Miaoqing Pan [Fri, 18 Mar 2016 09:54:56 +0000 (17:54 +0800)]
ath9k: fix rng high cpu load
If no valid ADC randomness output, ath9k rng will continuously
reading ADC, which will cause high CPU load. So increase the
delay to wait for ADC ready.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=114261 Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Mon, 21 Mar 2016 20:01:11 +0000 (22:01 +0200)]
wil6210: allow empty WMI commands in debugfs wmi_send
There are many valid WMI commands with only header without any
additional payload. Such WMI commands could not be sent using
the debugfs wmi_send facility. Fix the code to allow sending
of such commands.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Joseph Salisbury [Mon, 14 Mar 2016 18:51:48 +0000 (14:51 -0400)]
ath5k: Change led pin configuration for compaq c700 laptop
BugLink: http://bugs.launchpad.net/bugs/972604
Commit 09c9bae26b0d3c9472cb6ae45010460a2cee8b8d ("ath5k: add led pin
configuration for compaq c700 laptop") added a pin configuration for the Compaq
c700 laptop. However, the polarity of the led pin is reversed. It should be
red for wifi off and blue for wifi on, but it is the opposite. This bug was
reported in the following bug report:
http://pad.lv/972604
Fixes: 09c9bae26b0d3c9472cb6ae45010460a2cee8b8d ("ath5k: add led pin configuration for compaq c700 laptop") Signed-off-by: Joseph Salisbury <joseph.salisbury@canonical.com> Cc: stable@vger.kernel.org Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Peter Oh [Tue, 22 Mar 2016 22:44:53 +0000 (15:44 -0700)]
ath10k: parse Rx MAC timestamp in mgmt frame for FW 10.4
Check and parse Rx MAC timestamp when firmware sets its flag
to status variable.
10.4 firmware adds it in Rx beacon frame only at this moment.
Drivers and mac80211 may utilize it to detect such clockdrift
or beacon collision and use the result for beacon collision
avoidance.
Signed-off-by: Peter Oh <poh@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Since tx completion and rx indication processing are moved out
of txrx tasklet and rx ring lock contention also removed from txrx
for rx_ind messages, it would be efficient to combine both replenish
and txrx tasks. Refill threshold is adjusted for both AP135 and AP148
(low and high end systems). With this adjustment in AP135, TCP DL is
improved from 603 Mbps to 620 Mbps and UDP DL is improved from 758 Mbps
to 803 Mbps. Also no watchdog are observed on UDP BiDi.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Whenever htt rx indication i.e target to host messages are received
on rx copy engine (CE5), the message will be freed after processing
the response. Then CE 5 will be refilled with new descriptors at
post rx processing. This memory alloc and free operations can be avoided
by reusing the same descriptors.
During CE pipe allocation, full ring is not initialized i.e n-1 entries
are filled up. So for CE 5 full ring should be filled up to reuse
descriptors. Moreover CE 5 write index will be updated in single shot
instead of incremental access. This could avoid multiple pci_write and
ce_ring access. From experiments, It improves CPU usage by ~3% in IPQ4019
platform.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: cleanup copy engine receive next completion
The physical address necessary to unmap DMA ('bufferp') is stored
in ath10k_skb_cb as 'paddr'. For diag register read and write
operations, 'paddr' is stored in transfer context. ath10k doesn't rely
on the meta/transfer_id. So the unused output arguments {bufferp, nbytesp
and transfer_idp} are removed from CE recv_next completion.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Except qca61x4 family chips (qca6164, qca6174), copy engine 5 is used
for receiving target to host htt messages. In follow up patch, CE5
descriptors will be reused. In such case, same API can not be used as
htc layer callback where the response messages will be freed at the end.
Hence register new API for HTC layer that free up received message and
keep the message handler common for both HTC and HIF layers.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: speedup htt rx descriptor processing for rx_ind
In follow up patch, htt rx descriptors will be reused instead of
dealloc and refill. To achieve that htt rx indication messages
should not be deferred and should be processed in pci tasklet itself.
