openwrt/staging/blogic.git
10 years agoath10k: make target endianess more explicit
Michal Kazior [Tue, 26 Aug 2014 16:14:03 +0000 (19:14 +0300)]
ath10k: make target endianess more explicit

Some copy engine structures are target specific
and are uploaded to the device during
init/configuration.

This also cleans up a bit diag_mem_read/write
implicit byteswap mess leaving only
diag_access_read/write with an implicit endianess
byteswap.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: dont duplicate service-pipe mapping
Michal Kazior [Tue, 26 Aug 2014 16:14:03 +0000 (19:14 +0300)]
ath10k: dont duplicate service-pipe mapping

The mapping is already defined in a structure. It
makes little sense to duplicate information stored
in it within a function.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: move pci init structures
Michal Kazior [Tue, 26 Aug 2014 16:14:02 +0000 (19:14 +0300)]
ath10k: move pci init structures

It doesn't make much sense to have copy engine
configuration structures spread across the whole
source file.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix fw crash dumping
Michal Kazior [Mon, 25 Aug 2014 10:13:14 +0000 (12:13 +0200)]
ath10k: fix fw crash dumping

Recent crash dump patches introduced a regression.
If debugfs was disabled upon crash user could only
see the following:

 [  793.880000] ath10k: firmware crashed! (uuid n/a)
 [  793.890000] ath10k: qca988x hw2.0 (0x4100016c, 0x043202ff) fw 10.1.467.2-1 api 2 htt 2.1
 [  793.890000] ath10k: debug 0 debugfs 0 tracing 0 dfs 1

The report was missing register dump. Fix it by
printing registers regardless if crash_data is
present or not.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: improve logging to include dev id
Michal Kazior [Mon, 25 Aug 2014 10:09:38 +0000 (12:09 +0200)]
ath10k: improve logging to include dev id

This makes it a lot easier to log and debug
messages if there's more than 1 ath10k device on a
system.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: flush hif buffers before recovery
Michal Kazior [Fri, 22 Aug 2014 12:33:18 +0000 (14:33 +0200)]
ath10k: flush hif buffers before recovery

Transport buffers weren't flushed and processed
before queueing hw recovery request to mac80211.

This could in theory result in an unwanted htt/wmi
rx events being processed while mac80211 recovers
the device and possibly interfere or even crash
the system.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: remove ar_pci->started
Michal Kazior [Fri, 22 Aug 2014 12:33:17 +0000 (14:33 +0200)]
ath10k: remove ar_pci->started

There are basically no more uses for
ar_pci->started. It is also perfectly safe to call
hif_stop without hif_start now.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: ignore ar_pci->started in pipe cleanup
Michal Kazior [Fri, 22 Aug 2014 12:33:16 +0000 (14:33 +0200)]
ath10k: ignore ar_pci->started in pipe cleanup

Structures used by these functions are now
guaranteed to remain accessible until driver is
unregistered.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: update comment regarding warm reset
Michal Kazior [Fri, 22 Aug 2014 12:33:15 +0000 (14:33 +0200)]
ath10k: update comment regarding warm reset

The old comment was a little out of date. HTT Rx
ring is a more relevant problem when stopping
transport layer.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: rework posting pci rx buffers
Michal Kazior [Fri, 22 Aug 2014 12:33:14 +0000 (14:33 +0200)]
ath10k: rework posting pci rx buffers

It was possible on a host system running low on
memory to end up with no rx buffers on pci pipes.

This makes the driver more robust as it won't fail
to start if it can't allocate all rx buffers right
away. If it is fatal then upper layers will notice
trouble anyway.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: remove early irq handling
Michal Kazior [Fri, 22 Aug 2014 12:23:34 +0000 (14:23 +0200)]
ath10k: remove early irq handling

It's not really necessary to have a dedicated irq
handler just for the sake of catching early fw
crashes anymore. It is now safe to use one handler
even during early stages of device boot up.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: make sure to really disable irqs
Michal Kazior [Fri, 22 Aug 2014 12:23:33 +0000 (14:23 +0200)]
ath10k: make sure to really disable irqs

This fixes two corner cases.

One is a race between disabling copy engine
interrupts and unhandled pending interrupts on the
host. This could end up with a runaway tasklet and
consequently memory leak of a few copy engine
rx buffers.

The other one is an unexpected (and non-maskable
via device CSR) MSI fw indication interrupt during
teardown. This could trigger the same problem as
the first corner case.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: split ce irq/handler setup
Michal Kazior [Fri, 22 Aug 2014 12:23:32 +0000 (14:23 +0200)]
ath10k: split ce irq/handler setup

It doesn't make much sense to overwrite send_cb
and recv_cb callbacks over and over again whenever
transport starts. Just make sure to unmask copy
engine interrupts when starting.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: setup irq method in probe
Michal Kazior [Fri, 22 Aug 2014 12:23:31 +0000 (14:23 +0200)]
ath10k: setup irq method in probe

It doesn't make sense to re-init irqs completely
whenever transport is started/stopped. Do it just
once upon probing/removing.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix legacy irq workaround
Michal Kazior [Fri, 22 Aug 2014 12:23:30 +0000 (14:23 +0200)]
ath10k: fix legacy irq workaround

Wrong register was being set up. This could
prevent firmware from booting in some rare cases
when using legacy interrupts.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: move fw init print
Michal Kazior [Fri, 22 Aug 2014 12:23:29 +0000 (14:23 +0200)]
ath10k: move fw init print

Firmware probing is done only once when driver is
registered and firmware version is guaranteed to
remain the same until driver is unregistered.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: print more driver info when firmware crashes
Kalle Valo [Mon, 25 Aug 2014 05:37:45 +0000 (08:37 +0300)]
ath10k: print more driver info when firmware crashes

Sometimes users forget to include important info like firmware version,
so better to print all the info.

Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: rename ath10k_pci_hif_dump_area() to ath10k_pci_fw_crashed_dump()
Kalle Valo [Mon, 25 Aug 2014 05:37:37 +0000 (08:37 +0300)]
ath10k: rename ath10k_pci_hif_dump_area() to ath10k_pci_fw_crashed_dump()

Better to have a clear name for the function. While at it, clear up the title
for the register dump.

Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: provide firmware crash info via debugfs
Ben Greear [Mon, 25 Aug 2014 05:37:32 +0000 (08:37 +0300)]
ath10k: provide firmware crash info via debugfs

Store the firmware registers and other relevant data to a firmware crash dump
file and provide it to user-space via debugfs. Should help with figuring out
why the firmware crashed.

kvalo: remove dbglog support, rework and refactor the code to avoid ifdefs and
otherwise simplify it as well

Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: add ath10k_pci_diag_* helpers
Kalle Valo [Mon, 25 Aug 2014 05:37:26 +0000 (08:37 +0300)]
ath10k: add ath10k_pci_diag_* helpers

ath10k_pci_diag_read32() is for reading u32 from a device and ath10k_pci_diag_read_hi()
is a helper for reading data using "host interest" table.

Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix a conflict bug in wmi service bitmap
Michal Kazior [Tue, 19 Aug 2014 11:22:32 +0000 (13:22 +0200)]
ath10k: fix a conflict bug in wmi service bitmap

Service mapping for main firmware branch was incorrectly used for 10.x firmware
and vice-versa.  This caused wmi_services in debugfs to print wrong values.

This fixes commit cff990ce7ddd6 ("ath10k: fix wmi service bitmap debug") where
for some reason there was either a conflict that wasn't resolved properly or
git had a bad day.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath6kl: fix %d confusingly prefixed with 0x in format strings
Hans Wennborg [Sun, 17 Aug 2014 21:57:11 +0000 (14:57 -0700)]
ath6kl: fix %d confusingly prefixed with 0x in format strings

Signed-off-by: Hans Wennborg <hans@hanshq.net>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: unregister spectral before mac
Simon Wunderlich [Tue, 12 Aug 2014 15:12:17 +0000 (17:12 +0200)]
ath10k: unregister spectral before mac

If spectral is unregistered after mac80211, the relayfs file has already
been removed recursively by mac/cfg80211, and spectral tries to remove
the file once more, thus leading to double free problems. Better clean
up spectral before to avoid that problem.

Reported-by: Kalle Valo <kvalo@qca.qualcomm.com>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath6kl: Add SDIO device ID for QCA6234X Support
Srinivas Kandagatla [Tue, 12 Aug 2014 10:45:59 +0000 (11:45 +0100)]
ath6kl: Add SDIO device ID for QCA6234X Support

This patch adds device ID 402 to support QCA6234X found in APQ8064 SOC
in IFC6410 board.

Tested with mainline mmci sdio driver.

Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix typo in error message
Ben Greear [Tue, 12 Aug 2014 08:02:20 +0000 (11:02 +0300)]
ath10k: fix typo in error message

tranmist -> transmit

Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: improve vdev map handling
Ben Greear [Tue, 12 Aug 2014 08:02:19 +0000 (11:02 +0300)]
ath10k: improve vdev map handling

Check vdev map has space before calling ffs,
fix invalid cleanup in failure to create vdev
case.

Open-code the BIT() logic since BIT does not properly
handle 64-bit bitfields and future patches will make
use of larger bitfields.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: remove htc->stopped
Michal Kazior [Thu, 7 Aug 2014 09:03:31 +0000 (11:03 +0200)]
ath10k: remove htc->stopped

This is not necessary anymore. There are no more
uncontrolled htc tx entry points.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: group some pci probing helpers
Michal Kazior [Thu, 7 Aug 2014 09:03:30 +0000 (11:03 +0200)]
ath10k: group some pci probing helpers

Make probe/remove functions shorter and easier to
understand.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: remove pci features var
Michal Kazior [Thu, 7 Aug 2014 09:03:29 +0000 (11:03 +0200)]
ath10k: remove pci features var

The ATH10K_PCI_FEATURE_MSI_X was originally
introduced to support both chips QCA988Xv1 and
QCA988Xv2. Since v1 isn't supported anymore it
doesn't make sense to keep the feature flag
around. Since this is the last one remove the
whole thing.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: remove target soc ps code
Michal Kazior [Thu, 7 Aug 2014 09:03:28 +0000 (11:03 +0200)]
ath10k: remove target soc ps code

The soc powersave was disabled by default. It
never was fully tested. Some hw apparently had
problems with it and the implementation itself had
a possible race.

Just remove the refcounting and simply wake up the
device when probing and put to sleep when
removing.

kvalo: make ath10k_pci_wake() and _sleep() static

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: embed ar_pci inside ar
Michal Kazior [Thu, 7 Aug 2014 09:03:27 +0000 (11:03 +0200)]
ath10k: embed ar_pci inside ar

Use the common convention of embedding private
structures inside parent structures. This
reduces allocations and simplifies pci probing
code.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: introduce a stricter scan state machine
Michal Kazior [Tue, 5 Aug 2014 12:54:44 +0000 (14:54 +0200)]
ath10k: introduce a stricter scan state machine

This aims at fixing some rare scan bugs related to
firmware reporting unexpected scan event
sequences.

One such bug was if spectral scan phyerr reporting
prevented firmware from properly propagating scan
events to host. This led to scan timeout. After
that next scan would trigger scan completed event
first (before scan started event) leading to
ar->scan.in_progress and timeout timer states to
be overwritten incorrectly and making the very
next scan to hang forever.

