Raja Mani [Wed, 27 Jan 2016 09:54:28 +0000 (15:24 +0530)]
ath10k: add clock ctrl related functions in ahb
pre qca4019 chipsets has/uses internal clock generator for
the operation. But, qca4019 uses external clocks supplied from
outside of target (ie, outside of wifi core). Three different clocks
(cmd clock, ref clock, rtc clock) comes into picture in qca4019.
All those clocks needs to configured with help of global clock
controller (gcc) to make qca4019 functioning.
Add functions for clock init/deinit, clock enable/disable in ahb.
This is just a preparation, functions added in this patch will be
used in later patches.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 27 Jan 2016 09:54:27 +0000 (15:24 +0530)]
ath10k: add helper functions in ahb.c for reg rd/wr
qca4019 deals with below register memory region to control the clock,
reset, etc.
- Memory to control wifi core
- gcc (outside of wifi)
- tcsr (outside of wifi)
Add new helper functions to perform read/write in above registers
spaces. Actual ioremap for above registers are done in later patch.
Struct ath10k_ahb is introduced to maintain ahb specific info and
memory this struct will be allocated in the continuation of struct
ath10k_pci (again, memory ath10k_ahb is allocated in the later patch).
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 27 Jan 2016 09:54:25 +0000 (15:24 +0530)]
ath10k: add basic skeleton to support ahb
qca4019 uses ahb instead of pci where it slightly differs in device
enumeration, clock control, reset control, etc. Good thing is that
ahb also uses copy engine for the data transaction. So, the most of
the stuff implemented in pci.c/ce.c are reusable in ahb case too.
Device enumeration in ahb case comes through platform driver/device
model. All resource details like irq, memory map, clocks, etc for
qca4019 can be fetched from of_node of platform device.
Simply flow would look like,
device tree => platform device (kernel) => platform driver (ath10k)
Device tree entry will have all qca4019 resource details and the same
info will be passed to kernel. Kernel will prepare new platform device
for that entry and expose DT info to of_node in platform device.
Later, ath10k would register platform driver with unique compatible name
and then kernels binds to corresponding compatible entry & calls ath10k
ahb probe functions. From there onwards, ath10k will take control of it
and move forward.
New bool flag CONFIG_ATH10K_AHB is added in Kconfig to conditionally
enable ahb support in ath10k. On enabling this flag, ath10k_pci.ko
will have ahb support. This patch adds only basic skeleton and few
macros to support ahb in the context of qca4019.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 27 Jan 2016 09:54:24 +0000 (15:24 +0530)]
ath10k: pull reusable code from pci probe and remove for ahb
Some of the code present in ath10k_pci_{probe|remove} are reusable
in ahb case too. To avoid code duplication, move reusable code to
new functions. Later, those new functions can be called from ahb
module's probe and exit functions.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 27 Jan 2016 09:54:23 +0000 (15:24 +0530)]
ath10k: make ath10k_pci_read32/write32() ops more generic
ath10k_pci_read32/write32() does work more specific to
PCI by ensuring pci wake/sleep for every read and write.
There is a plan to use most of stuff available in pci.c
(irq stuff, copy engine, etc) for AHB case. Such kind
of pci wake/sleep for every read/write is not required
in AHB case (qca4019). All those reusable areas in pci.c
and ce.c calls ath10k_pci_read32/write32() for low level
read and write.
In fact, ath10k_pci_read32/write32() should do what it does
today for PCI case. But for AHB, it has to do differently.
To make ath10k_pci_read32/write32() more generic, new function
pointers are added in ar_pci for the function which does
operation more close to the bus. Later, corresponding bus
specific read and write function will be mapped to that.
ath10k_pci_read32/write32() are changed to call directly
those function pointers without worrying which bus underlying
to it. Also, the function to get number of bank is changed
in the same way.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 27 Jan 2016 09:54:22 +0000 (15:24 +0530)]
ath10k: make some of ath10k_pci_* func reusable
Some of static functions present in pci.c file are reusable
in ahb (qca4019) case. Remove static word for those reusable
functions and have those function prototype declaration in
pci.h file. So that, pci.h header file can be included in
ahb module and reused. There is no functionality changes done
in this patch.
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Thu, 21 Jan 2016 13:13:27 +0000 (14:13 +0100)]
ath10k: implement basic support for new tx path firmware
This allows to use the new firmware which
implements the new tx data path. Without this
patch firmware supporting new tx path stops
responding shortly after booting.
