For VHT peers, the maximum A-MPDU size has to be calculated
from the VHT capabilities element and not the HT-cap. The formula
is the same, but a higher value is used in VHT, allowing larger
aggregates to be transmitted.

The patch contains a workaround for some Netgear/Linksys APs that
report Rx A-MPDU factor incorrectly.
Tested-by: Kalle Valo <kvalo@qca.qualcomm.com>
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Due to oversight AP 10.X support was merged with
Ethernet RX decap mode.

Only Native Wifi RX decap mode guarantees IP
header is properly aligned and avoids sk_buff data
realignment (which is very expensive
performance-wise) in mac80211.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Some failpaths did `return` instead of a `goto`
leaking requested firmware.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

size_t corresponds to %zu not %d. Compiler was
complaining about it.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

The value provided by num_sends_allowed is now
derived from CE source ringbuffer state.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

It is completely pointless to keep this function
around. It doesn't do anything different than
ce_send except it introduces more overhead.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Tidy up ath10k_tx().
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Tidy up ath10k_tx().
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

It was observed few times, the artificial max_scan limit we
are using mainly to detect FW hangs, can be not enough for
instance while being associated and during heavy traffic.

What we do if the FW won't return with scan results within
the max_time time is a scan abort.

This is especially visible with 10.X fw which in combination
with dual band HW (scanning 32 channels) can end up with
hw_scan close to 10 seconds.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Since the WMI API has been added and we can detect from the
FW IEs what firmware variant we deal with, turn on support
for 10.x firmware branch in ath10k_wmi_attach().

kvalo: improve the commit log
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Fix the warning

drivers/net/wireless/ath/ath6kl/htc_pipe.c: In function
'ath6kl_htc_pipe_conn_service':
drivers/net/wireless/ath/ath6kl/htc_pipe.c:1293:26: warning: integer overflow
in expression [-Woverflow]

by giving a hint to compiler about unsigned nature of
HTC_CONN_FLGS_SET_RECV_ALLOC_MASK
Signed-off-by: Vladimir Murzin <murzin.v@gmail.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Firmware IE containers can dynamically provide various information what
firmware supports. Also it can embed more than one image so updating firmware
is easy, user just needs to update one file in /lib/firmware/.

The firmware API 2 or higher will use the IE container format, the current API
1 will not use the new format but it still is supported for some time. FW API 2
files are named as firmware-2.bin (which contains both firmware and otp images)
and API 1 files are firmware.bin and otp.bin.
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Needed for firmware IE formatted images.
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

I will use board_data for something else in the following patch
so I need to rename it.
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

The wmi_start_scan_cmd has an extra filed in our main
firmware track, reflact that to not have a mismatch in
case of 10.x track.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

This is done exactly the same way as for vdev.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Both firmwares (main and 10.x) have different set of vdev
parameters. To stay in sync with FW API, this patch introduces
a dynamic registering method.

ath10k_wmi_vdev_set_param() takes now indirect u32 value
to identify the Vdev parameter it want's to set.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Introduce all TARGET specific values for 10.x firmware.
Some of them are common for both firmwares we will support,
but to avoid confusion, define everything with prefix 10X_.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Due to API differences in initialization structures for
main and 10.x firmwares we need to split the wmi_init_cmd
and wmi_resource_config  structures.

This will be usefull also when setting the correct TARGET values,
like: number of peers, vdevs, pdevs etc.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

This is still the only way to submit mgmt frames in case
of 10.X firmware.

This patch introduces wmi_mgmt_tx queue, because of the
fact WMI command can block. This is a problem for
ath10k_tx_htt(), since it's called from atomic context.
The skb queue and worker are introduced to move the mgmt
frame handling out of .tx callback context and not block.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

10.X firmware can request a memory pool from host to offload
it's own resources. This is a feature designed especially
for AP mode where the target has to deal with large number
of peers.