Reported-by: Janusz Dziedzic <janusz.dziedzic@tieto.com>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: simplify scan debug prints
Michal Kazior [Tue, 5 Aug 2014 12:54:43 +0000 (14:54 +0200)]
ath10k: simplify scan debug prints

This also reduces the cruft of printing scan event
names in capitals.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix wmi service bitmap debug
Michal Kazior [Mon, 4 Aug 2014 06:18:33 +0000 (09:18 +0300)]
ath10k: fix wmi service bitmap debug

The 10.x and main firmware branches have
conflicting WMI service bitmap definitions.

This also fixes WMI services parsing on big-endian
hosts and changes debugfs output to be more human
friendly.

kvalo: remove braces and the last semicolon from SVCSTR()

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath6kl: convert a driver to use module_usb_driver()
Himangi Saraogi [Fri, 1 Aug 2014 16:11:15 +0000 (21:41 +0530)]
ath6kl: convert a driver to use module_usb_driver()

This converts a driver in drivers/net/* to use the
module_usb_driver() macro which makes the code smaller and a
bit simpler.

Signed-off-by: Himangi Saraogi <himangi774@gmail.com>
Acked-by: Julia Lawall <julia.lawall@lip6.fr>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: add spectral scan feature
Simon Wunderlich [Sat, 2 Aug 2014 06:12:54 +0000 (09:12 +0300)]
ath10k: add spectral scan feature

Adds the spectral scan feature for ath10k. The spectral scan is triggered by
configuring a mode through a debugfs control file. Samples can be gathered via
another relay debugfs file.

Essentially, to try it out:

ip link set dev wlan0 up
echo background > /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan_ctl
echo trigger > /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan_ctl
iw dev wlan0 scan
echo disable > /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan_ctl
cat /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan0 > samples

This feature is still experimental. Based on the original RFC patch of
Sven Eckelmann.

Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
Signed-off-by: Mathias Kretschmer <mathias.kretschmer@fokus.fraunhofer.de>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath: Move spectral debugfs structs to shared header
Sven Eckelmann [Sat, 2 Aug 2014 06:12:53 +0000 (09:12 +0300)]
ath: Move spectral debugfs structs to shared header

The ath9k and ath10k will share the definitions of the debugfs spectral
structures and enums. Having them in the same place helps to avoid conflicts.

Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Simon Wunderlich <sw@simonwunderlich.de>
Signed-off-by: Mathias Kretschmer <mathias.kretschmer@fokus.fraunhofer.de>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: improve channel switching
Michal Kazior [Tue, 29 Jul 2014 09:53:36 +0000 (12:53 +0300)]
ath10k: improve channel switching

In some cases during heavy tx vdev stop-start
would timeout on vdev synchronization causing
traffic to stall for a few seconds.

Instead of stop-starting use a dedicated vdev
restart command and down vdevs explicitly before
doing so.

This gets rid of the synchronization
warnings/timeouts and makes channel switching
smoother during traffic.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: extend debug code for RX path
Janusz Dziedzic [Mon, 28 Jul 2014 20:59:43 +0000 (23:59 +0300)]
ath10k: extend debug code for RX path

Print sequence number, AMSDU_MORE flag and AC when additional
debug enabled in RX path. This is usefull for debugging purpose.

Signed-off-by: Janusz Dziedzic <janusz.dziedzic@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix aggregated 4addr Rx
Michal Kazior [Mon, 28 Jul 2014 20:59:42 +0000 (23:59 +0300)]
ath10k: fix aggregated 4addr Rx

A-MSDU 4addr frames weren't reconstructed properly
and in some cases this resulted in a warning:

 br0: received packet on wlan0.sta1 with own address as source address

Since this was only related to A-MSDU it would
trigger when more intense traffic was generated.

Reported-by: Vu Hai NGUYEN <vh.nguyen@actiasodielec.fr>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: add support for 10.2 firmware
Michal Kazior [Fri, 25 Jul 2014 11:32:17 +0000 (13:32 +0200)]
ath10k: add support for 10.2 firmware

The 10.2 firmware is a successor of 10.1 firmware
(formerly identified as 10.x). Both share a lot
but have some slight ABI differences that need to
be taken care of.

The 10.2 firmware introduces some new features but
those can be added in subsequent patches. This
patch makes ath10k boot and work with 10.2 with
comparable functionality to 10.1.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: improve 'hard' simulate fw crash
Ben Greear [Fri, 25 Jul 2014 08:56:40 +0000 (11:56 +0300)]
ath10k: improve 'hard' simulate fw crash

Different firmware may support different numbers of
vdevs.  Use value that is always out of range for all
firmware.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: handle attention flags correctly when using A-MSDU
Janusz Dziedzic [Fri, 25 Jul 2014 08:28:50 +0000 (11:28 +0300)]
ath10k: handle attention flags correctly when using A-MSDU

In case of A-MSDU RX we should check attention flags
correctly to be sure we report correct FCS status for
A-MSDU subframes. Without a patch we could report A-MSDU
subframes with wrong FCS as a correct to the stack, next
get a lot of DUP ACK TCP packets. Finally TP drop is seen
and this drop depends on FCS errors ratio for A-MSDU frame.

