifier to mt76x2_mac_reset routine and remove the
prototype from mt76x2_mac.h since it is used just in mt76x2_init.c

Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt76x2: fix tssi initialization for 5GHz band
Lorenzo Bianconi [Tue, 3 Apr 2018 14:45:31 +0000 (16:45 +0200)]
mt76x2: fix tssi initialization for 5GHz band

Fix mcu initial configuration for tssi calibration on 5GHz band

Fixes: 7bc04215a66b ("mt76: add driver code for MT76x2e")
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agomt7601u: phy: mark expected switch fall-through
Gustavo A. R. Silva [Fri, 30 Mar 2018 21:12:23 +0000 (16:12 -0500)]
mt7601u: phy: mark expected switch fall-through

In preparation to enabling -Wimplicit-fallthrough, mark switch cases
where we are expecting to fall through.

Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Acked-by: Jakub Kicinski <kubakici@wp.pl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agortlwifi: btcoex: remove identical statements within if-else branches
Ping-Ke Shih [Mon, 9 Apr 2018 07:51:31 +0000 (15:51 +0800)]
rtlwifi: btcoex: remove identical statements within if-else branches

Since the statements are identical, we can safely remove the statements.
The commit 42e74946f016 ("rtlwifi: btcoexist: Fix if == else warnings in
halbtc8821a1ant.c") had fixed the statements, but the commit c6821613e653
("rtlwifi: btcoex: follow linux coding style") that converted coding style
and upgraded btcoex didn't include the fix.

Reported-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: Fix 'invalid vdd' warning in mmc
Siva Rebbagondla [Wed, 11 Apr 2018 06:43:32 +0000 (12:13 +0530)]
rsi: Fix 'invalid vdd' warning in mmc

While performing cleanup, driver is messing with card->ocr
value by not masking rocr against ocr_avail. Below panic
is observed with some of the SDIO host controllers due to
this. Issue is resolved by reverting incorrect modifications
to vdd.

[  927.423821] mmc1: Invalid vdd 0x1f
[  927.423925] Modules linked in: rsi_sdio(+) cmac bnep arc4 rsi_91x
       mac80211 cfg80211 btrsi rfcomm bluetooth ecdh_generic
[  927.424073] CPU: 0 PID: 1624 Comm: insmod Tainted: G W        4.15.0-1000-caracalla #1
[  927.424075] Hardware name: Dell Inc. Edge Gateway 3003/      , BIOS 01.00.06 01/22/2018
[  927.424082] RIP: 0010:sdhci_set_power_noreg+0xdd/0x190[sdhci]
[  927.424085] RSP: 0018:ffffac3fc064b930 EFLAGS:  00010282
[  927.424107] Call Trace:
[  927.424118]  sdhci_set_power+0x5a/0x60 [sdhci]
[  927.424125]  sdhci_set_ios+0x360/0x3b0 [sdhci]
[  927.424133]  mmc_set_initial_state+0x92/0x120
[  927.424137]  mmc_power_up.part.34+0x33/0x1d0
[  927.424141]  mmc_power_up+0x17/0x20
[  927.424147]  mmc_sdio_runtime_resume+0x2d/0x50
[  927.424151]  mmc_runtime_resume+0x17/0x20
[  927.424156]  __rpm_callback+0xc4/0x200
[  927.424161]  ? idr_alloc_cyclic+0x57/0xd0
[  927.424165]  ? mmc_runtime_suspend+0x20/0x20
[  927.424169]  rpm_callback+0x24/0x80
[  927.424172]  ? mmc_runtime_suspend+0x20/0x20
[  927.424176]  rpm_resume+0x4b3/0x6c0
[  927.424181]  __pm_runtime_resume+0x4e/0x80
[  927.424188]  driver_probe_device+0x41/0x490
[  927.424192]  __driver_attach+0xdf/0xf0
[  927.424196]  ? driver_probe_device+0x490/0x490
[  927.424201]  bus_for_each_dev+0x6c/0xc0
[  927.424205]  driver_attach+0x1e/0x20
[  927.424209]  bus_add_driver+0x1f4/0x270
[  927.424217]  ? rsi_sdio_ack_intr+0x50/0x50 [rsi_sdio]
[  927.424221]  driver_register+0x60/0xe0
[  927.424227]  ? rsi_sdio_ack_intr+0x50/0x50 [rsi_sdio]
[  927.424231]  sdio_register_driver+0x20/0x30
[  927.424237]  rsi_module_init+0x16/0x40 [rsi_sdio]

Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: fix nommu_map_sg overflow kernel panic
Siva Rebbagondla [Wed, 11 Apr 2018 06:43:31 +0000 (12:13 +0530)]
rsi: fix nommu_map_sg overflow kernel panic

Following overflow kernel panic is observed on some platforms while
loading the driver. It is fixed if dynamically allocated memory is
passed to SDIO instead of static one

[  927.513963] nommu_map_sg: overflow 17d54064ba7c+20 of device mask ffffffff
[  927.517712] Modules linked in: rsi_sdio(+) cmac bnep arc4 rsi_91x mac80211 cfg80211
       btrsi rfcomm bluetooth ecdh_generic snd_soc_sst_bytcr_rt5660
[  927.517861] CPU: 0 PID: 1624 Comm: insmod Tainted: G W 4.15.0-1000 #1
[  927.517870] RIP: 0010:sdhci_send_command+0x5f0/0xa90 [sdhci]
[  927.517873] RSP: 0000:ffffac3fc064b6d8 EFLAGS: 00010086
[  927.517895] Call Trace:
[  927.517908]  ? __schedule+0x3cd/0x890
[  927.517915]  ? mod_timer+0x17b/0x3c0
[  927.517922]  sdhci_request+0x7c/0xf0 [sdhci]
[  927.517928]  __mmc_start_request+0x5a/0x170
[  927.517932]  mmc_start_request+0x74/0x90
[  927.517936]  mmc_wait_for_req+0x87/0xe0
[  927.517940]  mmc_io_rw_extended+0x2fd/0x330
[  927.517946]  ? mmc_wait_data_done+0x30/0x30
[  927.517951]  sdio_io_rw_ext_helper+0x160/0x210
[  927.517956]  sdio_writesb+0x1d/0x20
[  927.517966] rsi_sdio_write_register_multiple+0x68/0x110 [rsi_sdio]
[  927.517976]  rsi_hal_device_init+0x357/0x910 [rsi_91x]
[  927.517983]  ? rsi_hal_device_init+0x357/0x910 [rsi_91x]
[  927.517990]  rsi_probe+0x2c6/0x450 [rsi_sdio]
[  927.517995]  sdio_bus_probe+0xfc/0x110
[  927.518000]  driver_probe_device+0x2b3/0x490
[  927.518005]  __driver_attach+0xdf/0xf0
[  927.518008]  ? driver_probe_device+0x490/0x490
[  927.518014]  bus_for_each_dev+0x6c/0xc0
[  927.518018]  driver_attach+0x1e/0x20
[  927.518021]  bus_add_driver+0x1f4/0x270
[  927.518028]  ? rsi_sdio_ack_intr+0x50/0x50 [rsi_sdio]
[  927.518031]  driver_register+0x60/0xe0
[  927.518038]  ? rsi_sdio_ack_intr+0x50/0x50 [rsi_sdio]
[  927.518041]  sdio_register_driver+0x20/0x30
[  927.518047]  rsi_module_init+0x16/0x40 [rsi_sdio]

Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: advertise 5GHz support based on device capability
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:17 +0000 (20:34 +0530)]
rsi: advertise 5GHz support based on device capability

Currently 5GHz gets advertised even for the device which supports
only 2.4Ghz band. This patch fixes the issue

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: enable power save by default for coex
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:16 +0000 (20:34 +0530)]
rsi: enable power save by default for coex

Power save is by default enabled for WLAN and BT coex mode.

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: parse TID from data frame correctly
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:15 +0000 (20:34 +0530)]
rsi: parse TID from data frame correctly

Currently TID is extracted by checking at specific offset in data frame.
This approach doesn't work for some of the frames. This patch uses mac80211
API and do it correctly

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: increase max supported aggregation subframes
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:14 +0000 (20:34 +0530)]
rsi: increase max supported aggregation subframes

Maximum number of supported aggregation subframes has been increased
to 8. This is the optimal number for the driver.

