| Message ID | 20250107021017.3857555-1-quic_periyasa@quicinc.com |
|---|---|
| Headers | show |
| Series | wifi: ath12k: Add monitor interface support on QCN9274 | expand |
On Tue Jan 7, 2025 at 3:10 AM CET, Karthikeyan Periyasamy wrote: > Currently, monitor interface not supported. Therefore, add the missed TLV > tags parsing in the monitor parser, configure the monitor vdev state > identifier with HTT filter setup. > > Depends-On: > [PATCH v3 0/3] wifi: ath12k: Refactor monitor Rx handler > https://lore.kernel.org/all/20241224143613.164921-1-quic_periyasa@quicinc.com/ > > [PATCH v2 00/14] wifi: ath12k: Enable monitor ring for updating station dump in QCN9274 > https://lore.kernel.org/all/20241223060132.3506372-1-quic_ppranees@quicinc.com/ > > [PATCH v2 00/10] wifi: ath12k: Add peer extended Rx statistics debugfs support > https://lore.kernel.org/all/20241224161442.263729-1-quic_periyasa@quicinc.com/ > > Hari Chandrakanthan (1): > wifi: ath12k: fix link valid field initialization in the monitor Rx > > Karthikeyan Periyasamy (1): > wifi: ath12k: Replace band define G with GHZ where appropriate > > P Praneesh (5): > wifi: ath12k: Add extra TLV tag parsing support in monitor Rx path > wifi: ath12k: Avoid fetch Error bitmap and decap format from Rx TLV > wifi: ath12k: change the status update in the monitor Rx > wifi: ath12k: Avoid packet offset and FCS length from Rx TLV > wifi: ath12k: add monitor interface support on QCN9274 > [...] Hello, There might be something wrong with this series. I applied it on top of ath-pending-202501150104 + the "ath12k: Add peer extended Rx statistics debugfs support" and now 'iw dev wlan0 station dump' shows me that the RSSI value seems to not be updated anymore after I stop a monitor interface on the same phy. Steps to reproduce: iw phy phy0 interface add mon0 type monitor ip link set mon0 up # then just for 1 sec but it might not be usefull at all tcpdump -nepi mon0 ip link set mon0 down watch -n1 iw dev wlan0 station dump # 'signal' seems to not be updating anymore, maybe some other fields too ? Don't know if this is from this series or something else broken before, but as we could not add a monitor interface before, I could not see it. There still is a possibility that I messed something up while applying it so I would greatly appreciate if you could check it. This is on split-phy QCN9274 hw2.0 PCI WLAN.WBE.1.3.1-00173-QCAHKSWPL_SILICONZ-1 Thanks
On 1/15/2025 9:55 AM, Nicolas Escande wrote:
> This is on split-phy QCN9274 hw2.0 PCI WLAN.WBE.1.3.1-00173-QCAHKSWPL_SILICONZ-1
I don't know if it will help, but I just pushed out
new firmware WLAN.WBE.1.4.1-00199-QCAHKSWPL_SILICONZ-1:
https://git.codelinaro.org/clo/ath-firmware/ath12k-firmware/-/commit/a289ab4de56345be86d2c89c81aa12577c2c5a30
new board file:
https://git.codelinaro.org/clo/ath-firmware/ath12k-firmware/-/commit/bd2ca865d78f3d7d899e06cef57e5bbb9da60bf7
On Wed Jan 15, 2025 at 8:56 PM CET, Jeff Johnson wrote: > On 1/15/2025 9:55 AM, Nicolas Escande wrote: > > This is on split-phy QCN9274 hw2.0 PCI WLAN.WBE.1.3.1-00173-QCAHKSWPL_SILICONZ-1 > > I don't know if it will help, but I just pushed out > > new firmware WLAN.WBE.1.4.1-00199-QCAHKSWPL_SILICONZ-1: > https://git.codelinaro.org/clo/ath-firmware/ath12k-firmware/-/commit/a289ab4de56345be86d2c89c81aa12577c2c5a30 > > new board file: > https://git.codelinaro.org/clo/ath-firmware/ath12k-firmware/-/commit/bd2ca865d78f3d7d899e06cef57e5bbb9da60bf7 Thanks for the tip, I just tried it but alas it doesn't change a thing. RSSI (and maybe rx/tx bitrates not so sure) is still broken for all connected STA after a monitor device goes up & down.