Also from rx indication message, mpdu_count alone is used. So it is
maintained as atomic variable and all rx amsdu handlers are done
processed from txrx tasklet. This change get rid of rx_compl_q usage.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: cleanup amsdu processing for rx indication
Make amsdu handlers (i.e amsdu_pop and rx_h_handler) common to both
rx_ind and frag_ind htt events. It is sufficient to hold rx_ring lock
for amsdu_pop alone and no need to hold it until the packets are
delivered to mac80211. This helps to reduce rx_lock contention as well.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
The fw descriptor was never used and probably never will be. It makes
little sense to maintain support for it. Remove it and simplify rx
processing. This will make it easier to optimize rx processing later
as well.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
To optmize CPU usage htt rx descriptors will be reused instead of
refilling it for htt rx copy engine (CE5). To support that all htt rx
indications should be proecssed at same context. Instead of queueing
actual indication message, queue copied message for txrx processing.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: speedup htt rx descriptor processing for tx completion
To optimize CPU usage htt rx descriptors will be reused instead of
refilling it for htt rx copy engine (CE5). To support that all htt rx
indications should be processed at same context. FIFO queue is used
to maintain tx completion status for each msdu. This helps to retain
the order of tx completion.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: enable parsing per station rx duration for 10.4
Rx duration support for per station is part of extended peer
stats, enable provision to parse the same and provide backward
compatibility based on the 'stats_id' event
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 16 Mar 2016 12:43:33 +0000 (18:13 +0530)]
ath10k: introduce Extended Resource Config support for 10.4
Add API support for Extended Resource Configuration for 10.4. This
is useful to enable new features like Peer Stats, LTEU etc if the
firmware advertises support for the service. This is also done to
provide backward compatibility with older firmware. Also for clarity
send default host platform type as 'WMI_HOST_PLATFORM_HIGH_PERF',
though this should not make any difference in functionality
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: enable debugfs provision to enable Peer Stats feature
Provide a debugfs entry to enable/ disable Peer Stats feature.
Peer Stats feature is for developers/users who are more interested
in studying in Rx/Tx stats with multiple clients connected, hence
disable this by default. Enabling this feature by default results
in unneccessary processing of Peer Stats event for every 500ms
and updating peer_stats list (allocating memory) and cleaning it up
ifexceeds the higher limit and this can be an unnecessary overhead
during long run stress testing.
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Fri, 18 Mar 2016 09:44:22 +0000 (11:44 +0200)]
ath10k: incorporate qca4019 cal data download sequence
qca4019 calibration data is stored in the host memory and it's mandatory
to download it even before reading board id and chip id from the target.
Also, there is a need to execute otp (download and run) twice, one after
cal data download and another one after board data download.
Existing cal data file name 'cal-<bus>-<id>.bin' and device tree entry
'qcom,ath10k-calibration-data' used in ath10k has assumption that it
carries other data (like board data) also along with the calibration data.
But, qca4019 cal data contains pure calibration data (doesn't include
any other info). So, using existing same cal file name and DT entry
in qca4019 case would alter the purpose of it. To avoid this, new cal
file name 'pre-cal-<bus>-<id>.bin' and new device tree entry name
'qcom,ath10k-pre-calibration-data are introduced.
Overall qca4019's firmware download sequence would look like,
1) Download cal data (either from a file or device tree entry)
at the address specified by target in the host interest area
member "hi_board_data".
2) Download otp and run with 0x10 (PARAM_GET_EEPROM_BOARD_ID)
as a argument.
At this point, otp will take back up of downloaded cal data
content in another location in the target and return valid
board id and chip id to the host.
3) Download board data at the address specified by target
in host interest area member "hi_board_data".
4) Download otp and run with 0x10000 (PARAM_FLASH_SECTION_ALL) as
a argument.
Now otp will apply cal data content from it's backup on top
of board data download in step 3 and prepare final data base.