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: use appropriate interface for power save configuration
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:13 +0000 (20:34 +0530)]
rsi: use appropriate interface for power save configuration

Power save request should be sent on station interface. Virtual
interface which is connected should be preferred. This patch
resolves device not entering power save problem in certain
situations

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: device bootup parameter configuration
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:12 +0000 (20:34 +0530)]
rsi: device bootup parameter configuration

Some device configuration flags need to be enabled while sending 'bootup
params' internal frame to firmware. This patch takes care of it.

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: disable fw watchdog timer during reset
Amitkumar Karwar [Tue, 10 Apr 2018 15:04:11 +0000 (20:34 +0530)]
rsi: disable fw watchdog timer during reset

Firmware's watchdog timer should be disabled as a part of reset
sequence. This change fixes a firmware hang issue observed during
stress tests.

Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Siva Rebbagondla <siva.rebbagondla@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: remove unecessary PTR_ALIGN()s
Dan Carpenter [Fri, 6 Apr 2018 08:37:17 +0000 (11:37 +0300)]
rsi: remove unecessary PTR_ALIGN()s

The issue here is that we allocate "data" and then set
"data = PTR_ALIGN(data, 8);" and then we free the aligned pointer
instead of the original pointer.

kmalloc() pointers are already ARCH_SLAB_MINALIGN aligned which is 8 or
more on everything except certain Xtensa variants.  We decided that if
the Xtensa people ever notice a bug here then we'll tell them the bug is
on their side.  ;)

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: roaming enhancements
Prameela Rani Garnepudi [Thu, 29 Mar 2018 14:14:52 +0000 (19:44 +0530)]
rsi: roaming enhancements

To support roaming below changes are done:
* Station notify frame is send to firmware after sending assoc
  request. This will avoid dropping of first EAPOL frame due to
  delay in creation of station control block in firmware.
* Data queues are unblocked after sending station notify in open
  mode, after configuring key in WEP mode, and after receiving
  EAPOL4 confirm in WPA mode.
* Initial EAPOL frames priority is chaged to MGMT, rekey EAPOL
  frames priority changed to VO.

Signed-off-by: Prameela Rani Garnepudi <prameela.j04cs@gmail.com>
Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: enable 80MHz clock by default
Prameela Rani Garnepudi [Thu, 29 Mar 2018 14:14:51 +0000 (19:44 +0530)]
rsi: enable 80MHz clock by default

80MHz clock for device should be enabled by default in
TX command frame radio capabilities.

Signed-off-by: Prameela Rani Garnepudi <prameela.j04cs@gmail.com>
Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: move descriptor preparation to core
Prameela Rani Garnepudi [Thu, 29 Mar 2018 14:14:50 +0000 (19:44 +0530)]
rsi: move descriptor preparation to core

Descriptors preparation is moved to core instead of HAL to
avoid synchronization issues in sending TX frames

Signed-off-by: Prameela Rani Garnepudi <prameela.j04cs@gmail.com>
Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agorsi: move xtend_desc structure from rsi_main.h to rsi_mgmt.h
Prameela Rani Garnepudi [Thu, 29 Mar 2018 14:14:49 +0000 (19:44 +0530)]
rsi: move xtend_desc structure from rsi_main.h to rsi_mgmt.h

All descriptor structures are in rsi_mgmt.h except this
extended descriptor structure. Hence moving it to rsi_mgmt.h
and also renaming to rsi_xtend_desc.