This patch doesn't implement the entire pull-push
logic available in the new firmware. This will be
done in subsequent patches.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Thu, 21 Jan 2016 13:13:25 +0000 (14:13 +0100)]
ath10k: add new FW_FEATURE_PEER_FLOW_CONTROL
This feature flag will be used for firmware which
supports pull-push model where host shares it's
software queue state with firmware and firmware
generates fetch requests telling host which queues
to dequeue tx from.
Primary function of this is improved MU-MIMO
performance with multiple clients.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Mon, 18 Jan 2016 01:33:50 +0000 (09:33 +0800)]
ath9k: fix data bus error on ar9300 and ar9580
One crash issue be found on ar9300: RTC_RC reg read leads crash, leading
the data bus error, due to RTC_RC reg write not happen properly.
Warm Reset trigger in continuous beacon stuck for one of the customer for
other chip, noticed the MAC was stuck in RTC reset. After analysis noticed
DMA did not complete when RTC was put in reset.
So, before resetting the MAC need to make sure there are no pending DMA
transactions because this reset does not reset all parts of the chip.
The 12th and 11th bit of MAC _DMA_CFG register used to do that.
12 cfg_halt_ack 0x0
0 DMA has not yet halted
1 DMA has halted
11 cfg_halt_req 0x0
0 DMA logic operates normally
1 Request DMA logic to stop so software can reset the MAC
The Bit [12] of this register indicates when the halt has taken effect or
not. the DMA halt IS NOT recoverable; once software sets bit [11] to
request a DMA halt, software must wait for bit [12] to be set and reset
the MAC.
So, the same thing we implemented for ar9580 chip.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Miaoqing Pan [Mon, 18 Jan 2016 01:33:47 +0000 (09:33 +0800)]
ath9k: clean up ANI per-channel pointer checking
commit c24bd3620c50 ("ath9k: Do not maintain ANI state per-channel")
removed per-channel handling, the code to check 'curchan' also
should be removed as never used.
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
In the current solution wil6210 configures the vring in a worker
and holds only one pending CID. This implementation may lead to
race conditions between connect and disconnect events of multiple
stations or fast connect/disconnect events of the same station.
In order to allow the removal of the connect worker and handling of
WMI_VRING_CFG_DONE_EVENTID in the connect event, the WMI replies
that provide the reply in a given buffer needs to be handled
immediately in the WMI event interrupt thread.
To prevent deadlocks, WMI replies that requires additional
handling are still handled via the events list.
Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Maya Erez [Sun, 17 Jan 2016 10:39:09 +0000 (12:39 +0200)]
wil6210: find TX vring only if vring_tx_data is enabled
In TX vring allocation, vring_tx->va is allocated before WMI command to
configure the vring is sent to the device. As the WMI command can take
time to complete, it can lead to scenarios where vring_tx->va is not NULL
but vring is still not enabled.
This patch adds a check that vring_tx_data is enabled before returning
a valid TX vring.
Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Maya Erez [Sun, 17 Jan 2016 10:39:08 +0000 (12:39 +0200)]
wil6210: handle tx completions only if vring tx data is enabled
wil_vring_fini_tx is called in disconnect in order to free the
allocated vrings.
wil_vring_fini_tx is disabling the vring_tx_data before napi_synchronize
is called in order to avoid the tx handling of this vring, while
wil_vring_free is called only after napi finished the current handling
of the tx completed packets.
Due to that, in case of disconnect, wil6210_netdev_poll_tx can be called
when vring->va is not NULL but vring_tx_data[i]->enabled is already
disabled.
This patch checks vring_tx_data[i]->enabled in wil6210_netdev_poll_tx
to prevent handling of disabled vrings.
Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Sun, 17 Jan 2016 10:39:07 +0000 (12:39 +0200)]
wil6210: use extra IEs from probe response
In the start_ap/change_beacon API, when we set up probe response
offloading, we only use the IE list from the probe response
template and not the IE list from the proberesp_ies argument.
As a result, we miss important IEs and it causes problems with
some scenarios such as P2P.
With this change, we merge the list of IEs from the probe response
template and proberesp_ies and send the merged list to the FW
for offloading. It is still FW responsibility to filter out
irrelevant IEs when sending probe response, based on the actual
contents of the probe request.
Also in case association response termplate is not provided,
we will use the merged list of IEs from probe response in the
association response as well.
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 [Sun, 17 Jan 2016 10:39:06 +0000 (12:39 +0200)]
wil6210: fix privacy flag calculation in change_beacon
Currently the privacy flag calculation in change_beacon
checks for RSN IE inside proberesp_ies, but normally it
is not found there. It works today because of code inside
the function wil_fix_bcon, but wil_fix_bcon is not directly
related to change_beacon, and it may be changed or removed
in the future and break the calculation.
