Currently ath11k will wait 11d scan complete while add interface in
ath11k_mac_op_add_interface(), when system resume without enable
wowlan, ath11k_mac_op_add_interface() is called for each resume, thus
it increase the resume time of system. And ath11k_mac_op_hw_scan()
after ath11k_mac_op_add_interface() also needs some time cost because
the previous 11d scan need more than 5 seconds when 6 GHz is enabled,
then the scan started event will indicated to ath11k after the 11d
scan completed.
While 11d scan/hw scan is running in firmware, if ath11k update channel
list to firmware by WMI_SCAN_CHAN_LIST_CMDID, then firmware will cancel
the current scan which is running, it lead the scan failed. The patch
commit 9dcf6808b2 ("ath11k: add 11d scan offload support") used
finish_11d_scan/finish_11d_ch_list/pending_11d to synchronize the 11d
scan/hw scan/channel list between ath11k/firmware/mac80211 and to avoid
the scan fail.
Add wait operation before ath11k update channel list, function
ath11k_reg_update_chan_list() will wait until the current 11d scan/hw
scan completed. And remove the wait operation of start 11d scan and
waiting channel list complete in hw scan. After these changes, resume
time cost reduce about 5 seconds and also hw scan time cost reduced
obviously, and scan failed not seen.
The 11d scan is sent to firmware only one time for each interface added
in mac.c, and it is moved after the 1st hw scan because 11d scan will
cost some time and thus leads the AP scan result update to UI delay.
Currently priority of ath11k's hw scan is WMI_SCAN_PRIORITY_LOW, and
priority of 11d scan in firmware is WMI_SCAN_PRIORITY_MEDIUM, then the
11d scan which sent after hw scan will cancel the hw scan in firmware,
so change the priority to WMI_SCAN_PRIORITY_MEDIUM for the hw scan which
is in front of the 11d scan, thus it will not happen scan cancel in
firmware.
Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3
Fixes: 9dcf6808b2 ("ath11k: add 11d scan offload support")
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220328035832.14122-1-quic_wgong@quicinc.com
Currently, the driver reports a tx_rate of 6.0 MBit/s no matter the true
rate:
root@linaro-developer:~# iw wlan0 link
Connected to 6c:f3:7f:eb:9b:92 (on wlan0)
SSID: SQ-DEVICETEST
freq: 5200
RX: 4141 bytes (32 packets)
TX: 2082 bytes (15 packets)
signal: -77 dBm
rx bitrate: 135.0 MBit/s MCS 6 40MHz short GI
tx bitrate: 6.0 MBit/s
bss flags: short-slot-time
dtim period: 1
beacon int: 100
This patch requests HAL_GLOBAL_CLASS_A_STATS_INFO via a hal_get_stats
firmware message and reports it via ieee80211_ops::sta_statistics.
root@linaro-developer:~# iw wlan0 link
Connected to 6c:f3:7f:eb:73:b2 (on wlan0)
SSID: SQ-DEVICETEST
freq: 5700
RX: 26788094 bytes (19859 packets)
TX: 1101376 bytes (12119 packets)
signal: -75 dBm
rx bitrate: 135.0 MBit/s MCS 6 40MHz short GI
tx bitrate: 108.0 MBit/s VHT-MCS 5 40MHz VHT-NSS 1
bss flags: short-slot-time
dtim period: 1
beacon int: 100
Tested on MSM8939 with WCN3680B running firmware CNSS-PR-2-0-1-2-c1-00083,
and verified by sniffing frames over the air with Wireshark to ensure the
MCS indices match.
Signed-off-by: Edmond Gagnon <egagnon@squareup.com>
Reviewed-by: Benjamin Li <benl@squareup.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220325224212.159690-1-egagnon@squareup.com
Currently ath11k does not send the country code to firmware after device
recovery, as a result the regdomain info is reported from firmware by
default. Regdomain info is important, so ath11k also need to restore
it to the value which was used before recovery.
Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220324011856.11014-1-quic_wgong@quicinc.com
This makes the code clearer since set11nRateFlags currently sets
both the rate flags and chain sel. This may also be required for
QCN550x support, where the rate flags and chain sel are in separate
fields.
This change does not appear to affect the final binary.