5) Download code swap and athwlan binary content.
Above sequences are implemented (step 1 to step 4) in the name of
pre calibration configuration.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Fri, 18 Mar 2016 09:44:22 +0000 (11:44 +0200)]
ath10k: move cal data len to hw_params
ath10k_download_cal_dt() compares obtained cal data content length
against QCA988X_CAL_DATA_LEN (2116 bytes). It was written by keeping
qca988x in mind. In fact, cal data length is more chip specific.
To make ath10k_download_cal_dt() more generic and reusable for other
chipsets (like qca4019), cal data length is moved to hw_params.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Fri, 18 Mar 2016 09:44:21 +0000 (11:44 +0200)]
ath10k: pass cal data location as an argument to ath10k_download_cal_{file|dt}
Both ath10k_download_cal_file() and ath10k_download_cal_dt() uses
hard coded file pointer (ar->cal_file) and device tree entry
(qcom,ath10k-calibration-data) respectively to get calibration
data content.
There is a need to use those two functions in qca4019 calibration
download sequence with different file pointer and device tree entry name.
Modify those two functions to take cal data location as an argument.
So that it can serve the purpose for other file pointer and device
tree entry.
This is just preparation before adding actual qca4019 calibration
download sequence. No functional changes.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Fri, 18 Mar 2016 09:44:21 +0000 (11:44 +0200)]
dt: bindings: add new dt entry for pre calibration in qcom, ath10k.txt
There two things done in this patch,
1) Existing device tree entry 'qcom,ath10k-calibration-data' carries
not only calibration data, it carries board specific data too.
So, make appropriate update in doc.
2) ipq4019 wifi needs new devie tree entry to carry calibration
data alone (called pre cal data, it doesn't include any other info).
Using 'qcom,ath10k-calibration-data' for ipq4019 would alter
the purpose of it. Hence, add new device tree entry called
'qcom,ath10k-pre-calibration-data' to carry only pre calibration data.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Thu, 17 Mar 2016 09:52:08 +0000 (10:52 +0100)]
ath10k: fix null deref if device crashes early
If device failed to init during early probing
(which is quite rare) it triggered driver to
compute crc before ar->firmware was ready causing
an oops.
Fixes: 3e58044b61a9 ("ath10k: print crc32 checksums for firmware and board files") Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Thu, 17 Mar 2016 09:51:05 +0000 (10:51 +0100)]
ath10k: fix pull-push tx threshold handling
This prevents tx hangs or hiccups if pull-push
supporting firmware defines per-txq thresholds or
switches modes dynamically.
Fixes: 299468782d94 ("ath10k: implement wake_tx_queue") Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Thu, 17 Mar 2016 09:51:04 +0000 (10:51 +0100)]
ath10k: fix tx hang
The wake_tx_queue/push_pending logic had a bug
which could stop queues indefinitely effectivelly
breaking traffic.
Fixes: 299468782d94 ("ath10k: implement wake_tx_queue") Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Results obtained from scan can be used for spectrum management by
doing something like building information of preferred channel
lists and sharing them with stations around. It is to be noted
that traffic to the connected stations would be affected during
the scan.
Signed-off-by: Vasanthakumar Thiagarajan <vthiagar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Anilkumar Kolli [Fri, 11 Mar 2016 06:16:39 +0000 (11:46 +0530)]
ath10k: fix debugfs pktlog_filter write
It is observed that, we are disabling the packet log if we write same
value to the pktlog_filter for the second time. Always enable pktlogs
on non zero filter.
Fixes: 90174455ae05 ("ath10k: add support to configure pktlog filter") Cc: stable@vger.kernel.org Signed-off-by: Anilkumar Kolli <akolli@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Thu, 10 Mar 2016 04:55:07 +0000 (10:25 +0530)]
ath10k: free cached fw bin contents when get board id fails
ath10k_core_probe_fw() simply returns error without freeing
cached firmware file content when get board id operation fails.