To fix this issue, change the privacy flag calculation
to check RSN IE inside the beacon itself. The new check is more
reliable and will not be sensitive to changes in wil_fix_bcon.
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>
Alexander Tsoy [Thu, 7 Jan 2016 22:26:03 +0000 (01:26 +0300)]
ath9k_htc: add device ID for Toshiba WLM-20U2/GN-1080
This device is available under different marketing names:
WLM-20U2 - Wireless USB Dongle for Toshiba TVs
GN-1080 - Wireless LAN Module for Toshiba MFPs.
Signed-off-by: Alexander Tsoy <alexander@tsoy.me> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Register for 500ms as periodic peer stats update period,
and parameters like rx_duration that needs to be tracked
in host can be achieved by this provision, also periodic
stats update is the future of fw_stats and shall be extended
for pdev / vdev stats irrespecitive PEER_STATS service is enabled
or not
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com>
[kvalo@qca.qualcomm.com: fix a checkpatch warning] Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: provision to support periodic peer stats update
Enable support for periodic peer stats update when peer stats
service is enabled. The host to update the peer stats received
from the firmware periodically, since firmware will reset this to zero
after sometime (due to memory constraints)
While enabling periodic peer / vdev stats cleanup the existing
list in debugfs if max limit is reached, so that the new stats is
updated.
Ideally speaking based on 'Michal Kazior's' suggestion
we need to completely move to periodic update of all the stats making
the 'ping - pong' method obselete. This requires a bit of re-work and
some testing as well, also confirmation regarding backward comptability
for various firmware and chipsets. Hence allow periodic update only
for peer_stats.
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Replace the local variable 'burst_enable' with 'param'
for mapping and setting pdev paraemeters and with this patch
pretty easy to extend support for new parameters adhering to
linux kernel coding guidelines
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: fix naming Peer stats rssi_changed field in 10.2.4
Fix naming of peer stats rssi_changed field in 10.2.4 to make it
more readable. As of now this field is not used, but necessary to
pull in fw_stats with appropriate length.
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
ath10k: add support for parsing per STA rx_duration for 10.2.4
Add support for parsing 'peer_rx_duration' as part
of 10.2.4 peer_stats. Also register PEER_STATS service
for parsing 'rx_duration' (and for any new fields to be added
as part of peer_stats). Have checks for backward compatibility with
older 10.2.4 firmware incase PEER_STATS service is not enabled
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Michal Kazior [Wed, 13 Jan 2016 13:52:52 +0000 (14:52 +0100)]
ath10k: prevent txrx running for too long
On multicore systems it was possible for the txrx
tasklet to keep on running for long periods of
time on a single CPU due to tx completion
processing. Another CPU could feed the running
tasklet for an indefinite amount of time.
The tasklet is now guaranteed to run a finite
amount of time and is limited by HTT CE Rx ring
depth.
This improves behavior when RPS is used on target
system and might improve TCP handling as well.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Raja Mani [Wed, 13 Jan 2016 05:56:44 +0000 (11:26 +0530)]
dt: bindings: add bindings for ipq4019 wifi block
Add device tree binding documentation details for wifi block present
in Qualcomm IPQ4019 SoC into "qcom,ath10k.txt".
Right now, ath10k wireless driver has support for PCI based wlan devices.
There is a plan to extend ath10k driver to support wifi devices which are
connected over AHB as well (enumeration will happen via device tree node).
For AHB based devices (ie, ipq4019), all wifi properties are supplied via
device tree (including irq, reg addr, cal data,etc).
Signed-off-by: Raja Mani <rmani@qti.qualcomm.com> Acked-by: Rob Herring <robh@kernel.org>
[kvalo@qca.qualcomm.com: fixed typos noticed by Rob] Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Hante Meuleman [Tue, 5 Jan 2016 10:05:50 +0000 (11:05 +0100)]
brcmfmac: Do not handle link downs for ibss.
Sometimes on module reload and reconnect to ibss a deauth from
other station can be received. This is treated as a link down but
for ibss this is wrong. It will close the interface and no data
is possible. Ignore the firmware generated link down events in
ibss mode, as ibss is always teared down by cfg80211.
