In order to allow user space based QOS control we use the available debugfs
infrastructure. With this feature, user can make changes to qos parameters,
thereby allowing creation and deletion of user defined priority streams and
features like uapsd. This feature has been added for testing purposes.

All 21 parameters for the create_qos command are mandatory in the correct
order. They have to be written to the create_qos file in
the ath6kl debug directory. These parameters(in order) are:

1)user priority
2)direction
3)traffic class
4)traffic type
5)voice PS capability
6)min service intvl
7)max service intvl
8)inactivity intvl
9)suspension intvl
10)serv start time
11)tsid
12)nominal msdu
13)max msdu
14)min data rate
15)mean data rate
16)peak data rate
17)max burst size
18)delay bound
19)min phy rate
20)surplus bw allowance
21)medium time

To create a qos stream:

echo "6 2 3 1 1 9999999 9999999 9999999 7777777 0 6 45000 200 56789000
56789000 5678900 0 0 9999999 20000 0" > create_qos

delete_qos requires 2 parameters:

1)traffic class
2)tsid

To delete a qos stream:

echo "3 1" > delete_qos

kvalo: minor commit log cleanup
Signed-off-by: NRishi Panjwani <rpanjwan@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

The previous version did not really make much sense and the theoretical
maximum length would be a bit longer. Calculate the length more
accurately. In addition, there is no need to clear the buffer, so use
kmalloc instead of kzalloc. For bonus points, add the forgotten
cred_rpt_from_other value to the file.
Reported-by: NJoe Perches <joe@perches.com>
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Add missing {}'s that caused ath6kl_sdio_suspend to always return -EINVAL
causing suspend to be aborted.

kvalo: I broke this in commit f7325b85 ("ath6kl: add sdio debug messages")
Signed-off-by: NSam Leffler <sleffler@chromium.org>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Sangwook found out that commit 639d0b89 ("ath6kl: read firmware start
address from hardware") broke firmware boot on ar6003 hw2.0 as it seems
it's not posible to automatically query the address from hardware. So
we need to hardcode the address for hw2.0.
Reported-by: NSangwook Lee <sangwook.lee@linaro.org>
Tested-by: NSangwook Lee <sangwook.lee@linaro.org>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Use vendor-specific suite selectors to allow CCKM to be configured.
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

The new debugfs files keepalive and disconnect_timeout can be used to
fetch the current values and to change the values for keepalive and
disconnect event timeout (both in seconds).
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Roaming mode can be changed by writing roam mode (default, bssbias, or
lock) to roam_mode. Forced roam can be requested by writing the BSSID
into force_roam.
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

The new roam_table debugfs file can be used to display the current
roam table from the target.
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

This file can be used to fetch endpoint statistics counters and
to clear them by writing 0 to it.
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

For some WPS test items, such as item "5.1.14"
STAUT must include the WSC IE in the 802.11 Association Request frame.
Therefore, add the corresponding IE in association message.
Signed-off-by: NKevin Fang <kevin.fang@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

smatch found that skb might be null in some cases in ath6kl_rx():

ath6kl/txrx.c +1252 ath6kl_rx(222) error: potential null derefence 'skb'.

This will happen when ath6kl is in AP mode and two clients send traffic
to each other.
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Now htc_hif.h can be removed.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Also remove some cache line optimisation. It was using hardcoded values
which is wrong.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Currently only a warning is emitted but no error is returned.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It was missing endian annotation.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

There's no use for these, at least right now, so better to remove them.
If some of them are ever needed, we can always add them back.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Power save is enabled during ath6kl init. But when user space disables power
save, the system will go into suspend with power save disabled. The ath6kl
driver will now explicitly enable power save prior to entering suspend and
restore its previous setting upon resume
Signed-off-by: NChilam Ng <chilamng@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It is just a four byte information of the received message
from ath6kl_htc_rxmsg_pending_handler(). Remove unnecessary
array representaion.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It is not necessary to process an htc_packet which is allocated for rx
but failed in sdio rx. Though it does not fix any real issue, it does
not make much sense to process the failed frame.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It is found during the code review. As the leak would happen only
in failure case, the imapct is not easily visible.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Packet is not reclaimed when ath6kl_htc_rx_process_hdr() fails.
Fix this by deferring the packet deletion from comp_pktq till
ath6kl_htc_rx_process_hdr() returns success. This bug is found
in code review, impact is not easily visible as the leak happens
only in failure cases.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

There are places where tx_complete callbacks are called with
claiming the sdio function. It is not necessary to hold the
sdio func for longer. This may even affect the host side power
save, if it is supported by the controller.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It is not necessary to use spinlock primitive to
protect data which is accessed in hard irq context as
nothing is running in hard irq with this driver. The
spinlock primitive meant to protect data in softirq
context is more appropriate.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

This was disabled beacause there was a network stall
issue when scan is issued. This issue does not happen
with the new firmware (3.1.1.609), enable it back.
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

For probe responses it can be useful to not wait for ACK to
avoid retransmissions if the station that sent the probe is
already on the next channel, so allow userspace to request
not caring about the ACK with a new nl80211 flag.