Free cached fw bin data in failure case to avoid memory leak.
Fixes: db0984e51a18 ("ath10k: select board data based on BMI chip id and board id") Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: move mgmt descriptor limit handle under mgmt_tx
Frames that are transmitted via MGMT_TX are using reserved descriptor
slots in firmware. This limitation is for the htt_mgmt_tx path itself,
not for mgmt frames per se. In 16 MBSSID scenario, these reserved slots
will be easy exhausted due to frequent probe responses. So for 10.4
based solutions, probe responses are limited by a threshold (24).
management tx path is separate for all except tlv based solutions. Since
tlv solutions (qca6174 & qca9377) do not support 16 AP interfaces, it is
safe to move management descriptor limitation check under mgmt_tx
function. Though CPU improvement is negligible, unlikely conditions or
never hit conditions in hot path can be avoided on data transmission.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Whenever firmware is configuring operating channel during scan or
home channel, channel change event will be indicated to host. In some
cases (device probe/ last vdev down), target will be configured to
default channel whereas host is unaware of target's operating channel.
This leads to packet drop due to unknown channel and kernel log will be
filled up with "no channel configured; ignoring frame(s)!". Fix that
by handling HTT_T2H_MSG_TYPE_CHAN_CHANGE event.
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Steve deRosier [Tue, 8 Mar 2016 00:58:50 +0000 (16:58 -0800)]
ath6kl: ignore WMI_TXE_NOTIFY_EVENTID based on fw capability flags
Certain 6004 firmware releases redefine the WMI_TXE_NOTIFY_EVENTID event
number and sends the new event frequently. However it doesn't have the
tx-err-notify feature and thus this firmware capability flag isn't set on
the firmware package. By guarding the processing of this event by the same
method we guard the sending of the WMI_SET_TXE_NOTIFY_CMDID command, we
can ignore the spurious event that we don't know how to process.
Without this change we call cfg80211_cqm_txe_notify() with possibly bad
data.
Signed-off-by: Steve deRosier <steve.derosier@lairdtech.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Mon, 7 Mar 2016 02:38:17 +0000 (10:38 +0800)]
ath9k: cleanup led_pin initial
Make ath_init_leds() and ath_deinit_leds() pairs as the only
API to set leds, also removed direction configuration from
ath9k_start() and ath9k_stop(). So the initial is more clear
now.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Mon, 7 Mar 2016 02:38:15 +0000 (10:38 +0800)]
ath9k: make GPIO API to support both of WMAC and SOC
commit 61b559dea40e ("ath9k: add extra GPIO led support")
added ath9k to support access SOC's GPIOs, but implemented
in a separated API: ath9k_hw_request_gpio().
So this patch make the APIs more common, to support both
of WMAC and SOC GPIOs. The new APIs as below,
Miaoqing Pan [Mon, 7 Mar 2016 02:38:14 +0000 (10:38 +0800)]
ath9k: define correct GPIO numbers and bits mask
Define correct GPIO numbers and MASK bits to indicate the WMAC
GPIO resource.
Allow SOC chips(AR9340, AR9531, AR9550, AR9561) to access all GPIOs
which rely on gpiolib framework. But restrict SOC AR9330 only to
access WMAC GPIO which has the same design with the old chips.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:42 +0000 (16:08 +0800)]
ath9k: use AR_SREV_9003_PCOEM to identify PCOEM chips
commit f49c90db4d23 ("ath9k: Add a macro to identify PCOEM chips")
defined AR_SREV_9003_PCOEM macro, its more clear to use the macro
instead of checking one by one. Also removed PCOEM chips checking
in the callback of ar9003_hw_do_pcoem_manual_peak_cal() which only
for PCOEM chips.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:40 +0000 (16:08 +0800)]
ath9k: Update AR9580 initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:39 +0000 (16:08 +0800)]
ath9k: Update QCA956x initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:38 +0000 (16:08 +0800)]
ath9k: Update AR9565 initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:37 +0000 (16:08 +0800)]
ath9k: Update AR955x initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:36 +0000 (16:08 +0800)]
ath9k: Update AR9485 initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:35 +0000 (16:08 +0800)]
ath9k: Update AR9462 initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:34 +0000 (16:08 +0800)]
ath9k: Update AR9340 initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:33 +0000 (16:08 +0800)]
ath9k: Update AR933x initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Fri, 26 Feb 2016 08:08:32 +0000 (16:08 +0800)]
ath9k: Update AR9003 2.2 initvals
HW peak detect calibration would fail for AR9300 chips and
we went for implementing the SW way of doing it instead of
HW doing the peak detect calibration.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Peter Oh [Tue, 1 Mar 2016 17:52:49 +0000 (09:52 -0800)]
ath10k: set MAC timestamp in management Rx frame
Check and set Rx MAC timestamp when firmware indicates it.