Reviewed-by: Arend Van Spriel <arend@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The btcoex uses a timeout which was in milliseconds and got
converted to jiffies upon using timer api. Instead, convert
it to jiffies and treat it as such further.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
brcmfmac: use msecs_to_jiffies() in macro definitions
Instead to having macro definition for millisecond timeout have
the definition directly in jiffies. This makes the unit of the
value immediately clear and may result in code that is bit more
compact.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
The function brcmf_sdio_wd_timer() has wdtick parameter. However, it
is only called with two values and as such the parameter is replaced
with boolean value indicating the timer should be active or not.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Kalle Valo [Thu, 7 Jan 2016 21:04:07 +0000 (23:04 +0200)]
Merge tag 'iwlwifi-next-for-kalle-2016-01-07_2' of https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/iwlwifi-next
* bug fixes and improvements for firmware debug system (Golan and myself)
* fixes for D0i3 (Eliad)
* prevent muliple stations with the same MAC address
* advertise support for Rx A-MSDU in A-MPDU
* scan related fixes
* support -20.ucode
* fix WoWLAN for iwldvm
* preparations towards multiple Rx queues
* platform power improvements for GO mode when no clients are associated
iwlwifi: pcie: properly configure the debug buffer size for 8000
8000 device family has a new debug engine that needs to be
configured differently than 7000's.
The debug engine's DMA works in chunks of memory and the
size of the buffer really means the start of the last
chunk. Since one chunk is 256-byte long, we should
configure the device to write to buffer_size - 256.
This fixes a situation were the device would write to
memory it is not allowed to access.
iwlwifi: mvm: fix memory leaks in error paths upon fw error dump
When iwl_mvm_fw_error_dump fails, it needs to clear the
state in mvm, which includes:
* clear IWL_MVM_STATUS_DUMPING_FW_LOG
* set mvm->fw_dump_trig to NULL
* free the description
While at it, remove a NULL check in
iwl_mvm_free_fw_dump_desc since kfree is NULL safe.
Johannes Berg [Tue, 5 Jan 2016 15:16:31 +0000 (16:16 +0100)]
iwlwifi: mvm: check minimum temperature notification length
This notification will be extended with extra data, so just
check that it has a minimum length, not the exact length;
we might later add handling for the extra fields added and
have more code to handle both versions.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Johannes Berg [Sun, 6 Dec 2015 12:58:08 +0000 (14:58 +0200)]
iwlwifi: mvm: check PN for CCMP/GCMP in the driver
As we're working on multi-queue RX, we want to parallelise checking
the PN in order to avoid having to serialise the RX processing.
It may seem that doing parallel PN checking is insecure, but it turns
out to be OK because queue assignment is done based on the data in the
frame (IP/TCP) and thus cannot be manipulated by an attacker, since
the data is encrypted and must first have been decrypted successfully.
There are some corner cases, in particular when the peer starts using
fragmentation which redirects the packet to the default queue. However
this redirection is remembered (for the STA, per TID) and thus cannot
be exploited by an attacker either.
Leave checking on the default queue (queue 0) to mac80211, since we
get fragmented packets there and those are subject to stricter checks
during reassembly.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Sara Sharon <sara.sharon@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Ayala Beker [Sun, 20 Dec 2015 07:27:50 +0000 (09:27 +0200)]
iwlwifi: mvm: don't ask beacons when P2P GO vif and no assoc sta
The commit below called iwl_mvm_mac_ctxt_changed() to handle
a case that the vif is a P2P GO.
However iwl_mvm_mac_ctxt_cmd_go() ignores the number of
associated stations and asks the FW to pass beacons anyways.
Fix this by checking ap_assoc_sta_count parameter, in iwl_mvm_mac_ctxt_cmd_go()
as well, and ask the FW to pass beacons only when there's
at least one associated station.
Johannes Berg [Thu, 24 Sep 2015 16:14:55 +0000 (18:14 +0200)]
iwlwifi: mvm: prevent multiple stations with the same address
As the device (and parts of the driver) cannot deal with having the
same MAC address for two stations (on two virtual interfaces), add
some explicit code to prevent this case.
Note that in practice this cannot happen since the device doesn't
support operating with two AP/GO interfaces at the same time either,
and other scenarios for this are, while not impossible, not going to
happen in practice.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Johannes Berg [Mon, 21 Sep 2015 12:09:17 +0000 (14:09 +0200)]
iwlwifi: mvm: support A-MSDU in A-MPDU
Since A-MPDU deaggregation is done in hardware, and A-MSDU deaggregation
is done in software, there's no reason not to support A-MSDU in A-MPDU;
set the flag to support it.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Golan Ben-Ami [Wed, 25 Nov 2015 09:44:57 +0000 (11:44 +0200)]
iwlwifi: mvm: add a non-trigger window to fw dbg triggers
Allow the user to configure a non-trigger session - a window
between triggers in which the driver won't collect fw debug data.
This can be useful when the frequent collection of fw data
has an impact on the performance, such as debugging
tx flows.