Example test case when TP drop is seen:
- ath10k configured as an AP
- used ath10k station
- forced A-MSDU (7 frames) on STA
- other traffic on channel (often FCS errors)
- monitor iface added on AP
- TCP STA -> AP traffic (iperf)

a) Iperf logs for case without the patch:

echo "1 64" > htt_max_amsdu_ampdu // disable A-MSDU
[ ID] Interval       Transfer     Bandwidth
[  3]  0.0- 5.0 sec  56.6 MBytes  95.0 Mbits/sec
[  3]  5.0-10.0 sec  60.4 MBytes   101 Mbits/sec
[  3] 10.0-15.0 sec  60.2 MBytes   101 Mbits/sec
[  3] 15.0-20.0 sec  60.2 MBytes   101 Mbits/sec
[  3] 20.0-25.0 sec  63.8 MBytes   107 Mbits/sec
[  3] 25.0-30.0 sec  64.9 MBytes   109 Mbits/sec

echo "7 64" > htt_max_amsdu_ampdu  // set 7 A-MSDU subframes
[  3] 30.0-35.0 sec  40.0 MBytes  67.1 Mbits/sec
[  3] 35.0-40.0 sec  35.9 MBytes  60.2 Mbits/sec
[  3] 40.0-45.0 sec  36.9 MBytes  61.9 Mbits/sec
[  3] 45.0-50.0 sec  37.9 MBytes  63.5 Mbits/sec
[  3] 50.0-55.0 sec  34.5 MBytes  57.9 Mbits/sec
[  3] 55.0-60.0 sec  25.4 MBytes  42.6 Mbits/sec
[  3] 60.0-65.0 sec  48.2 MBytes  81.0 Mbits/sec
[  3] 65.0-70.0 sec  28.8 MBytes  48.2 Mbits/sec
[  3] 70.0-75.0 sec  29.2 MBytes  49.1 Mbits/sec
[  3] 75.0-80.0 sec  22.9 MBytes  38.4 Mbits/sec
[  3] 80.0-85.0 sec  26.4 MBytes  44.2 Mbits/sec
[  3] 85.0-90.0 sec  31.5 MBytes  52.8 Mbits/sec

b) Iperf logs for case with patch:

echo "1 64" > htt_max_amsdu_ampdu // disable A-MSDU
[  3] local 192.168.12.2 port 57512 connected with 192.168.12.1 port 5001
[ ID] Interval       Transfer     Bandwidth
[  3]  0.0- 5.0 sec  60.8 MBytes   102 Mbits/sec
[  3]  5.0-10.0 sec  62.2 MBytes   104 Mbits/sec
[  3] 10.0-15.0 sec  60.9 MBytes   102 Mbits/sec

echo "7 64" > htt_max_amsdu_ampdu  // set 7 A-MSDU subframes
[  3] 15.0-20.0 sec  68.1 MBytes   114 Mbits/sec
[  3] 20.0-25.0 sec  80.5 MBytes   135 Mbits/sec
[  3] 25.0-30.0 sec  83.0 MBytes   139 Mbits/sec
[  3] 30.0-35.0 sec  79.1 MBytes   133 Mbits/sec
[  3] 35.0-40.0 sec  77.1 MBytes   129 Mbits/sec
[  3] 40.0-45.0 sec  77.4 MBytes   130 Mbits/sec

Reported-by: Denton Gentry <denton.gentry@gmail.com>
Signed-off-by: Janusz Dziedzic <janusz.dziedzic@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: don't advertise IBSS iftype for 10.x
Michal Kazior [Thu, 24 Jul 2014 17:07:00 +0000 (20:07 +0300)]
ath10k: don't advertise IBSS iftype for 10.x

The 10.x firmware does not support IBSS mode at
all. It can't beacon and it crashes when trying to
scan.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: fix Rx aggregation reordering
Michal Kazior [Wed, 23 Jul 2014 10:20:33 +0000 (12:20 +0200)]
ath10k: fix Rx aggregation reordering

Firmware doesn't perform Rx reordering so it is
left to the host driver to do that.

Use mac80211 to perform reordering instead of
re-inventing the wheel.

This fixes TCP throughput issues in some
environments.

Reported-by: Denton Gentry <denton.gentry@gmail.com>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoMerge remote-tracking branch 'wireless-next/master' into ath-next
Kalle Valo [Thu, 24 Jul 2014 17:06:37 +0000 (20:06 +0300)]
Merge remote-tracking branch 'wireless-next/master' into ath-next

10 years agob43: enable 5 GHz support for N-PHY devices
Rafał Miłecki [Wed, 23 Jul 2014 16:54:50 +0000 (18:54 +0200)]
b43: enable 5 GHz support for N-PHY devices

This has been tested on 14e4:4328 (BCM4321), 14e4:432b (BCM4322),
14e4:4353 (BCM43224) and 14e4:4359 (BCM43228) which is an almost
complete list of 5 GHz capable device (only BCM43222 is missing).

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: register limited amount of 5G channels for BCM43228
Rafał Miłecki [Wed, 23 Jul 2014 16:54:49 +0000 (18:54 +0200)]
b43: register limited amount of 5G channels for BCM43228

We don't have all needed channel tables due to RE process for this
device.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: don't calculate values for TSSI if we can't transmit
Rafał Miłecki [Wed, 23 Jul 2014 16:54:48 +0000 (18:54 +0200)]
b43: N-PHY: don't calculate values for TSSI if we can't transmit

This process requires sending some sample tone, so make sure we're
allowed to transmit first.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: report correct rate to mac80211 for 5 GHz packets
Rafał Miłecki [Wed, 23 Jul 2014 14:55:44 +0000 (16:55 +0200)]
b43: report correct rate to mac80211 for 5 GHz packets

So far we were assuming only A-PHY supports 5 GHz.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agomac80211_hwsim: fix compiler warning on MIPS
Andrew Bresticker [Tue, 22 Jul 2014 21:43:51 +0000 (14:43 -0700)]
mac80211_hwsim: fix compiler warning on MIPS

The dividend in do_div() is expected to be an unsigned 64-bit integer,
which leads to the following warning when building for 32-bit MIPS:

  drivers/net/wireless/mac80211_hwsim.c: In function 'mac80211_hwsim_set_tsf':
  drivers/net/wireless/mac80211_hwsim.c:664:98: warning: comparison of distinct pointer types lacks a cast [enabled by default]
    data->bcn_delta = do_div(delta, bcn_int);

Since we care about the signedness of delta when adjusting tsf_offset
and bcm_delta, use the absolute value for the division and compare
the two timestamps to determine the sign.