Firmware adds it in Rx beacon frame only at this moment.
Driver and mac80211 may utilize it to detect such clockdrift
or beacon collision and use the result for beacon collision
avoidance.
Signed-off-by: Peter Oh <poh@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: deliver mgmt frames from htt to monitor vifs only
Until now only WMI originating mgmt frames were
reported to mac80211. Management frames on HTT
were basically dropped (except frames which looked
like management but had FCS error).
To allow sniffing all frames (including offloaded
frames) without interfering with mac80211
operation and states a new rx_flag was introduced
and is not being used to distinguish frames and
classify them for mac80211.
Signed-off-by: Grzegorz Bajorski <grzegorz.bajorski@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:18 +0000 (19:18 +0200)]
wil6210: add oob_mode module parameter
Add module parameter oob_mode. Takes effect the next time
the interface is brought up and FW is loaded. Puts the FW
in special "out of the box" (OOB) mode which is used for
diagnostics and certification.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:17 +0000 (19:18 +0200)]
wil6210: pass is_go flag to firmware
When starting a PCP, pass the is_go flag to firmware in
wmi_pcp_start. This flag indicates whether we started
a PCP which is also a GO(P2P group owner) or just a regular
PCP.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:16 +0000 (19:18 +0200)]
wil6210: fix no_fw_recovery mode with change_virtual_intf
When FW crashed with no_fw_recovery mode enabled, user space
could still call wil_cfg80211_change_iface quickly to change
interface type, and this would cause recovery to proceed and
FW crash logs may be lost.
Fix this problem by not resetting the FW in case no_fw_recovery
is enabled.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:15 +0000 (19:18 +0200)]
wil6210: clean ioctl debug message
Fix a debug message related to IOCTL that was incorrectly logged
with the MISC category, and move it inside wil_ioctl so it will
always be logged even if we call wil_ioctl from other places.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:14 +0000 (19:18 +0200)]
wil6210: fix race conditions in p2p listen and search
Fix 2 race conditions found during test runs of P2P discovery:
1. Because wil_p2p_cancel_listen was not protected, user space
could start a new P2P listen/search before wmi_stop_discovery
completed. This caused a crash in the firmware.
2. In P2P listen, when listen timer expires and user space calls
cancel_remain_on_channel at the same time, code could send the
cfg80211_remain_on_channel_expired notification twice.
Added protections with wil->mutex to several places that call
wmi_stop_discovery.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:13 +0000 (19:18 +0200)]
wil6210: P2P_DEVICE virtual interface support
Added support for the P2P_DEVICE virtual interface. This interface
is intended for P2P management operations such as discovery and
GO negotiation. Normally it is implemented by drivers to allow
a separate interface for P2P management with its own MAC address,
but for 11ad drivers it is needed to support P2P search, since it
cannot otherwise be separated from normal scan.
Since we only support a single interface/MAC address, we can't
easily separate between primary and P2P_DEVICE interfaces.