David Spinadel [Sun, 3 Jan 2016 15:08:32 +0000 (17:08 +0200)]
iwlwifi: mvm: fix extended dwell time
FW adds 10 msec for every dwell time in low band, so we need
to set 10 msec less.
Don't use extended dwell time when fragmented scan is needed
because FW adds 3 msec per probe and it can easily exceed
max out of channel time.
iwlwifi: mvm: reset mvm->scan_type when firmware is started
If we don't reset the scan type when the firmware is
started, we will think the firmware is still configured
after the interface has been brought down. When we will
bring it up again, we will not configure the scan type
in firmware and it will crash with the following assert:
0x0000100A | ADVANCED_SYSASSERT
Fixes: 355346ba3050 ("iwlwifi: mvm: configure scheduled scan according to traffic conditions") Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
iwlwifi: mvm: let the firmware choose the antenna for beacons
The firmware knows better what antenna to choose.
Old firmware still need the setting, so use a flag to know
if the driver should choose the antenna or if the firmware
can do it iself.
My commit below introduced a mutex in the transport to
prevent concurrent operations. To do so, it added a flag
(is_down) to make sure the transport is in the right state.
This uncoverred an bug that didn't cause any harm until
now: iwldvm calls stop_device and then starts the firmware
without calling start_hw in between. While this flow is
fine from the device configuration point of view (register,
etc...), it is now forbidden by the new is_down flag.
This led to this error to appear:
iwlwifi 0000:05:00.0: Can't start_fw since the HW hasn't been started
and the suspend would fail.
This fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=109591
Fengwei Yin [Sun, 20 Dec 2015 13:20:41 +0000 (21:20 +0800)]
wcn36xx: split DMA mask register writing.
Per comments from Bjorn Andersson <bjorn.andersson@sonymobile.com>,
split DMA mask register writing as seperate patch in case we need
bi-sect in the furture.
Signed-off-by: Fengwei Yin <fengwei.yin@linaro.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Fengwei Yin [Sun, 20 Dec 2015 13:20:40 +0000 (21:20 +0800)]
wcn36xx: handle rx skb allocation failure to avoid system crash
Lawrence reported that git clone could make system crash on a
Qualcomm ARM soc based device (DragonBoard, 1G memory without
swap) running 64bit Debian.
It's turned out the crash is related with rx skb allocation
failure. git could consume more than 600MB anonymous memory.
And system is in extremely memory shortage case.
But driver didn't handle the rx allocation failure case. This patch
doesn't submit skb to upper layer if rx skb allocation fails.
Instead, it reuse the old skb for rx DMA again. It's more like
drop the packets if system is in memory shortage case.
With this change, git clone is OOMed instead of system crash.
Reported-by: King, Lawrence <lking@qti.qualcomm.com> Signed-off-by: Fengwei Yin <fengwei.yin@linaro.org> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Dan Carpenter [Sat, 19 Dec 2015 10:59:19 +0000 (13:59 +0300)]
ath9k: fix ath9k_hw_nvram_check_version()
There is a type bug so it always returns success.
Fixes: 6fa658fd5ab2 ('ath9k: Simplify and fix eeprom endianness swapping') Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Pawel Kulakowski [Fri, 18 Dec 2015 09:48:57 +0000 (10:48 +0100)]
ath9k: Enable support for cloned SKBS
Ath9k driver does not modify tx skbs, so SUPPORTS_CLONED_SKBS
flag can be set. Enabling this flag significant reduce number
of copy operation during TCP Tx. This is especially noticeable
on platforms with slower CPU (lower CPU usage brings
profits in better TCP Tx troughput results).
Tested on MIPS with 560 MHz clock
Without CLONED_SKBS flag:
TCP Tx 145 Mb/s (iperf result)
__copy_user_common consumes 12.9% of CPU (result from perf tool)
0% CPU Idle
With CLONED_SKBS flag:
TCP Tx 170 Mb/s (iperf result)
__copy_user_common consumes 1.8% of CPU (result from perf tool)
12% CPU Idle
Signed-off-by: Pawel Kulakowski <pawel.kulakowski@tieto.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Lior David [Wed, 16 Dec 2015 15:51:46 +0000 (17:51 +0200)]
wil6210: support for platform specific crash recovery
Added a simple interface for platform to perform crash
recovery.
When firmware crashes, wil driver can notify the platform
which can trigger a crash recovery process. During
the process the platform can request a ram dump
from the wil driver as well as control when firmware
recovery will start. This interface allows the platform
to implement a more advanced crash recovery, for
example to reset dependent subsystems in proper order, or
to provide its own notifications during the recovery process.