Signed-off-by: Andrew Bresticker <abrestic@chromium.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: fix rev7+ typos at random places
Rafał Miłecki [Tue, 22 Jul 2014 19:31:05 +0000 (21:31 +0200)]
b43: N-PHY: fix rev7+ typos at random places

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agoMerge branch 'for-john' of git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi...
John W. Linville [Wed, 23 Jul 2014 17:06:17 +0000 (13:06 -0400)]
Merge branch 'for-john' of git://git./linux/kernel/git/iwlwifi/iwlwifi-next

10 years agoiwlwifi: mvm: teardown TDLS peers during chan-switch and AP DCM
Arik Nemtsov [Sun, 15 Jun 2014 13:03:55 +0000 (16:03 +0300)]
iwlwifi: mvm: teardown TDLS peers during chan-switch and AP DCM

The DCM condition was not checked well for channel switch in both AP and
station scenarios. Teardown was also not done for AP/GO DCM. Add the
missing checks.

Reported-by: Peer, Ilan <ilan.peer@intel.com>
Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoMerge remote-tracking branch 'wireless-next/master' into iwlwifi-next
Emmanuel Grumbach [Tue, 22 Jul 2014 19:08:29 +0000 (22:08 +0300)]
Merge remote-tracking branch 'wireless-next/master' into iwlwifi-next

10 years agoath10k: prevent endless pci rx loop
Michal Kazior [Mon, 21 Jul 2014 18:03:10 +0000 (21:03 +0300)]
ath10k: prevent endless pci rx loop

It was possible to enter an endless loop while
processing a single pci copy engine pipe. This
could effectively render ath10k incapable of
responding to any requests.

An example case when this could happen is when
firmware generates a lot of events, e.g. spectral
scan phyerr via WMI.

Reported-by: Janusz Dziedzic <janusz.dziedzic@tieto.com>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agob43: extract one more radio parameter: version
Rafał Miłecki [Sun, 20 Jul 2014 10:57:45 +0000 (12:57 +0200)]
b43: extract one more radio parameter: version

Some radios may share the same ID and revision but differ by a version.
E.g. radio in BCM5357B0 is version 1 and requires specific handling.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: enable radio 0x2057 rev 14 support (AKA BCM43217)
Rafał Miłecki [Sat, 19 Jul 2014 10:52:47 +0000 (12:52 +0200)]
b43: enable radio 0x2057 rev 14 support (AKA BCM43217)

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: final fixes to rev7+ workarounds
Rafał Miłecki [Sat, 19 Jul 2014 10:52:46 +0000 (12:52 +0200)]
b43: N-PHY: final fixes to rev7+ workarounds

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: add rev7+ workarounds for radio revs 9 and 14
Rafał Miłecki [Sat, 19 Jul 2014 10:52:45 +0000 (12:52 +0200)]
b43: N-PHY: add rev7+ workarounds for radio revs 9 and 14

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: allow applying separated workarounds per core
Rafał Miłecki [Sat, 19 Jul 2014 10:52:44 +0000 (12:52 +0200)]
b43: N-PHY: allow applying separated workarounds per core

Newer devices need different workarounds for cores 0 and 1.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: update generic rev7+ workarounds
Rafał Miłecki [Sat, 19 Jul 2014 10:52:43 +0000 (12:52 +0200)]
b43: N-PHY: update generic rev7+ workarounds

Add some missing ops and prepare for new devices support. This patch is
a great stability improvement for BCM43217. Earlier Tenda W322E used to
disconnect every 2 minutes (16 times over 30 minutes). With this fix I
got it running for 4 hours (with iperf) without any disconnection.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: update digital filters setup
Rafał Miłecki [Sat, 19 Jul 2014 10:52:42 +0000 (12:52 +0200)]
b43: N-PHY: update digital filters setup

This fixes handling channel 14 and adds code for BCM43217.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agob43: N-PHY: add helper for setting digital filters
Rafał Miłecki [Sat, 19 Jul 2014 10:52:41 +0000 (12:52 +0200)]
b43: N-PHY: add helper for setting digital filters

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agomwifiex: card reset: enable rescan of non-removable card
Andreas Fenkart [Mon, 21 Jul 2014 08:01:52 +0000 (10:01 +0200)]
mwifiex: card reset: enable rescan of non-removable card

mmc_rescan will scan for non-removable cards only once, hence the card
will not be rediscovered.

Signed-off-by: Andreas Fenkart <afenkart@gmail.com>
Acked-by: Bing Zhao <bzhao@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
10 years agoMerge branch 'for-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetoot...
John W. Linville [Tue, 22 Jul 2014 17:50:23 +0000 (13:50 -0400)]
Merge branch 'for-upstream' of git://git./linux/kernel/git/bluetooth/bluetooth-next

10 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next
John W. Linville [Tue, 22 Jul 2014 17:49:34 +0000 (13:49 -0400)]
Merge git://git./linux/kernel/git/jberg/mac80211-next

10 years agoMerge branch 'iwlwifi-fixes' into iwlwifi-next
Emmanuel Grumbach [Tue, 22 Jul 2014 16:26:41 +0000 (19:26 +0300)]
Merge branch 'iwlwifi-fixes' into iwlwifi-next

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Conflicts:
drivers/net/wireless/iwlwifi/mvm/mac80211.c

10 years agoiwlwifi: mvm: reset beacon filtering and BT Coex data upon FW restart
Emmanuel Grumbach [Sun, 13 Jul 2014 06:12:11 +0000 (09:12 +0300)]
iwlwifi: mvm: reset beacon filtering and BT Coex data upon FW restart

When the firmware asserts, we restart the device and reset
the relevant data we hold in the driver. BT Coex data was
not reset and because of that, the driver wouldn't
reconfigure the firmware properly after firmware restart.
Same for beacon filtering. Fix that.

Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: mvm: update smart fifo / beacon filtering upon association
Eran Harary [Tue, 15 Jul 2014 11:04:23 +0000 (14:04 +0300)]
iwlwifi: mvm: update smart fifo / beacon filtering upon association

When we associate, we may have heard the beacon before the
association. In that case, BSS_CHANGED_BEACON_INFO will be
set along with BSS_CHANGED_ASSOC in changes in
bss_info_change.

In this case, we didn't update the smart fifo nor beacon
filtering leaving those two feature disabled.

Signed-off-by: Eran Harary <eran.harary@intel.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: dump CSRs to fw-error-dump
Emmanuel Grumbach [Tue, 8 Jul 2014 16:44:25 +0000 (19:44 +0300)]
iwlwifi: dump CSRs to fw-error-dump

Add the Control Status Registers to the firmware error dump
infrastructure.

Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: dump periphery registers to fw-error-dump
Emmanuel Grumbach [Thu, 26 Jun 2014 08:27:51 +0000 (11:27 +0300)]
iwlwifi: dump periphery registers to fw-error-dump

Use the fw-error-dump infrastructure to dump the periphery
registers. Only certain ranges are readable, so dump only
these.

Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: split fw-error-dump between transport and mvm
Emmanuel Grumbach [Tue, 8 Jul 2014 16:45:17 +0000 (19:45 +0300)]
iwlwifi: split fw-error-dump between transport and mvm

The mvm op_mode won't allocate the buffer for the transport
any more. The transport allocates its own buffer and mvm
is in charge of splicing the buffers in the debugfs hook.

This makes the repartition easier to handle.

Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: fix inconsistency about power_save module parameter
Emmanuel Grumbach [Mon, 21 Jul 2014 08:44:12 +0000 (11:44 +0300)]
iwlwifi: fix inconsistency about power_save module parameter

modinfo and kerneldoc disagreed on the meaning of this field.

Reported-by: Andrea Oliveri <oliveriandrea@gmail.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: mvm: minor change in debug print
Eran Harary [Tue, 1 Jul 2014 14:33:29 +0000 (17:33 +0300)]
iwlwifi: mvm: minor change in debug print

Add OTP to the string: "can't parse empty OTP/NVM section"
NVM usually refers to nvm_file while the problem can be in
the OTP.

Signed-off-by: Eran Harary <eran.harary@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: mvm: wait for handlers when stopping scans
Eliad Peller [Sun, 22 Jun 2014 15:15:59 +0000 (18:15 +0300)]
iwlwifi: mvm: wait for handlers when stopping scans

The recent unified scan api change introduced issues
when stopping ongoing scans, since both regular and
sched scan now use same stopped notification.

When issuing a new scan right after a running one,
we get the "old" notification and handle it wrongly
as notification for the current scan.

Fix it by introducing a new function that make sure
we consume the pending notifications before issuing
a new scan.

Signed-off-by: Eliad Peller <eliad@wizery.com>
Reviewed-by: ArikX Nemtsov <arik@wizery.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: mvm: BT Coex - don't change AP SMPS mode
Emmanuel Grumbach [Mon, 7 Jul 2014 11:38:39 +0000 (14:38 +0300)]
iwlwifi: mvm: BT Coex - don't change AP SMPS mode

Leave it to default instead - regardless of the BT activity.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: mvm: BT Coex - fix the ACK / CTS kill mask
Emmanuel Grumbach [Sun, 6 Jul 2014 06:34:38 +0000 (09:34 +0300)]
iwlwifi: mvm: BT Coex - fix the ACK / CTS kill mask

According to new requirements, the ACK / CTS kill mask is
not related to reduced TX power anymore. This allows to
remove the code that tracked reduced TX power enablement
across different interfaces.
The ACK / CTS kill mask is now fetch from a table. It
depends on the Activity grading (activity from BT) and on
the Look Up Table (LUT) type.

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoiwlwifi: mvm: add Aux ROC request/response flow
Ariej Marjieh [Wed, 16 Jul 2014 18:11:12 +0000 (21:11 +0300)]
iwlwifi: mvm: add Aux ROC request/response flow

The Remain On Channel framework added to the firmare is
a bit like time events. It allows the driver to request
the firmware to be on a certain channel for a certain time.
Unlike the time events, the ROC infrastructure doesn't need
a MAC context in the firmware - it uses a generic context
called "auxiliary framework".

This is useful for any offchannel activity that is not bound
to a specific MAC.
The flow is synchronized much like with time events:
1) The driver receives an action frame from the wpa_supplicant
   via nl80211 that requests to be sent offchannel.
2) The driver sends an Aux ROC command (0x53) to the firmware.
3) The firmware responds with the unique id of the time event.
4) When time event starts, the driver puts the frame in the
   Aux queue.

Special care needs to be taken when the time events ends:
the queue needs to be cleaned-up.

Signed-off-by: Ariej Marjieh <ariej.marjieh@intel.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
10 years agoath10k: workaround qos nullfunc bug
Michal Kazior [Mon, 21 Jul 2014 18:03:10 +0000 (21:03 +0300)]
ath10k: workaround qos nullfunc bug

Apparently fw/hw generates a corrupted QoS Control
Field in Qos NullFunc frames. The only way to
workaround this is to downgrade frames to
NullFunc. This should be okay since powersave is
done by fw/hw and these frames are only used for
CQM purposes (e.g. from hostapd to check if
station is still connected).

This doesn't fix any user visible bug that I know
of. It just prevents from sending out funky frames
on the air.

Reported-by: Janusz Dziedzic <janusz.dziedzic@tieto.com>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: simplify tx helpers
Michal Kazior [Mon, 21 Jul 2014 18:03:09 +0000 (21:03 +0300)]
ath10k: simplify tx helpers

It always bugged me how tid is computed and stored
in a temporary var before written to the control
buffer. It was confusing and it made it difficult
to work with tx helpers.