On 1/15/2025 11:25 PM, Nicolas Escande wrote: > On Tue Jan 7, 2025 at 3:10 AM CET, Karthikeyan Periyasamy wrote: >> Currently, monitor interface not supported. Therefore, add the missed TLV >> tags parsing in the monitor parser, configure the monitor vdev state >> identifier with HTT filter setup. >> >> Depends-On: >> [PATCH v3 0/3] wifi: ath12k: Refactor monitor Rx handler >> https://lore.kernel.org/all/20241224143613.164921-1-quic_periyasa@quicinc.com/ >> >> [PATCH v2 00/14] wifi: ath12k: Enable monitor ring for updating station dump in QCN9274 >> https://lore.kernel.org/all/20241223060132.3506372-1-quic_ppranees@quicinc.com/ >> >> [PATCH v2 00/10] wifi: ath12k: Add peer extended Rx statistics debugfs support >> https://lore.kernel.org/all/20241224161442.263729-1-quic_periyasa@quicinc.com/ >> >> Hari Chandrakanthan (1): >> wifi: ath12k: fix link valid field initialization in the monitor Rx >> >> Karthikeyan Periyasamy (1): >> wifi: ath12k: Replace band define G with GHZ where appropriate >> >> P Praneesh (5): >> wifi: ath12k: Add extra TLV tag parsing support in monitor Rx path >> wifi: ath12k: Avoid fetch Error bitmap and decap format from Rx TLV >> wifi: ath12k: change the status update in the monitor Rx >> wifi: ath12k: Avoid packet offset and FCS length from Rx TLV >> wifi: ath12k: add monitor interface support on QCN9274 >> > [...] > > Hello, > > There might be something wrong with this series. > > I applied it on top of ath-pending-202501150104 + the "ath12k: Add peer extended > Rx statistics debugfs support" and now 'iw dev wlan0 station dump' shows me > that the RSSI value seems to not be updated anymore after I stop a monitor > interface on the same phy. > > Steps to reproduce: > iw phy phy0 interface add mon0 type monitor > ip link set mon0 up > # then just for 1 sec but it might not be usefull at all > tcpdump -nepi mon0 > ip link set mon0 down > watch -n1 iw dev wlan0 station dump > # 'signal' seems to not be updating anymore, maybe some other fields too ? > Not able to find the wlan0 interface details from the above steps ?
On Fri Jan 17, 2025 at 5:37 AM CET, Karthikeyan Periyasamy wrote: > On 1/15/2025 11:25 PM, Nicolas Escande wrote: > > On Tue Jan 7, 2025 at 3:10 AM CET, Karthikeyan Periyasamy wrote: [...] > > I applied it on top of ath-pending-202501150104 + the "ath12k: Add peer extended > > Rx statistics debugfs support" and now 'iw dev wlan0 station dump' shows me > > that the RSSI value seems to not be updated anymore after I stop a monitor > > interface on the same phy. > > > > Steps to reproduce: > > iw phy phy0 interface add mon0 type monitor > > ip link set mon0 up > > # then just for 1 sec but it might not be usefull at all > > tcpdump -nepi mon0 > > ip link set mon0 down > > watch -n1 iw dev wlan0 station dump > > # 'signal' seems to not be updating anymore, maybe some other fields too ? > > > > Not able to find the wlan0 interface details from the above steps ? Arf maybe I wasn't clear, my bad. I have a running AP on wlan0. I constantly monitor connected stations infos using 'iw dev wlan0 station dump' on the side in another terminal, and then I add a monitor interface with the above steps, and thats when I see that the stations connected to my AP have frozen values in station dump. I suspect that ath12k_dp_tx_htt_rx_monitor_mode_ring_config() should reset the rx filters to ath12k_mac_mon_status_filter_default when disabling monitor. But I'm not so sure, you guys know better.