Since mac80211 needs to be updated for the new function
prototype anyway implement it right away -- it's just a
few lines of code.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Add the ability to advertise that the device
contains the AP SME and what features it can
support. There are currently no features in
the bitmap -- probe response offload will be
advertised by a few patches Arik is working
on now (who took over from Guy Eilam) and a
device with AP SME will typically implement
and require response offload.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

  CC [M]  drivers/net/wireless/ath/ath6kl/cfg80211.o
drivers/net/wireless/ath/ath6kl/cfg80211.c:1838:2: warning: initialization from incompatible pointer type

Caused by commit e9f935e3...
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Fixes compilation errors when compiling for ARM:

ath6kl/debug.c:312: error: implicit declaration of function 'vmalloc'
ath6kl/debug.c:312: warning: assignment makes pointer from integer without a cast
ath6kl/debug.c:342: error: implicit declaration of function 'vfree'
ath6kl/debug.c:696: warning: assignment makes pointer from integer without a cast
ath6kl/debug.c:871: warning: assignment makes pointer from integer without a cast
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

My earlier debug log additions added these warnings when compiling 64 bit
kernels:

ath6kl/init.c:962: warning: format '%d' expects type 'int',
  but argument 3 has type 'size_t'
ath6kl/init.c:975: warning: format '%d' expects type 'int',
  but argument 3 has type 'size_t'
ath6kl/init.c:988: warning: format '%d' expects type 'int',
  but argument 3 has type 'size_t'
ath6kl/init.c:1009: warning: format '%d' expects type 'int',
  but argument 3 has type 'size_t'
ath6kl/init.c:1192: warning: format '%d' expects type 'int',
  but argument 4 has type 'size_t'
ath6kl/init.c:1236: warning: format '%d' expects type 'int',
  but argument 4 has type 'size_t'
ath6kl/init.c:1267: warning: format '%d' expects type 'int',
  but argument 4 has type 'size_t'
Reported-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Otheriwse the module.h split up fails like this:

drivers/net/wireless/ath/ath6kl/init.c:27:26: error: expected ')' before 'uint'
Signed-off-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Add a new debug level ATH6KL_DBG_WMI_DUMP and other minor
improvements to the wmi debug messages.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Just to make it easier to find out why boot fails.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Add extensive debug messages to sdio.c. Makes it easier to debug
various problems.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Makes it easier to recognise longs dumps.

Obligatory screenshot using "rx" prefix:

ath6kl: ath6kl_rx
rx 00000000: 10 10 00 00 00 00 08 30 00 00 00 00 00 00 f9 0b  .......0........
rx 00000010: 2c 44 08 30 00 00 f9 0b 0c a4 02 00 00 00 73 d2  ,D.0..........s.
rx 00000020: 94 00 f9 0b 04 8c 01 00 02 00 07 02 02 00 f9 0b  ................
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

For some reason firmware is sending invalid rates when we try to
query current bitrate from ath6kl_get_station() and a warning is issued:

[ 3810.415720] ath6kl: invalid rate: 1935633515
[ 3811.105493] ath6kl: invalid rate: 1935633515
[ 3811.556063] ath6kl: invalid rate: 1935633515

As the warning happens way too often, convert the warning to a debug
message once we have a proper fix. But to make it easy to follow
how often the problem appears, add a debugfs to print
various statistics about workarounds and make this issue the first WAR.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

In some firmware versions their patch address has changed. If the firmware
provides one, use it to override the default address.
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Commit 94e532d1 ("ath6kl: Fix system freeze under heavy data load")
aligns the skb data without checking if the skb is cloned. Because of
this ath6kl can corrupt the local TCP stack information that can result
in TCP retransmission failing and TCP connections stalling.

To avoid the corruption we need to copy the skb. Now the alignment
in ath6kl_htc_tx_buf_align() doesn't corrupt TCP packets anymore (and is
not even used for the cloned skb's that got copied since the alignment
of the data is handled at the copy time).
Signed-off-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

It is unnecessary to take the first list entry from queue again for
transmission. Sometimes it may look racy when the head of the list
changes between subsequent retrival, but should not happen in
practical.
Reported-by: NJouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: NVasanthakumar Thiagarajan <vthiagar@qca.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>