[PATCH ath-next v2] wifi: ath12k: support average ack rssi in station dump

[Jeff Johnson]

On 6/12/2025 5:10 AM, Nicolas Escande wrote:
> On Thu Jun 12, 2025 at 1:46 PM CEST, Sowmiya Sree Elavalagan wrote:
>> From: Sivashankari Madhavan <quic_sivamadh at quicinc.com>
>>
>> Currently, the ACK RSSI value is not shown in station dump. Enable WMI
>> resource flag for ACK RSSI in WMI INIT command to add ACK RSSI value in
>> management TX completion event from WMI. Update ACK RSSI value obtained
>> in management and data frame completion path to ieee80211_tx_info. Also
>> advertise NL80211_EXT_FEATURE_ACK_SIGNAL_SUPPORT flag during hardware
>> register to mac80211 layer so that ACK RSSI is added to station dump
>> message.
>>
>> Example output :
>> Station 00:03:7f:01:5c:4b (on wlp88s0)
>>         inactive time:  46584 ms
>>         rx bytes:       955
>>         rx packets:     10
>>         tx bytes:       769
>>         tx packets:     6
>>         tx retries:     81
>>         tx failed:      0
>>         rx drop misc:   0
>>         signal:         -39 dBm
>>         signal avg:     -40 dBm
>>         tx bitrate:     6.0 MBit/s
>>         tx duration:    1185 us
>>         rx bitrate:     309.7 MBit/s 40MHz HE-MCS 6 HE-NSS 2 HE-GI 0 HE-DCM 0
>>         rx duration:    0 us
>>         last ack signal:-41 dBm
>>         avg ack signal: -40 dBm
>>         authorized:     yes
>>         authenticated:  yes
>> 	.......
>>
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.4.1-00199-QCAHKSWPL_SILICONZ-1
>> Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.0.c5-00481-QCAHMTSWPL_V1.0_V2.0_SILICONZ-3
>>
>> Signed-off-by: Sivashankari Madhavan <quic_sivamadh at quicinc.com>
>> Signed-off-by: Sowmiya Sree Elavalagan <quic_ssreeela at quicinc.com>
> [...]
>> @@ -5651,6 +5655,8 @@ static int ath12k_pull_mgmt_tx_compl_param_tlv(struct ath12k_base *ab,
>>  	param->pdev_id = ev->pdev_id;
>>  	param->desc_id = ev->desc_id;
>>  	param->status = ev->status;
>> +	param->ppdu_id = ev->ppdu_id;
> Hello,
> It seems this ppdu_id is retrieved from frimware but not used in this patchset
> Is this by design ?

I'm guessing it is for completeness (parse all members even if they aren't
used). Do you think this is a problem?

/jeff