For example when a management packet arrives we can't tell for
which interface it is intended. To work around this, we store
a pointer to the interface where the last "radio operation" was
triggered such as scan or remain on channel, and we forward
management packets and scan results to this interface.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:10 +0000 (19:18 +0200)]
wil6210: basic PBSS/PCP support
PBSS (Personal Basic Service Set) is a new BSS type for DMG
networks. It is similar to infrastructure BSS, having an AP-like
entity called PCP (PBSS Control Point), but it has few differences.
For example, stations inside a PBSS can communicate directly, and
the PCP role can be transferred between stations.
This change adds PBSS support, and has 2 main parts:
1. When starting an AP, add an option to start as a PCP instead.
This is implemented by a new PBSS flag which is passed as part of
the cfg80211_ap_settings structure.
2. When connecting to a BSS, add an option to connect to a PCP
instead of an AP. This is again implemented by a new PBSS flag,
added to the cfg80211_connect_params structure.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:09 +0000 (19:18 +0200)]
wil6210: switch to generated wmi.h
Switch to auto-generated version of wmi.h which is maintained
by FW team. This will allow better sync between teams in the
future and avoid bugs because of unexpected API changes.
The wmi.h will have many differences but most are cosmetic.
It also includes these real differences:
1. is_go parameter added to BCON_CTRL and START_PCP commands.
2. max_rx_pl_per_desc added to CFG_RX_CHAIN command.
3. various small API updates that are not currently used by
driver.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Tue, 1 Mar 2016 17:18:08 +0000 (19:18 +0200)]
wil6210: add support for discovery mode during scan
Add support for discovery mode during scan. When discovery mode
is active, station transmits special beacons while scanning.
This can optimize the scan mainly when there is only one AP/PCP
around.
Discovery mode is implicitly used by firmware during P2P search.
Since there is currently no use case where user space has a
reason to directly control discovery mode, we expose it only
through a debugfs flag.
Also fix name confusion in the wmi_scan_type enumeration.
The type previously called WMI_LONG_SCAN is actually
WMI_ACTIVE_SCAN.
Signed-off-by: Lior David <qca_liord@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Maya Erez [Tue, 1 Mar 2016 17:18:05 +0000 (19:18 +0200)]
wil6210: AP: prevent connecting to already connected station
wmi_evt_connect doesn't check if the connect event is received for
an already connected station.
This can lead to memory leak as a new vring is allocated without
freeing the previously allocated vring and to unexpected behavior
of nl80211 layer due to unexpected notification of a new station.
Add a check in wmi_evt_connect in AP mode to verify that the requested
CID is not associated to an already connected station.
Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Maya Erez [Tue, 1 Mar 2016 17:18:04 +0000 (19:18 +0200)]
wil6210: remove BACK RX and TX workers
WMI synchronous handling has changed and WMI calls that provide
a buffer for the reply are completed in the WMI interrupt context.
This allows sending the RX and TX BACK commands from the WMI event
handler without the need for the worker thread.
This is a better approach as it can decrease the handshake time
in the connect flow and prevent race conditions in case of fast
disconnects. An example for such a race is handling of wil_back_rx_handle
during a disconnect event, as wil_back_rx_handle is not protected by
the wil mutex and a disconnect can be handled after sta->status is
verified as connected.
Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
commit 166de3f1895d ("ath10k: remove supported chain mask") had revealed
an issue on monitor mode. Configuring NSS upon monitor interface
creation is causing target assert in all qca9888x and qca6174 firmware.
Firmware assert issue can be reproduced by below sequence even after
reverting commit 166de3f1895d ("ath10k: remove supported chain mask").
ip link set wlan0 down
iw wlan0 set type monitor
iw phy0 set antenna 7
ip link set wlan0 up
This issue is originally reported on qca9888 with 10.1 firmware.
Fixes: 5572a95b4b ("ath10k: apply chainmask settings to vdev on creation") Cc: stable@vger.kernel.org Reported-by: Janusz Dziedzic <janusz.dziedzic@tieto.com> Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
The number of HTT Tx descriptors and qcache peer
limit aren't hw-specific. In fact they are
firmware specific and should not be placed in
hw_params.