Signed-off-by: Prameela Rani Garnepudi <prameela.j04cs@gmail.com>
Signed-off-by: Amitkumar Karwar <amit.karwar@redpinesignals.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agocw1200: fix spelling mistake: "Mailformed" -> "Malformed"
Colin Ian King [Thu, 29 Mar 2018 15:38:18 +0000 (16:38 +0100)]
cw1200: fix spelling mistake: "Mailformed" -> "Malformed"

Trivial fix to spelling mistake in wiphy_warn warning message text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoqtnfmac: add DFS offload support
Dmitry Lebed [Mon, 26 Mar 2018 13:36:33 +0000 (16:36 +0300)]
qtnfmac: add DFS offload support

DFS offload support implemented:
- DFS_OFFLOAD feature is advertised depending on HW capabilities
- CAC_STARTED event forwarding from HW implemented
- start_radar_detection() callback now returning -ENOTSUPP
  if DFS_OFFLOAD is enabled

Signed-off-by: Dmitry Lebed <dlebed@quantenna.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agortlwifi: cleanup 8723be ant_sel definition
Ping-Ke Shih [Fri, 20 Apr 2018 02:30:09 +0000 (10:30 +0800)]
rtlwifi: cleanup 8723be ant_sel definition

Some HP laptops have only a single wifi antenna. This would not be a
problem except that they were shipped with an incorrectly encoded
EFUSE. It should have been possible to open the computer and transfer
the antenna connection to the other terminal except that such action
might void the warranty, and moving the antenna broke the Windows
driver. The fix was to add a module option that would override the
EFUSE encoding. That was done with commit c18d8f509571 ("rtlwifi:
rtl8723be: Add antenna select module parameter"). There was still a
problem with Bluetooth coexistence, which was addressed with commit
baa170229095 ("rtlwifi: btcoexist: Implement antenna selection").
There were still problems, thus there were commit 0ff78adeef11
("rtlwifi: rtl8723be: fix ant_sel code") and commit 6d6226928369
("rtlwifi: btcoexist: Fix antenna selection code"). Despite all these
attempts at fixing the problem, the code is not yet right. A proper
fix is important as there are now instances of laptops having
RTL8723DE chips with the same problem.

The module parameter ant_sel is used to control antenna number and path.
At present enum ANT_{X2,X1} is used to define the antenna number, but
this choice is not intuitive, thus change to a new enum ANT_{MAIN,AUX}
to make it more readable. This change showed examples where incorrect
values were used. It was also possible to remove a workaround in
halbtcoutsrc.c.

The experimental results with single antenna connected to specific path
are now as follows:
  ant_sel  ANT_MAIN(#1)  ANT_AUX(#2)
     0        -8            -62
     1        -62           -10
     2        -6            -60

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Fixes: c18d8f509571 ("rtlwifi: rtl8723be: Add antenna select module parameter")
Fixes: baa170229095 ("rtlwifi: btcoexist: Implement antenna selection")
Fixes: 0ff78adeef11 ("rtlwifi: rtl8723be: fix ant_sel code")
Fixes: 6d6226928369 ("rtlwifi: btcoexist: Fix antenna selection code")
Cc: Stable <stable@vger.kernel.org> # 4.7+
Reviewed-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoiwlwifi: mvm: fix old scan version sizes
Luca Coelho [Mon, 23 Apr 2018 13:01:31 +0000 (16:01 +0300)]
iwlwifi: mvm: fix old scan version sizes

When version 8 of the scan command API was introduced, only the size
of version 7 was updated, causing older versions of the firmware to
throw BAD_COMMAND errors.

Calculating the old version based on the size of the latest version
got too complicated and the size of the older versions will never
change anyway, so it's better to just hardcoded the sizes.

Fixes: 66fa2424df16 ("iwlwifi: fw api: support the new scan request FW API version")
Reported-by: Scott Register <sreg@sreg.io>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoMerge tag 'iwlwifi-next-for-kalle-2018-04-20' of git://git.kernel.org/pub/scm/linux...
Kalle Valo [Tue, 24 Apr 2018 10:00:03 +0000 (13:00 +0300)]
Merge tag 'iwlwifi-next-for-kalle-2018-04-20' of git://git./linux/kernel/git/iwlwifi/iwlwifi-next

iwlwifi patches for 4.18

* implement Traffic Condition Monitor;
* use TCM for scan and BT coex;
* use TCM to detect when the AP doesn't support UAPSD properly;
* some more work for the 22000 family of devices;
* introduce AMSDU rate control offload;
* a couple of clean-ups and bugfixes.