While at it rename the qos workaround function as
it was misleading - it's not a workaround but
preparation for nwifi tx mode.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: prevent some tx flushing failures
Michal Kazior [Mon, 21 Jul 2014 17:52:59 +0000 (20:52 +0300)]
ath10k: prevent some tx flushing failures

Firmware could request inspection of some
submitted tx requests. Since the callback wasn't
implemented it was possible to bleed tx msdu_ids
which could translate to tx flushing timeouts.

There's nothing ath10k can do to help firmware
with tx processing now so just report all tx
frames as already inspected to prevent firmware
from sending up inspection events and force it to
report regular tx completion indications with
discard status.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agoath10k: workaround boot issues with KVM/PCI-passthrough
Michal Kazior [Wed, 16 Jul 2014 13:01:39 +0000 (15:01 +0200)]
ath10k: workaround boot issues with KVM/PCI-passthrough

Apparently iomap writes that unmask CE irqs aren't
propagated properly sometimes. Before failing try
to poll for the control response message as it may
have been delivered without an interrupt.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
10 years agomac80211: add support for Rx reordering offloading
Michal Kazior [Wed, 16 Jul 2014 10:12:15 +0000 (12:12 +0200)]
mac80211: add support for Rx reordering offloading

Some drivers may be performing most of Tx/Rx
aggregation on their own (e.g. in firmware)
including AddBa/DelBa negotiations but may
otherwise require Rx reordering assistance.

The patch exports 2 new functions for establishing
Rx aggregation sessions in assumption device
driver has taken care of the necessary
negotiations.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
[fix endian bug]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: fix Rx reordering with RX_FLAG_AMSDU_MORE
Michal Kazior [Wed, 16 Jul 2014 10:09:31 +0000 (12:09 +0200)]
mac80211: fix Rx reordering with RX_FLAG_AMSDU_MORE

Some drivers (e.g. ath10k) report A-MSDU subframes
individually with identical seqno. The A-MPDU Rx
reorder code did not account for that which made
it practically unusable with drivers using
RX_FLAG_AMSDU_MORE because it would end up
dropping a lot of frames resulting in confusion in
upper network transport layers.

Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: remove useless NULL checks
Eytan Lifshitz [Mon, 21 Jul 2014 12:18:41 +0000 (15:18 +0300)]
mac80211: remove useless NULL checks

sdata can't be NULL, and key being NULL is really not possible
unless the code is modified.

The sdata check made a static analyze (klocwork) unhappy because
we would get pointer to local (sdata->local) and only then check
if sdata is non-NULL.

Signed-off-by: Eytan Lifshitz <eytan.lifshitz@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
[remove !key check as well]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agoBluetooth: Add support for Broadcom device of Asus Z97-DELUXE motherboard
Marcel Holtmann [Mon, 21 Jul 2014 12:02:33 +0000 (14:02 +0200)]
Bluetooth: Add support for Broadcom device of Asus Z97-DELUXE motherboard

The Asus Z97-DELUXE motherboard contains a Broadcom based Bluetooth
controller on the USB bus. However vendor and product ID are listed
as ASUSTek Computer.

T:  Bus=01 Lev=01 Prnt=01 Port=01 Cnt=02 Dev#=  3 Spd=12   MxCh= 0
D:  Ver= 2.00 Cls=ff(vend.) Sub=01 Prot=01 MxPS=64 #Cfgs=  1
P:  Vendor=0b05 ProdID=17cf Rev= 1.12
S:  Manufacturer=Broadcom Corp
S:  Product=BCM20702A0
S:  SerialNumber=54271E910064
C:* #Ifs= 4 Cfg#= 1 Atr=e0 MxPwr=  0mA
I:* If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=81(I) Atr=03(Int.) MxPS=  16 Ivl=1ms
E:  Ad=82(I) Atr=02(Bulk) MxPS=  64 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS=  64 Ivl=0ms
I:* If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=   0 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=   0 Ivl=1ms
I:  If#= 1 Alt= 1 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=   9 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=   9 Ivl=1ms
I:  If#= 1 Alt= 2 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  17 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  17 Ivl=1ms
I:  If#= 1 Alt= 3 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  25 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  25 Ivl=1ms
I:  If#= 1 Alt= 4 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  33 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  33 Ivl=1ms
I:  If#= 1 Alt= 5 #EPs= 2 Cls=ff(vend.) Sub=01 Prot=01 Driver=btusb
E:  Ad=83(I) Atr=01(Isoc) MxPS=  49 Ivl=1ms
E:  Ad=03(O) Atr=01(Isoc) MxPS=  49 Ivl=1ms
I:* If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=(none)
E:  Ad=84(I) Atr=02(Bulk) MxPS=  32 Ivl=0ms
E:  Ad=04(O) Atr=02(Bulk) MxPS=  32 Ivl=0ms
I:* If#= 3 Alt= 0 #EPs= 0 Cls=fe(app. ) Sub=01 Prot=01 Driver=(none)

Reported-by: Jerome Leclanche <jerome@leclan.ch>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
10 years agoBluetooth: Prefer sizeof(*ptr) when allocating memory
Johan Hedberg [Mon, 21 Jul 2014 07:50:06 +0000 (10:50 +0300)]
Bluetooth: Prefer sizeof(*ptr) when allocating memory

It's safer practice to use sizeof(*ptr) instead of sizeof(ptr_type) when
allocating memory in case the type changes. This also fixes the
following style of warnings from static analyzers:

CHECK: Prefer kzalloc(sizeof(*ie)...) over kzalloc(sizeof(struct inquiry_entry)...)
+ ie = kzalloc(sizeof(struct inquiry_entry), GFP_KERNEL);

Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
10 years agomac80211: fix a potential NULL access in ieee80211_crypto_hw_decrypt
Max Stepanov [Wed, 9 Jul 2014 13:55:32 +0000 (16:55 +0300)]
mac80211: fix a potential NULL access in ieee80211_crypto_hw_decrypt

The NULL pointer access could happen when ieee80211_crypto_hw_decrypt
is called from ieee80211_rx_h_decrypt with the following condition:
1. rx->key->conf.cipher is not WEP, CCMP, TKIP or AES_CMAC
2. rx->sta is NULL

When ieee80211_crypto_hw_decrypt is called, it verifies
rx->sta->cipher_scheme and it will cause Oops if rx->sta is NULL.