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>
Hamad Kadmany [Wed, 16 Dec 2015 15:51:45 +0000 (17:51 +0200)]
wil6210: fix kernel OOPS when stopping interface during Rx traffic
When network interface is stopping, some resources may
be already released by the network stack, and Rx frames
cause kernel OOPS (observed one is in netfilter code)
Proper solution is to drop packets pending in reorder buffer.
Signed-off-by: Hamad Kadmany <qca_hkadmany@qca.qualcomm.com> Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Signed-off-by: Maya Erez <qca_merez@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Peter Oh [Thu, 31 Dec 2015 13:26:20 +0000 (15:26 +0200)]
ath10k: set SM power save disabled to default value
Use SMPS disabled as default because FW does not indicate
any support of SMPS.
This change will help STAs out that don’t support SMPS from
sticking on 1SS, since they don’t have method to change it
back to multiple chains.
This change also should not affect power consumption of STAs
supporting SMPS, because they are capable to switch the mode
to dynamic or static either at the end of frame sequence or
by using SMPS action frame.
Signed-off-by: Peter Oh <poh@qca.qualcomm.com> Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Jia-Ju Bai [Mon, 4 Jan 2016 07:55:38 +0000 (15:55 +0800)]
rt2x00pci: Disable memory-write-invalidate when the driver exits
The driver calls pci_set_mwi to enable memory-write-invalidate when it
is initialized, but does not call pci_clear_mwi when it is removed. Many
other drivers calls pci_clear_mwi when pci_set_mwi is called, such as
r8169, 8139cp and e1000.
This patch adds pci_clear_mwi in error handling and removal procedure,
which can fix the problem.
Signed-off-by: Jia-Ju Bai <baijiaju1990@163.com> Acked-by: Helmut Schaa <helmut.schaa@googlemail.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
brcmfmac: introduce module parameter to force successful probe
The module parameter can be used to ensure the probe succeeds thus
claiming the device and allowing post-mortem debugging in case of
firmware crash. It is only available when select CONFIG_BRCMDBG.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Hante Meuleman [Sat, 2 Jan 2016 08:41:41 +0000 (09:41 +0100)]
brcmfmac: Move all module parameters to one place
Module parameters are defined in several files. Move them in one
place and make them device specific or global. This makes it
easier to override device specific settings by external data like
platform data in the future.
Reviewed-by: Arend Van Spriel <arend@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Hante Meuleman [Sat, 2 Jan 2016 08:41:40 +0000 (09:41 +0100)]
brcmfmac: Fix warn trace on module unload while in ibss mode
When the driver is being unloaded a situation can occur where the
wirelesss core (cfg80211) wants to remove the ibss, but the state
of brcmfmac has already been set to down. When an error is
returned in that situation then that will result in a stack
trace on removal of the wiphy object. This is avoided by
returning 0 when device is down on a leave_ibss call.
Reviewed-by: Arend Van Spriel <arend@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
brcmfmac: obtain feature info using 'cap' firmware command
Several features in the driver directly map to a firmware feature
listed in response of the 'cap' firmware command. For those features
this response will be examined instead of attempting individual
firmware commands.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Reviewed-by: Mathy Vanhoef <vanhoefm@gmail.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Hante Meuleman [Sat, 2 Jan 2016 08:41:37 +0000 (09:41 +0100)]
brcmfmac: Add get_station support for IBSS
When get_station is requested for IBSS then an error will be
printed and no information will be returned. This patch adds
IBSS specific get station information function.
Reviewed-by: Arend Van Spriel <arend@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Amitkumar Karwar [Thu, 31 Dec 2015 14:24:15 +0000 (06:24 -0800)]
mwifiex: reduce cloned skb queue size
Driver supports Tx status ack feature. When hostapd/
wpa_supplicant asks for ack status of an EAPOL/ACTION
frame, driver maintains a cloned skb for the packet
until TX_STATUS event is received from firmware.
Cloned skb queue gets flushed when connection is terminated
or driver is unloaded.
Let's reduce the queue size to avoid unnecessarily
keeping memory allocated when environment is busy.
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Xinming Hu [Thu, 31 Dec 2015 14:24:12 +0000 (06:24 -0800)]
mwifiex: fix missing debug messages
Some critical messages are missed until "adapter->dev"
gets initialized in mwifiex_register_dev().
We will use pr_* print message instead of mwifiex_dbg at
those places to resolve the problem.