Signed-off-by: Wenli Looi <wlooi@ucalgary.ca>
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220320233010.123106-3-wlooi@ucalgary.ca
When more clients (128) are connected, the UL data traffic
KPI measurement is low compared to single client. This issue
is due to more CPU cycles spent on the peer lookup operation
with more clients. So reduce the peer lookup operation by
modifying the linear based lookup operation into the rhash
based lookup operation. This improve the peak throughput
measurement. Since this is a software algorithm change, it is
applicable for all the platforms.
TCP UL 128 Clients test case Observation (64bit system):
Previous: ~550 Mbps
Now : ~860 Mbps
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.5.0.1-01067-QCAHKSWPL_SILICONZ-1
Signed-off-by: Karthikeyan Periyasamy <quic_periyasa@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/1644036628-5334-1-git-send-email-quic_periyasa@quicinc.com
In Qualcomm ARM platforms there is WL_EN pin and other power regulators
which can be controlled at platform side to completely reset the chip.
For most of x86 and other platforms, the chip is connected via PCIe M.2
interface, and there is no way to control WL_EN pin. Instead the host
driver needs to reset the chip via PCIE_SOC_GLOBAL_RESET hardware
register, just like ath11k does currently.
But when using PCIE_SOC_GLOBAL_RESET there are some hardware registers
which are not cleared/restored. To handle those cases we can enable
PLATFORM_CAP_PCIE_GLOBAL_RESET QMI host capability to tell the firmware
to do some platform specific operations after firmware download.
This does not fix any known issues, but is recommended by the firmware
team, so enable the capability on QCA6390 and WCN6855 PCI devices. It is
currently unclear if this should be enabled also on QCN9074, so leave it
disabled for now. On AHB devices this is not needed as they don't use
PCIE_SOC_GLOBAL_RESET.
Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-01720.1-QCAHSPSWPL_V1_V2_SILICONZ_LITE-1
Signed-off-by: Baochen Qiang <bqiang@codeaurora.org>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20211011055602.77342-1-bqiang@codeaurora.org
Add hw-restart to directly restart wlan. Like UTF mode start it will
restart hardware and download firmware again.
Usage:
1. Run command:
echo hw-restart > /sys/kernel/debug/ath11k/qca6390\ hw2.0/simulate_fw_crash
echo hw-restart > /sys/kernel/debug/ath11k/wcn6855\ hw2.0/simulate_fw_crash
2. wlan will be restart and do recovery process and success.
Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-03003-QCAHSPSWPL_V1_V2_SILICONZ_LITE-2
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220228064606.8981-4-quic_wgong@quicinc.com
ieee80211_reconfig() of mac80211 is the main function for recovery of
each ieee80211_hw and ath11k, and ath11k_core_reconfigure_on_crash()
is the main function for recovery of ath11k_base, it has more than
one ieee80211_hw and ath11k for each ath11k_base, so it need to add
synchronization between them, otherwise it has many issue.
For example, when ath11k_core_reconfigure_on_crash() is not complete,
mac80211 send a hw scan request to ath11k, it leads firmware crash,
because firmware has not been initialized at that moment, firmware
is only finished downloaded and loaded, it can not receive scan
command.
Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-03003-QCAHSPSWPL_V1_V2_SILICONZ_LITE-2
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220228064606.8981-3-quic_wgong@quicinc.com
Currently ath11k has device recovery logic, it is introduced by this
patch "ath11k: Add support for subsystem recovery" which is upstream
by https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git/commit/?h=ath11k-bringup&id=3a7b4838b6f6f234239f263ef3dc02e612a083ad.
The patch is for AHB devices such as IPQ8074, it has remote proc module
which is used to download the firmware and boots the processor which
firmware is running on. If firmware crashed, remote proc module will
detect it and download and boot firmware again. Below command will
trigger a firmware crash, and then user can test feature of device
recovery.
Test command:
echo assert > /sys/kernel/debug/ath11k/qca6390\ hw2.0/simulate_fw_crash
echo assert > /sys/kernel/debug/ath11k/wcn6855\ hw2.0/simulate_fw_crash
Unfortunately, QCA6390 is PCIe bus, it does not have the remote proc
module, it use mhi module to communicate between firmware and ath11k.
So ath11k does not support device recovery for QCA6390 currently.
This patch is to add the extra logic which is different for QCA6390.