The QCA4019 limits were submitted with the peer
flow control firmware only and to my understanding
there's no non-peer-flow-ctrl QCA4019 firmware.
However QCA99X0 is planned to run firmware
supporting the feature as well. Therefore this
patch enables QCA99X0 to use 2500 tx descriptors
whenever possible instead of just 1424.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Tue, 1 Mar 2016 12:16:10 +0000 (13:16 +0100)]
ath10k: fix HTT Tx CE ring size
QCA4019 can queue up to 2500 frames at a time.
This means it requires roughly 5000 entires on the
ring to work properly. Otherwise random tx failure
may occur.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Sun, 6 Mar 2016 14:14:43 +0000 (16:14 +0200)]
ath10k: implement push-pull tx
The current/old tx path design was that host, at
its own leisure, pushed tx frames to the device.
For HTT there was ~1000-1400 msdu queue depth.
After reaching that limit the driver would request
mac80211 to stop queues. There was little control
over what packets got in there as far as
DA/RA was considered so it was rather easy to
starve per-station traffic flows.
With MU-MIMO this became a significant problem
because the queue depth was insufficient to buffer
frames from multiple clients (which could have
different signal quality and capabilities) in an
efficient fashion.
Hence the new tx path in 10.4 was introduced: a
pull-push mode.
Firmware and host can share tx queue state via
DMA. The state is logically a 2 dimensional array
addressed via peer_id+tid pair. Each entry is a
counter (either number of bytes or packets. Host
keeps it updated and firmware uses it for
scheduling Tx pull requests to host.
This allows MU-MIMO to become a lot more effective
with 10+ clients.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Sun, 6 Mar 2016 14:14:36 +0000 (16:14 +0200)]
ath10k: implement updating shared htt txq state
Firmware 10.4.3 onwards can support a pull-push Tx
model where it shares a Tx queue state with the
host.
The host updates the DMA region it pointed to
during HTT setup whenever number of software
queued from (on host) changes. Based on this
information firmware issues fetch requests to the
host telling the host how many frames from a list
of given stations/tids should be submitted to the
firmware.
The code won't be called because not all
appropriate HTT events are processed yet.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Sun, 6 Mar 2016 14:14:30 +0000 (16:14 +0200)]
ath10k: add fast peer_map lookup
The pull-push functionality of 10.4 will be based
on peer_id and tid. These will need to be mapped,
eventually, to ieee80211_txq to be used with
ieee80211_tx_dequeue().
Iterating over existing stations every time
peer_id needs to be mapped to a station would be
inefficient wrt CPU time.
The new firmware, which will be the only user of
the code flow-wise, will guarantee to use low
peer_ids first so despite peer_map's apparent huge
size d-cache thrashing should not be a problem.
Older firmware hot paths will effectively not use
peer_map.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Sun, 6 Mar 2016 14:14:27 +0000 (16:14 +0200)]
ath10k: maintain peer_id for each sta and vif
The 10.4.3 firmware with congestion control
guarantees that each peer has only a single
peer_id mapping.
The 1:1 mapping isn't the case for older firmwares
(e.g. 10.4.1, 10.2, 10.1) but it should not
matter. This 1:1 mapping is going to be only used
by future code which inherently (flow-wise) is for
10.4.3.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Sun, 6 Mar 2016 14:14:25 +0000 (16:14 +0200)]
ath10k: refactor tx pending management
Tx pending counter logic assumed that the sk_buff
is already known and hence was performed in HTT
functions themselves.
However, for the sake of future wake_tx_queue()
usage the driver must be able to tell whether it
can submit more frames to firmware before it
dequeues frame from ieee80211_txq (and thus long
before HTT Tx functions are called) because once a
frame is dequeued it cannot be requeud back to
mac80211.