6 years agowcn36xx: pass information elements in scan requests
Daniel Mack [Tue, 17 Apr 2018 13:23:37 +0000 (15:23 +0200)]
wcn36xx: pass information elements in scan requests

When the wifi driver core passes IE elements in the scan request, append
them to the firmware message. The driver currently tells the core that
it is capable of attaching up to WCN36XX_MAX_SCAN_IE_LEN octets, but
doesn't actually pass them to the the hardware.

Note that this patch doesn't fix a bug that was observed. The change is
merely done for the sake of completeness as the hardware supports
appending IEs in scans. Tests show that network scans work fine with
this patch applied.

Some defines were moved around to avoid cyclic include dependencies.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agowcn36xx: send bss_type in scan requests
Daniel Mack [Tue, 17 Apr 2018 13:23:36 +0000 (15:23 +0200)]
wcn36xx: send bss_type in scan requests

Pass the bss_type of the currently configured BSS in the message for the
scan request. Therefore, that setting needs to be kept in struct
wcn36xx_vif.

This seems to be only interesting when scanning for a specific SSID
and doesn't matter for regular wildcard scans.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agowcn36xx: handle scan cancellation when firmware support is missing
Daniel Mack [Tue, 17 Apr 2018 13:23:35 +0000 (15:23 +0200)]
wcn36xx: handle scan cancellation when firmware support is missing

For firmwares that don't have the SCAN_OFFLOAD feature bit set, do
not call into wcn36xx_smd_stop_hw_scan(). Instead, stop the asynchronous
work and call into ieee80211_scan_completed() immediately.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agowcn36xx: cancel pending scan request when interface goes down
Daniel Mack [Tue, 17 Apr 2018 13:23:34 +0000 (15:23 +0200)]
wcn36xx: cancel pending scan request when interface goes down

When the network interface goes down while a scan request is still
pending that can't be stopped due to firmware hickups, wcn->scan_req
remains set, even though the hardware is deinitialized. This results
in -EBUSY for all scan requests after the interface was brought up
again.

Fix this by explicitly completing pending scan requests in
wcn36xx_stop().

Signed-off-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agowcn36xx: abort scan request when 'dequeued' indicator is sent
Daniel Mack [Tue, 17 Apr 2018 13:23:33 +0000 (15:23 +0200)]
wcn36xx: abort scan request when 'dequeued' indicator is sent

When the firmware sends a WCN36XX_HAL_SCAN_IND_DEQUEUED indication,
the request is apparently no longer valid. Attempts to stop the hardware
scan request subsequently will lead to the following error message, and
the hardware is no longer able to communicate with any AP:

[   57.917186] wcn36xx: ERROR hal_stop_scan_offload response failed err=5

Interpreting this indicator message as scan abortion fixes this.

While at it, add a newline to a debug print.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: enable sta idle power save
Govind Singh [Tue, 17 Apr 2018 12:07:01 +0000 (17:37 +0530)]
ath10k: enable sta idle power save

Enable sta power save in fw for the targets that
supports idle power save. The idle ps enable command
will be ignored by the firmware which does not support
this feature.

Signed-off-by: Govind Singh <govinds@codeaurora.org>
Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: enable SRRI/DRRI support on ddr for WCN3990
Govind Singh [Tue, 17 Apr 2018 12:07:00 +0000 (17:37 +0530)]
ath10k: enable SRRI/DRRI support on ddr for WCN3990

SRRI/DRRI are not mapped in the HW Shadow block and can lead
to un-clocked access if common subsystem in the target is
powered down due to idle mode.

To mitigate this problem SRRI/DRRI can be read from
DDR instead of doing an actual hardware read.
Host allocates non cached memory on ddr and configures
the physical address of this memory to the CE hardware.
The hardware updates the RRI on this particular location.
Read SRRI/DRRI from DDR location instead of
direct target read.

Enable retention restore on ddr using hw params to enable
in specific targets.