When firmware crashed, MHI_CB_EE_RDDM event will be indicate by
firmware and then ath11k_mhi_op_status_cb which is the callback of
mhi_controller will receive the MHI_CB_EE_RDDM event, then ath11k
will start to do recovery process, ath11k_core_reset() calls
ath11k_hif_power_down()/ath11k_hif_power_up(), then the mhi/ath11k
will start to download and boot firmware. There are some logic to
avoid deadloop recovery and two simultaneous recovery operations.
And because it has muti-radios for the soc, so it add some logic
in ath11k_mac_op_reconfig_complete() to make sure all radios has
reconfig complete and then complete the device recovery.
Also it add workqueue_aux, because ab->workqueue is used when receive
ATH11K_QMI_EVENT_FW_READY in recovery process(queue_work(ab->workqueue,
&ab->restart_work)), and ath11k_core_reset will wait for max
ATH11K_RESET_TIMEOUT_HZ for the previous restart_work finished, if
ath11k_core_reset also queued in ab->workqueue, then it will delay
restart_work of previous recovery and lead previous recovery fail.
ath11k recovery success for QCA6390/WCN6855 after apply this patch.
Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1
Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-03003-QCAHSPSWPL_V1_V2_SILICONZ_LITE-2
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220228064606.8981-2-quic_wgong@quicinc.com
ath11k_peer_find_by_addr states via lockdep that ab->base_lock must be
held when calling that function in order to protect the list. All
callers except ath11k_mac_op_unassign_vif_chanctx have that lock
acquired when calling ath11k_peer_find_by_addr. That lock is also not
transitively held by a path towards ath11k_mac_op_unassign_vif_chanctx.
The solution is to acquire the lock when calling
ath11k_peer_find_by_addr inside ath11k_mac_op_unassign_vif_chanctx.
I am currently working on a static analyser to detect missing locks and
this was a reported case. I manually verified the report by looking at
the code, but I do not have real hardware so this is compile tested
only.
Fixes: 701e48a43e ("ath11k: add packet log support for QCA6390")
Signed-off-by: Niels Dossche <dossche.niels@gmail.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220314215253.92658-1-dossche.niels@gmail.com
Currently fw build id is printed during init as follows.
fw_version 0x250684a5 fw_build_timestamp 2021-07-13 10:57
fw_build_id QC_IMAGE_VERSION_STRING=WLAN.HK.2.5.0.1-01100-QCAHKSWPL_SILICONZ-1
The string "QC_IMAGE_VERSION_STRING=" is removed from the log
to improve readability.
With this patch the fw build id is printed during init as follows.
fw_version 0x250684a5 fw_build_timestamp 2021-07-13 10:57
fw_build_id WLAN.HK.2.5.0.1-01100-QCAHKSWPL_SILICONZ-1
Signed-off-by: Hari Chandrakanthan <quic_haric@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/1646825065-7736-1-git-send-email-quic_haric@quicinc.com
Currently after the hardware restart triggered from the driver,
the station interface connection remains intact, since a disconnect
trigger is not sent to userspace. This can lead to a problem in
targets where the wifi mac sequence is added by the firmware.
After the target restart, its wifi mac sequence number gets
reset to zero. Hence AP to which our device is connected will receive
frames with a wifi mac sequence number jump to the past, thereby
resulting in the AP dropping all these frames, until the frame
arrives with a wifi mac sequence number which AP was expecting.
To avoid such frame drops, its better to trigger a station disconnect
upon target hardware restart which can be done with API
ieee80211_reconfig_disconnect exposed to mac80211.
The other targets are not affected by this change, since the hardware
params flag is not set.
Tested-on: WCN3990 hw1.0 SNOC WLAN.HL.3.1-01040-QCAHLSWMTPLZ-1
Tested-on: QCA6174 hw3.2 PCI WLAN.RM.4.4.1-00110-QCARMSWP-1
Tested-on: QCA6174 hw3.2 SDIO WLAN.RMH.4.4.1-00048
Signed-off-by: Youghandhar Chintala <youghand@codeaurora.org>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220315082944.12406-3-youghand@codeaurora.org
It is not recommened to use platform_get_resource(pdev, IORESOURCE_IRQ)
for requesting IRQ's resources any more, as they can be not ready yet in
case of DT-booting.
platform_get_irq() instead is a recommended way for getting IRQ even if
it was not retrieved earlier.