Signed-off-by: Xinming Hu <huxm@marvell.com> Signed-off-by: Amitkumar Karwar <akarwar@marvell.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Craig Gallek [Tue, 5 Jan 2016 15:57:13 +0000 (10:57 -0500)]
soreuseport: change consume_skb to kfree_skb in error case
Fixes: 538950a1b752 ("soreuseport: setsockopt SO_ATTACH_REUSEPORT_[CE]BPF") Suggested-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Craig Gallek <kraig@google.com> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Craig Gallek [Tue, 5 Jan 2016 20:08:07 +0000 (15:08 -0500)]
soreuseport: pass skb to secondary UDP socket lookup
This socket-lookup path did not pass along the skb in question
in my original BPF-based socket selection patch. The skb in the
udpN_lib_lookup2 path can be used for BPF-based socket selection just
like it is in the 'traditional' udpN_lib_lookup path.
udpN_lib_lookup2 kicks in when there are greater than 10 sockets in
the same hlist slot. Coincidentally, I chose 10 sockets per
reuseport group in my functional test, so the lookup2 path was not
excersised. This adds an additional set of tests with 20 sockets.
Fixes: 538950a1b752 ("soreuseport: setsockopt SO_ATTACH_REUSEPORT_[CE]BPF") Fixes: 3ca8e4029969 ("soreuseport: BPF selection functional test") Suggested-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Craig Gallek <kraig@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ido Schimmel [Tue, 5 Jan 2016 10:36:40 +0000 (11:36 +0100)]
mlxsw: pci: Adjust value of CPU egress traffic class
During initialization, when creating the send descriptor queues (SDQs),
we specify the CPU egress traffic class of each SDQ. The maximum number
of classes of this type is different in the two ASICs supported by this
PCI driver.
New firmware versions check this value is set correctly, which causes
errors on the Spectrum ASIC, as its max exposed egress traffic class is
lower than 7.
Solve this by setting this field to 3, which is an acceptable value for
both ASICs.
Note that we currently do not expose the QoS capabilities of the ASICs,
so setting this to an hardcoded value is OK for now.
Signed-off-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 6 Jan 2016 05:05:04 +0000 (00:05 -0500)]
Merge tag 'wireless-drivers-next-for-davem-2016-01-05' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next
Kalle Valo says:
====================
brcfmac
* fix IBSS which got broken over time
* new USB id for bcm43242 dongle
* arp offload configuration through inet notifier
ath9k
* add random number generator support (CONFIG_ATH9K_HWRNG)
iwlwifi
* Make scan parameters low latency aware
* Fix in the NL80211_FEATURE_FULL_AP_CLIENT_STATE state case
* Fix enable injection mode (Chaya Rachel)
* Various cleanups (Dan / Julia / myself)
* Allow to stay more time on popular channels (David Spinadel)
* Bug fixes for D0i3 (Eliad / Luca)
* Fixes for GO uAPSD (myself)
* Start of TSO support (myself)
* Rate control bug fixes (Eyal / Gregory)
* Start the work on 9000 devices (Johannes / Sara / Oren)
* Start the work on a new Tx queue allocation model (Liad)
* Debug infrastructure enhancements (Golan)
mwifiex
* add a debugfs file for chip reset
* advertise SMS4 cipher suite
* increase ap and station interface limit to 3
* enable MSI support on newer pcie devices (8897 onwards)
rtlwifi
* fix lots of module parameter usage
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Arnd Bergmann [Fri, 1 Jan 2016 22:27:57 +0000 (23:27 +0100)]
net: hns: avoid uninitialized variable warning:
gcc fails to see that the use of the 'last_offset' variable
in hns_nic_reuse_page() is used correctly and issues a bogus
warning:
drivers/net/ethernet/hisilicon/hns/hns_enet.c: In function 'hns_nic_reuse_page':
drivers/net/ethernet/hisilicon/hns/hns_enet.c:541:6: warning: 'last_offset' may be used uninitialized in this function [-Wmaybe-uninitialized]
This simplifies the function to make it more obvious what is
going on to both readers and compilers, which makes the warning
go away.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch series introduces the support for ConnectX-4 timestamping
and the PTP kernel interface.
Changes from V2:
net/mlx5_core: Introduce access function to read internal_timer
- Remove one line function
- Change function name
net/mlx5e: Add HW timestamping (TS) support:
- Data path performance optimization (caching tstamp struct in rq,sq)
- Change read/write_lock_irqsave to read/write_lock
- Move ioctl functions to en_clock file
- Changed overflow start algorithm according to comments from Richard
- Move timestamp init/cleanup to open/close ndos.
In details:
1st patch prevents the driver from modifying skb->data and SKB CB in
device xmit function.