Signed-off-by: Govind Singh <govinds@codeaurora.org>
Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: add support for shadow register for WNC3990
Rakesh Pillai [Tue, 17 Apr 2018 12:06:59 +0000 (17:36 +0530)]
ath10k: add support for shadow register for WNC3990

WCN3990 needs shadow register write operation support
for copy engine for regular operation in powersave mode.
Add support for copy engine shadow register write in
datapath tx for WCN3990

Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: add hw params for shadow register support
Rakesh Pillai [Tue, 17 Apr 2018 12:06:58 +0000 (17:36 +0530)]
ath10k: add hw params for shadow register support

wcn3990 supports shadow register for ce write.

Add a hw param for shadow register support.

Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: enable hw checksum for wcn3990
Rakesh Pillai [Tue, 17 Apr 2018 09:24:26 +0000 (14:54 +0530)]
ath10k: enable hw checksum for wcn3990

By default ath10k driver enables the support for HW_CHECKSUM
(NETIF_F_HW_CSUM). Since the TCP/UDP checksum calculation is not enabled
in the wcn3990 firmware the checksum is incorrect in the TCP/UDP packets
and all patckets are dropped. But due note that wcn3990 support in
ath10k is still incomplete so this isn't a critical fix (yet).

Enable hw checksum calculations in wcn3990 hardware by
setting the proper flags in msdu descriptor tso flags.

Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agowcn36xx: pass correct BSS index when deleting BSS keys
Daniel Mack [Thu, 19 Apr 2018 16:40:11 +0000 (19:40 +0300)]
wcn36xx: pass correct BSS index when deleting BSS keys

The firmware message to delete BSS keys expects a BSS index to be passed.
This field is currently hard-coded to 0. Fix this by passing in the index
we received from the firmware when the BSS was configured.

The encryption type in that message also needs to be set to what was used
when the key was set, so the assignment of vif_priv->encrypt_type is now
done after the firmware command was sent. This reportedly fixes the
following error in AP mode:

  wcn36xx: ERROR hal_remove_bsskey response failed err=6

Also, AFAIU, when a BSS is deleted, the firmware apparently drops all the
keys associated with it. Trying to remove the key explicitly afterwards
will hence lead to the following message:

  wcn36xx: ERROR hal_remove_bsskey response failed err=16

This is now suppressed with an extra check for the BSS index validity.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: convert wow pattern from 802.3 to 802.11
Wen Gong [Thu, 19 Apr 2018 16:40:07 +0000 (19:40 +0300)]
ath10k: convert wow pattern from 802.3 to 802.11

When trying to set wow wakeup patterns it fails with this command:

iw phyxx wowlan enable patterns offset xx+ IP address xx.xx.xx.xx

The reason is that the wow pattern from upper layer is in 802.3 format
for this case, it need to convert it to 802.11 format. The input
offset parameter is used for 802.3, but the actual offset firmware
need depends on rx_decap_mode, so that it needs to be recalculated.
Pattern of 802.3 packet is not same with 802.11 packet. If the
rx_decap_mode is ATH10K_HW_TXRX_NATIVE_WIFI, then firmware will
receive data packet with 802.11 format from hardware.

Tested with QCA6174 hw3.0 with firmware
WLAN.RM.4.4.1-00099-QCARMSWPZ-1, but this will also affect QCA9377.
This has always failed, so it's not a regression with new firmware
releases.

Signed-off-by: Wen Gong <wgong@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: support MAC address randomization in scan
Carl Huang [Thu, 19 Apr 2018 16:39:40 +0000 (19:39 +0300)]
ath10k: support MAC address randomization in scan

The ath10k reports the random_mac_addr capability to upper layer
based on the service bit firmware reported. Driver sets the
spoofed flag in scan_ctrl_flag to firmware if upper layer has
enabled this feature in scan request.

Test with QCA6174 hw3.0 and firmware-6.bin_WLAN.RM.4.4.1-00102-QCARMSWP-1,
but QCA9377 is also affected.

Signed-off-by: Carl Huang <cjhuang@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agoath10k: add WMI_SERVICE_AVAILABLE_EVENT support
Carl Huang [Thu, 19 Apr 2018 16:39:38 +0000 (19:39 +0300)]
ath10k: add WMI_SERVICE_AVAILABLE_EVENT support

Add WMI_SERVICE_AVAILABLE_EVENT to extend WMI_SERVICE_READY_EVENT,
the 128bit service map in WMI_SERVICE_READY_EVENT is not enough
for firmware to notice new WLAN service to host driver. Hereby,
for thoese new WLAN service, firmware will notice host driver by
WMI_SERVICE_AVAILABLE_EVENT.