So we allocate and map a consistent DMA memory which FW can
use to store e.g. peer rate contol maps.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

It was corrupted and leading to compilation warning
core.c: In function 'ath10k_check_fw_version':
core.c:79: warning: comparison is always true due to limited range of data type

Since we are going to support new FW track in the driver,
the sanity check for fw version number would be an overkill
then. This is just for information purposes anyway.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Since the both firmwares we are going to support,
have significantly different APIs (WMI and shared
structures), it's easier to actually split the whole
event handling functions, instead cutting them inside.

The fork starts now on ath10k_wmi_process_rx().
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

We still use the same function handlers for both main and 10.X
paths. Next step is to track down and make the split needed.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

This will show and make it easier to track the API
differences in the new AP firmware.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Introduce the cmd and events definitions for 10.X FW API
and make up the wmi_10x_cmd_map. This is the core of
host-firmware WMI interface for 10.X FW branch.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

This is the initial framework to implement dynamic
WMI API in order to support new firmware (from so
called 10.X branch)

The realization is to have a static WMI cmd map for
each of the firmwares, registered upom wmi init.

This patch creates such map for MAIN FW, updates
wmi_cmd_send() calls to take as a parameter
the map value instead of direct WMI enum.

As soon as complete 10.X API will be on place,
we will introduce the FW IE mechanics to dynamicaly
identify which FW is being used and based on that
we will use correct map, API, structures, etc.
Signed-off-by: Bartosz Markowski <bartosz.markowski@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Following sequence causes FW crash:
-monitor vdev up,
-monitor vdev stop,
-monitor vdev delete.
Making monitor vdev down before stoping it works ok:
-monitor vdev up,
-monitor vdev down,
-monitor vdev stop,
-monitor vdev delete.
Signed-off-by: Marek Puzyniak <marek.puzyniak@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

This starves FW RX ring buffer in case of
excessive RX. This prevents from CPU being
overwhelmed by RX indications/completions by
naturally forbiddin FW to submit more RX.

This fixes RX starvation on slow machines when
under heavy RX traffic.

kvalo: remove extra newline
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Ethernet-like decapping mode leaves IP protocol
frame not aligned to 4-byte boundaries. This leads
to re-aligning in mac80211 which in turn leads to
poor CPU cache behaviour on some machines.

Since HW doesn't allow to change payload offset
properly the solution is to force HW to decap in
Native Wifi mode which always has 24-bytes long
802.11 header (even for QoS frames). This means IP
frame is properly aligned in this decap mode.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

NWifi decap mode always reports 802.11 Data
Frames, even when QoS Data Frames are actually
received.

This made mac80211 not report frame priority
properly (since there was no QoS Control field).
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Simplify decapping code and make it easier to
understand.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Clarify how each decap mode works in one place.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

HW reports each A-MSDU subframe as a separate
sk_buff. It is impossible to configure it to
behave differently.

Until now ath10k was reconstructing A-MSDUs from
subframes which involved a lot of memory
operations. This proved to be a significant
contributor to degraded RX performance.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Instead of allocating sk_buff for a mere 16-byte
tx fragment list buffer use headroom of the
original msdu sk_buff.

This decreases CPU cache pressure and improves
performance.

Measured improvement on AP135 is 560mbps ->
590mbps of UDP TX briding traffic.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Use a saner goto scheme for failure handling. Also
group operations more sensibly.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Until now the all MSDU transfer related structures
were freed when all resources were unreferenced.

Now HTC transfer is freed independently and HTT
transfer is so too.

This yields a way more simpler ath10k_skb_cb and
should possibly enable parallel pipe processing
(which is now serialized in
ath10k_pci_process_ce routine).
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

This reduces number of memory accesses and
hopefully contributes to better performance in the
future.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

It's more efficient to simply check num_pending_tx
value instead of traversing whole bitmap of
msdu ids.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>

Commit e9bb0aa3 ("ath10k: delete struct ce_sendlist") broke
num_sends_allowed incrementing. num_sends_allowed
exceeded initial values and could overflow.

This code was supposed to replenish
num_sends_allowed for partial sendlist items (i.e.
before final sendlist item from a sendlist was
completed and could be processed by completion
handlers).

Fortunately it seems it did not cause any major breakage,
yet.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>