2nd patch adds the needed low level helpers for:
- Fetching the hardware clock (hardware internal timer)
- Parsing CQEs timestamps
- Device frequency capability
3rd patch adds new en_clock.c file that handles all needed timestamping
operations:
- Internal clock structure initialization and other helper functions
- Added the needed ioctl for setting/getting the current timestamping
configuration.
- used this configuration in RX/TX data path to fill the SKB with
the timestamp.
Eran Ben Elisha [Tue, 29 Dec 2015 12:58:32 +0000 (14:58 +0200)]
net/mlx5e: Add PTP Hardware Clock (PHC) support
Add a PHC support to the mlx5_en driver. Use reader/writer spinlocks to
protect the timecounter since every packet received needs to call
timecounter_cycle2time() when timestamping is enabled. This can become
a performance bottleneck with RSS and multiple receive queues if normal
spinlocks are used.
The driver has been tested with both Documentation/ptp/testptp and the
linuxptp project (http://linuxptp.sourceforge.net/) on a Mellanox
ConnectX-4 card.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Cc: Richard Cochran <richardcochran@gmail.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eran Ben Elisha [Tue, 29 Dec 2015 12:58:31 +0000 (14:58 +0200)]
net/mlx5e: Add HW timestamping (TS) support
Add support for enable/disable HW timestamping for incoming and/or
outgoing packets. To enable/disable HW timestamping appropriate
ioctl should be used. Currently HWTSTAMP_FILTER_ALL/NONE and
HWTSAMP_TX_ON/OFF only are supported. Make all relevant changes in
RX/TX flows to consider TS request and plant HW timestamps into
relevant structures.
Add internal clock for converting hardware timestamp to nanoseconds. In
addition, add a service task to catch internal clock overflow, to make
sure timestamping is accurate.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eran Ben Elisha [Tue, 29 Dec 2015 12:58:30 +0000 (14:58 +0200)]
net/mlx5_core: Introduce access function to read internal timer
A preparation step which adds support for reading the hardware
internal timer and the hardware timestamping from the CQE.
In addition, advertize device_frequency_khz HCA capability.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Achiad Shochat [Tue, 29 Dec 2015 12:58:29 +0000 (14:58 +0200)]
net/mlx5e: Do not modify the TX SKB
If the SKB is cloned, or has an elevated users count, someone else
can be looking at it at the same time.
Signed-off-by: Achiad Shochat <achiad@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 5 Jan 2016 17:24:06 +0000 (12:24 -0500)]
Merge branch 'sctp-transport-rhashtable'
Xin Long says:
====================
sctp: use transport hashtable to replace association's with rhashtable
for telecom center, the usual case is that a server is connected by thousands
of clients. but if the server with only one enpoint(udp style) use the same
sport and dport to communicate with every clients, and every assoc in server
will be hashed in the same chain of global assoc hashtable due to currently we
choose dport and sport as the hash key.
when a packet is received, sctp_rcv try to find the assoc with sport and dport,
since that chain is too long to find it fast, it make the performance turn to
very low, some test data is as follow:
in server:
$./ss [start a udp style server there]
in client:
$./cc [start 2500 sockets to connect server with same port and different ip,
and use one of them to send data to server]
===== test on net-next
-- perf top
server:
55.73% [kernel] [k] sctp_assoc_is_match
6.80% [kernel] [k] sctp_assoc_lookup_paddr
4.81% [kernel] [k] sctp_v4_cmp_addr
3.12% [kernel] [k] _raw_spin_unlock_irqrestore
1.94% [kernel] [k] sctp_cmp_addr_exact
we need to change the way to calculate the hash key, to use lport +
rport + paddr as the hash key can avoid this issue.
besides, this patchset will use transport hashtable to replace
association hashtable to lookup with rhashtable api. get transport
first then get association by t->asoc. and also it will make tcp
style work better.
===== test with this patchset:
-- perf top
server:
15.98% [kernel] [k] _raw_spin_unlock_irqrestore
9.92% [kernel] [k] __pv_queued_spin_lock_slowpath
7.22% [kernel] [k] copy_user_generic_string
2.38% libpthread-2.17.so [.] __recvmsg_nocancel
1.88% [kernel] [k] sctp_recvmsg
Xin Long [Wed, 30 Dec 2015 15:50:50 +0000 (23:50 +0800)]
sctp: remove the local_bh_disable/enable in sctp_endpoint_lookup_assoc
sctp_endpoint_lookup_assoc is called in the protection of sock lock
there is no need to call local_bh_disable in this function. so remove
them.
Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / linux-beck.git / log