Signed-off-by: Alan Liu <alanliu@codeaurora.org>
Signed-off-by: Carl Huang <cjhuang@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
6 years agonet: fib_rules: fix l3mdev netlink attr processing
Roopa Prabhu [Tue, 24 Apr 2018 03:08:41 +0000 (20:08 -0700)]
net: fib_rules: fix l3mdev netlink attr processing

Fixes: b16fb418b1bf ("net: fib_rules: add extack support")
Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoselftests: net: update .gitignore with missing test
Anders Roxell [Mon, 23 Apr 2018 14:00:50 +0000 (16:00 +0200)]
selftests: net: update .gitignore with missing test

Fixes: 192dc405f308 ("selftests: net: add tcp_mmap program")
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodca: make function dca_common_get_tag static
Colin Ian King [Mon, 23 Apr 2018 12:49:38 +0000 (13:49 +0100)]
dca: make function dca_common_get_tag static

Function dca_common_get_tag is local to the source and does not need to be
in global scope, so make it static.

Cleans up sparse warning:
drivers/dca/dca-core.c:273:4: warning: symbol 'dca_common_get_tag' was
not declared. Should it be static?

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'ipv6-couple-of-fixes-for-rcu-change-to-from'
David S. Miller [Mon, 23 Apr 2018 20:12:55 +0000 (16:12 -0400)]
Merge branch 'ipv6-couple-of-fixes-for-rcu-change-to-from'

David Ahern says:

====================
net/ipv6: couple of fixes for rcu change to from

So many details... I am thankful for all the robots running the
permutations and tools.

Two bug fixes from the rcu change to rt->from:
1. missing rcu lock in ip6_negative_advice
2. rcu dereferences in 2 sites
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet/ipv6: Fix missing rcu dereferences on from
David Ahern [Mon, 23 Apr 2018 18:32:07 +0000 (11:32 -0700)]
net/ipv6: Fix missing rcu dereferences on from

kbuild test robot reported 2 uses of rt->from not properly accessed
using rcu_dereference:
1. add rcu_dereference_protected to rt6_remove_exception_rt and make
   sure it is always called with rcu lock held.

2. change rt6_do_redirect to take a reference on 'from' when accessed
   the first time so it can be used the sceond time outside of the lock

Fixes: a68886a69180 ("net/ipv6: Make from in rt6_info rcu protected")
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet/ipv6: add rcu locking to ip6_negative_advice
David Ahern [Mon, 23 Apr 2018 18:32:06 +0000 (11:32 -0700)]
net/ipv6: add rcu locking to ip6_negative_advice

syzbot reported a suspicious rcu_dereference_check:
  __dump_stack lib/dump_stack.c:77 [inline]
  dump_stack+0x1b9/0x294 lib/dump_stack.c:113
  lockdep_rcu_suspicious+0x14a/0x153 kernel/locking/lockdep.c:4592
  rt6_check_expired+0x38b/0x3e0 net/ipv6/route.c:410
  ip6_negative_advice+0x67/0xc0 net/ipv6/route.c:2204
  dst_negative_advice include/net/sock.h:1786 [inline]
  sock_setsockopt+0x138f/0x1fe0 net/core/sock.c:1051
  __sys_setsockopt+0x2df/0x390 net/socket.c:1899
  SYSC_setsockopt net/socket.c:1914 [inline]
  SyS_setsockopt+0x34/0x50 net/socket.c:1911

Add rcu locking around call to rt6_check_expired in
ip6_negative_advice.