It also makes code simpler because we're getting "int" value right away
and no conversion from resource to int is required.
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Minghao Chi <chi.minghao@zte.com.cn>
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220314064501.2114002-1-chi.minghao@zte.com.cn
Implement net-detect feature by setting flag
WIPHY_WOWLAN_NET_DETECT if firmware supports this
feature. Driver sets the related PNO configuration
to firmware before entering WoW and firmware then
scans periodically and wakes up host if a specific
SSID is found.
Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1
Signed-off-by: Carl Huang <quic_cjhuang@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/1644308006-22784-3-git-send-email-quic_cjhuang@quicinc.com
linux-can-next-for-5.18-20220313
Marc Kleine-Budde says:
====================
pull-request: can-next 2022-03-13
this is a pull request of 13 patches for net-next/master.
The 1st patch is by me and fixes the freeing of a skb in the vxcan
driver (initially added in this net-next window).
The remaining 12 patches are also by me and target the mcp251xfd
driver. The first patch fixes a printf modifier (initially added in
this net-next window). The remaining patches add ethtool based ring
and RX/TX IRQ coalescing support to the driver.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds RX IRQ coalescing support to the driver.
The mcp251xfd chip doesn't support proper hardware based coalescing,
so this patch tries to implemented it in software. The RX-FIFO offers
a "FIFO not empty" interrupt, which is used if no coalescing is
active.
With activated RX IRQ coalescing the "FIFO not empty" interrupt is
disabled in the RX IRQ handler and the "FIFO half full" or "FIFO full
interrupt" (depending on RX max coalesced frames IRQ) is used instead.
To avoid RX CAN frame starvation a hrtimer is setup with RX coalesce
usecs IRQ,on timer expiration the "FIFO not empty" is enabled again.
Support for ethtool configuration is added in the next patch.
Link: https://lore.kernel.org/20220313083640.501791-9-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
So far the configuration of the hardware FIFOs is hard coded and
depend only on the selected CAN mode (CAN-2.0 or CAN-FD).
This patch updates the macros describing the ring, FIFO and RAM layout
to prepare for the next patches that add support for runtime
configurable ring parameters via ethtool.
Link: https://lore.kernel.org/20220313083640.501791-7-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Tony Nguyen says:
====================
ice: GTP support in switchdev
Marcin Szycik says:
Add support for adding GTP-C and GTP-U filters in switchdev mode.
To create a filter for GTP, create a GTP-type netdev with ip tool, enable
hardware offload, add qdisc and add a filter in tc:
ip link add $GTP0 type gtp role <sgsn/ggsn> hsize <hsize>
ethtool -K $PF0 hw-tc-offload on
tc qdisc add dev $GTP0 ingress
tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \
action mirred egress redirect dev $VF1_PR
By default, a filter for GTP-U will be added. To add a filter for GTP-C,
specify enc_dst_port = 2123, e.g.:
tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \
enc_dst_port 2123 action mirred egress redirect dev $VF1_PR
Note: outer IPv6 offload is not supported yet.
Note: GTP-U with no payload offload is not supported yet.
ICE COMMS package is required to create a filter as it contains GTP
profiles.
Changes in iproute2 [1] are required to be able to add GTP netdev and use
GTP-specific options (QFI and PDU type).
[1] https://lore.kernel.org/netdev/20220211182902.11542-1-wojciech.drewek@intel.com/T
---
v2: Add more CC
v3: Fix mail thread, sorry for spam
v4: Add GTP echo response in gtp module
v5: Change patch order
v6: Add GTP echo request in gtp module
v7: Fix kernel-docs in ice
v8: Remove handling of GTP Echo Response
v9: Add sending of multicast message on GTP Echo Response, fix GTP-C dummy
packet selection
v10: Rebase, fixed most 80 char line limits
v11: Rebase, collect Harald's Reviewed-by on patch 3
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
In case external PHY is used, we need to take care of embedded PHY.
Since there are no methods to disable this PHY from the MAC side and
keeping RMII reference clock, we need to suspend it.
This patch will reduce electrical noise (PHY is continuing to send FLPs)
and power consumption by 0,22W.
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