This prepares the driver for future changes.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Zhang Shengju [Thu, 3 Mar 2016 01:16:57 +0000 (01:16 +0000)]
wireless: use reset to set mac header
Since offset is zero, it's not necessary to use set function. Reset
function is straightforward, and will remove the unnecessary add
operation in set function.
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Zhang Shengju [Thu, 3 Mar 2016 01:16:56 +0000 (01:16 +0000)]
mac80211: use reset to set header pointer
Since offset is zero, it's not necessary to use set function. Reset
function is straightforward, and will remove the unnecessary add
operation in set function.
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Zhang Shengju [Thu, 3 Mar 2016 01:16:55 +0000 (01:16 +0000)]
mac80211_hwsim: use reset to set mac header
Since offset is zero, it's not necessary to use set function. Reset
function is straightforward, and will remove the unnecessary add
operation in set function.
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Zhang Shengju [Thu, 3 Mar 2016 01:16:54 +0000 (01:16 +0000)]
vxlan: use reset to set header pointers
Since offset is zero, it's not necessary to use set function. Reset
function is straightforward, and will remove the unnecessary add operation
in set function.
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 4 Mar 2016 22:19:16 +0000 (17:19 -0500)]
Merge tag 'rxrpc-rewrite-20160304' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs
David Howells says:
====================
RxRPC: Rewrite part 1
Here's the first set of patches from my RxRPC rewrite, aimed at net-next.
These do some clean ups and bug fixes. Most of the changes are small, but
there are a couple of bigger changes:
(*) Convert call flag numbers and event numbers into enums. Then rename
the event numbers to all have _EV_ in their name to stop confusion.
Fix one instance of an event bit being used instead of a flag bit.
(*) A copy of the Rx protocol header is kept in the sk_buff private data.
Keep this in host byte order rather than network byte order as it
makes more sense. A number of other fields then get converted into
host byte order too.
Conversion between host and network byte order is then done at the
packet reception/generation stage.
This is based on net-next/master
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Fri, 4 Mar 2016 10:52:19 +0000 (11:52 +0100)]
ethtool.h: define INT_MAX for userland
INT_MAX needs limits.h in userland.
When ethtool.h is included by a userland app, we got the following error:
.../usr/include/linux/ethtool.h: In function 'ethtool_validate_speed':
.../usr/include/linux/ethtool.h:1471:18: error: 'INT_MAX' undeclared (first use in this function)
return speed <= INT_MAX || speed == SPEED_UNKNOWN
^
Fixes: e02564ee334a ("ethtool: make validate_speed accept all speeds between 0 and INT_MAX") CC: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Acked-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Fri, 4 Mar 2016 10:52:17 +0000 (11:52 +0100)]
cxgb4i: don't redefine DIV_ROUND_UP
let's use the common definition to avoid the following warning during the
compilation:
drivers/scsi/cxgbi/cxgb4i/cxgb4i.c:161:0: warning: "DIV_ROUND_UP" redefined
#define DIV_ROUND_UP(n, d) (((n) + (d) - 1) / (d))
^
In file included from include/linux/list.h:8:0,
from include/linux/module.h:9,
from drivers/scsi/cxgbi/cxgb4i/cxgb4i.c:16:
include/linux/kernel.h:67:0: note: this is the location of the previous definition
#define DIV_ROUND_UP __KERNEL_DIV_ROUND_UP
^
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Let's add a common definition in uapi and use it everywhere.
Fixes: ac2c7ad0e5d6 ("net/ethtool: introduce a new ioctl for per queue setting") CC: Kan Liang <kan.liang@intel.com> Suggested-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Simon Horman [Fri, 4 Mar 2016 06:58:07 +0000 (15:58 +0900)]
nfp: remove Rolf Neugebauer as co-maintainer
Rolf is no longer in his previous role at Netronome and as far as I know no
longer working on the NFP driver. Thus it does not seem appropriate for him
to be a co-maintainer anymore.
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com> Signed-off-by: Simon Horman <simon.horman@netronome.com> Signed-off-by: David S. Miller <davem@davemloft.net>