Fixes: a68886a69180 ("net/ipv6: Make from in rt6_info rcu protected")
Reported-by: syzbot+2422c9e35796659d2273@syzkaller.appspotmail.com
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'qed-debug-data'
David S. Miller [Mon, 23 Apr 2018 16:05:58 +0000 (12:05 -0400)]
Merge branch 'qed-debug-data'

Denis Bolotin says:

====================
Add configuration information to register dump and debug data

The purpose of this patchset is to add configuration information to the
debug data collection, which already contains register dump.
The first patch (removing the ptt) is essential because it prevents the
unnecessary ptt acquirement when calling mcp APIs.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Add configuration information to register dump and debug data
Denis Bolotin [Mon, 23 Apr 2018 11:56:05 +0000 (14:56 +0300)]
qed: Add configuration information to register dump and debug data

Configuration information is added to the debug data collection, in
addition to register dump.
Added qed_dbg_nvm_image() that receives an image type, allocates a
buffer and reads the image. The images are saved in the buffers and the
dump size is updated.

Signed-off-by: Denis Bolotin <denis.bolotin@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqed: Delete unused parameter p_ptt from mcp APIs
Denis Bolotin [Mon, 23 Apr 2018 11:56:04 +0000 (14:56 +0300)]
qed: Delete unused parameter p_ptt from mcp APIs

Since nvm images attributes are cached during driver load, acquiring ptt
is not needed when calling qed_mcp_get_nvm_image().

Signed-off-by: Denis Bolotin <denis.bolotin@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: stmmac: Implement logic to automatically select HW Interface
Jose Abreu [Mon, 23 Apr 2018 08:05:15 +0000 (09:05 +0100)]
net: stmmac: Implement logic to automatically select HW Interface

Move all the core version detection to a common place ("hwif.c") and
implement a table which can be used to lookup the correct callbacks for
each IP version.

This simplifies the initialization flow of each IP version and eases
future implementation of new IP versions.

Signed-off-by: Jose Abreu <joabreu@synopsys.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Joao Pinto <jpinto@synopsys.com>
Cc: Vitor Soares <soares@synopsys.com>
Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agor8169: don't use netif_info et al before net_device has been registered
Heiner Kallweit [Sun, 22 Apr 2018 15:15:15 +0000 (17:15 +0200)]
r8169: don't use netif_info et al before net_device has been registered

There's no benefit in using netif_info et al before the net_device has
been registered. We get messages like
r8169 0000:03:00.0 (unnamed net_device) (uninitialized): [message]
Therefore use dev_info/dev_err instead.

As a side effect we don't need parameter dev for function
rtl8169_get_mac_version() any longer.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: init sk_cookie for inet socket
Yafang Shao [Sun, 22 Apr 2018 13:50:04 +0000 (21:50 +0800)]
net: init sk_cookie for inet socket

With sk_cookie we can identify a socket, that is very helpful for
traceing and statistic, i.e. tcp tracepiont and ebpf.
So we'd better init it by default for inet socket.
When using it, we just need call atomic64_read(&sk->sk_cookie).

Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'fib-rules-extack-support'
David S. Miller [Mon, 23 Apr 2018 14:21:25 +0000 (10:21 -0400)]
Merge branch 'fib-rules-extack-support'

Roopa Prabhu says:

====================
fib rules extack support

First patch refactors code to move fib rule netlink handling
into a common function. This became obvious when adding
duplicate extack msgs in add and del paths. Second patch
adds extack msgs.

v2 - Dropped the ip route get support and selftests from
     the series to look at the input path some more (as pointed
     out by ido). Will come back to that next week when i have
     some time. resending just the extack part for now.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: fib_rules: add extack support
Roopa Prabhu [Sat, 21 Apr 2018 16:41:31 +0000 (09:41 -0700)]
net: fib_rules: add extack support

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agofib_rules: move common handling of newrule delrule msgs into fib_nl2rule
Roopa Prabhu [Sat, 21 Apr 2018 16:41:30 +0000 (09:41 -0700)]
fib_rules: move common handling of newrule delrule msgs into fib_nl2rule

This reduces code duplication in the fib rule add and del paths.
Get rid of validate_rulemsg. This became obvious when adding duplicate
extack support in fib newrule/delrule error paths.

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agotc-testing: updated ife test cases
Roman Mashak [Sat, 21 Apr 2018 03:56:51 +0000 (23:56 -0400)]
tc-testing: updated ife test cases

Signed-off-by: Roman Mashak <mrv@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>