This path adds an addirional rx->sta == NULL verification in
ieee80211_crypto_hw_decrypt for this case.

Signed-off-by: Max Stepanov <Max.Stepanov@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agowireless: fixup genregdb.awk for remove of antenna gain from wireless-regd
Luis R. Rodriguez [Mon, 14 Jul 2014 21:19:49 +0000 (14:19 -0700)]
wireless: fixup genregdb.awk for remove of antenna gain from wireless-regd

Since "wireless-regdb: remove antenna gain" was merged in the
wireless-regdb tree, the awk script parser has been incompatible
with the 'official' regulatory database.  This fixes that up.
Without this change the max EIRP is set to 0 making 802.11 devices
useless.

The fragile nature of the awk parser must be replaced, but ideas
over how to do that in the most scalable way are being reviewed.
In the meantime update the documentation for CFG80211_INTERNAL_REGDB
so folks are aware of expectations for now.

Reported-by: John Walker <john@x109.net>
Reported-by: Krishna Chaitanya <chaitanya.mgit@gmail.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@suse.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: remove redundant IEEE80211_STA_CSA_RECEIVED flag
Luciano Coelho [Sat, 5 Jul 2014 08:43:01 +0000 (11:43 +0300)]
mac80211: remove redundant IEEE80211_STA_CSA_RECEIVED flag

The csa_active flag was added in sdata a while ago and made
IEEE80211_STA_CSA_RECEIVED redundant.  The new flag is also used to
mark when CSA is ongoing on other iftypes and took over the old one as
the preferred method for checking whether we're in the middle of a
channel switch.  Remove the old, redundant flag.

Signed-off-by: Luciano Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: make sure TDLS teardown packet is sent on time
Liad Kaufman [Thu, 17 Jul 2014 14:14:31 +0000 (17:14 +0300)]
mac80211: make sure TDLS teardown packet is sent on time

Since the teardown packet is created while the queues are
stopped, it isn't sent immediately, but rather is pending.
To be sure that when we flush the queues prior to destroying
the station we also send this packet - the tasklet handling
pending packets is invoked to flush the packets.

Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Reviewed-by: ArikX Nemtsov <arik@wizery.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: ignore frames between TDLS peers when operating as AP
Arik Nemtsov [Thu, 17 Jul 2014 14:14:30 +0000 (17:14 +0300)]
mac80211: ignore frames between TDLS peers when operating as AP

If the AP receives actions frames destined for other peers, it may
mistakenly toggle BA-sessions from itself to a peer.
Ignore TDLS data packets as well - the AP should not handle them.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agocfg80211: fix TDLS setup with VHT peers
Arik Nemtsov [Thu, 17 Jul 2014 14:14:29 +0000 (17:14 +0300)]
cfg80211: fix TDLS setup with VHT peers

Some VHT TDLS peers (Google Nexus 5) include the VHT-AID IE in their
TDLS setup request/response. Usermode passes this aid as the station
aid, causing it to fail verifiction, since this happens in the
"set_station" stage. Make an exception for the TDLS use-case.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: disable VHT for TDLS
Arik Nemtsov [Thu, 17 Jul 2014 14:14:28 +0000 (17:14 +0300)]
mac80211: disable VHT for TDLS

TDLS VHT support requires some more information elements during setup.
While these are not there, mask out the peer's VHT capabilities so that
VHT rates are not mistakenly used.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: set Rx highest rate in ht_cap
Arik Nemtsov [Thu, 17 Jul 2014 14:14:27 +0000 (17:14 +0300)]
mac80211: set Rx highest rate in ht_cap

Set for completeness mostly, currently unused in the code.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: support HT for TDLS stations
Arik Nemtsov [Thu, 17 Jul 2014 14:14:26 +0000 (17:14 +0300)]
mac80211: support HT for TDLS stations

Add the HT capabilities and HT operation information elements to TDLS
setup packets where appropriate.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: move TDLS data to mgd private part
Arik Nemtsov [Thu, 17 Jul 2014 14:14:25 +0000 (17:14 +0300)]
mac80211: move TDLS data to mgd private part

We can only be a station for TDLS connections. Also fix a bug where
a delayed work could be left scheduled if the station interface was
brought down during TDLS setup.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: add TDLS QoS param IE on setup-confirm
Arik Nemtsov [Thu, 17 Jul 2014 14:14:24 +0000 (17:14 +0300)]
mac80211: add TDLS QoS param IE on setup-confirm

When TDLS QoS is supported by the the peer and the local card, add
the WMM parameter IE to the setup-confirm frame. Take the QoS settings
from the current AP, or if unsupported, use the default values from
the specification. This behavior is mandated by IEEE802.11-2012 section
10.22.4.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Liad Kaufman <liad.kaufman@intel.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
10 years agomac80211: add QoS IE during TDLS setup start
Arik Nemtsov [Thu, 17 Jul 2014 14:14:23 +0000 (17:14 +0300)]
mac80211: add QoS IE during TDLS setup start

If QoS is supported by the card, add an appropriate IE to TDLS setup-
request and setup-response frames.

Consolidate the setting of the WMM info IE across mac80211.

Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Reviewed-by: Liad Kaufman <liad.kaufman@intel.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>