[PATCH] ath10k: Add sta rx packet stats per tid
Vasanthakumar Thiagarajan
vthiagar at qti.qualcomm.com
Fri Mar 2 00:21:25 PST 2018
On Wednesday 28 February 2018 06:47 PM, Kalle Valo wrote:
> Vasanthakumar Thiagarajan <vthiagar at codeaurora.org> writes:
>
>> Added per tid sta counters for the following
>>
>> - Total number MSDUs received from firmware
>> - Number of MSDUs received with errors like decryption, crc, mic ,etc.
>> - Number of MSDUs dropped in the driver
>> - A-MPDU/A-MSDU subframe stats
>> - Number of MSDUS passed to mac80211
>>
>> All stats other than A-MPDU stats are only for received data frames.
>> A-MPDU stats might have stats for management frames when monitor
>> interface is active where management frames are notified both in wmi
>> and HTT interfaces.
>>
>> These per tid stats can be enabled with tid bitmask through a debugfs
>> like below
>>
>> echo <tid_bitmask> > /sys/kernel/debug/ieee80211/phyX/ath10k/sta_tid_stats_mask
>>
>> tid 16 (tid_bitmask 0x10000) is used for non-qos data/management frames
>>
>> The stats are read from
>> /sys/kernel/debug/ieee80211/phyX/netdev\:wlanX/stations/<sta_mac>/dump_tid_stats
>>
>> Sample output:
>>
>> To enable rx stats for tid 0, 5 and 6,
>>
>> echo 0x00000061 > /sys/kernel/debug/ieee80211/phy0/ath10k/sta_tid_stats_mask
>>
>> cat /sys/kernel/debug/ieee80211/phy0/netdev\:wlan15/stations/8c\:fd\:f0\:0a\:8e\:df/dump_tid_stats
>>
>> Driver Rx pkt stats per tid, ([tid] count)
>> ------------------------------------------
>> MSDUs from FW [00] 2567 [05] 3178 [06] 1089
>> MSDUs unchained [00] 0 [05] 0 [06] 0
>> MSDUs locally dropped:chained [00] 0 [05] 0 [06] 0
>> MSDUs locally dropped:filtered [00] 0 [05] 0 [06] 0
>> MSDUs queued for mac80211 [00] 2567 [05] 3178 [06] 1089
>> MSDUs with error:fcs_err [00] 0 [05] 0 [06] 2
>> MSDUs with error:tkip_err [00] 0 [05] 0 [06] 0
>> MSDUs with error:crypt_err [00] 0 [05] 0 [06] 0
>> MSDUs with error:peer_idx_inval [00] 0 [05] 0 [06] 0
>>
>> A-MPDU num subframes upto 10 [00] 2567 [05] 3178 [06] 1087
>> A-MPDU num subframes 11-20 [00] 0 [05] 0 [06] 0
>> A-MPDU num subframes 21-30 [00] 0 [05] 0 [06] 0
>> A-MPDU num subframes 31-40 [00] 0 [05] 0 [06] 0
>> A-MPDU num subframes 41-50 [00] 0 [05] 0 [06] 0
>> A-MPDU num subframes 51-60 [00] 0 [05] 0 [06] 0
>> A-MPDU num subframes >60 [00] 0 [05] 0 [06] 0
>>
>> A-MSDU num subframes 1 [00] 2567 [05] 3178 [06] 1089
>> A-MSDU num subframes 2 [00] 0 [05] 0 [06] 0
>> A-MSDU num subframes 3 [00] 0 [05] 0 [06] 0
>> A-MSDU num subframes 4 [00] 0 [05] 0 [06] 0
>> A-MSDU num subframes >4 [00] 0 [05] 0 [06] 0
>>
>> Signed-off-by: Vasanthakumar Thiagarajan <vthiagar at codeaurora.org>
>
> This added new ath10k-check warnings, I fixed those in the pending
> branch:
>
> drivers/net/wireless/ath/ath10k/debug.h:203: Alignment should match open parenthesis
> drivers/net/wireless/ath/ath10k/debug.h:226: Alignment should match open parenthesis
> drivers/net/wireless/ath/ath10k/debugfs_sta.c:24: Alignment should match open parenthesis
> drivers/net/wireless/ath/ath10k/debugfs_sta.c:40: Alignment should match open parenthesis
> drivers/net/wireless/ath/ath10k/debugfs_sta.c:60: Alignment should match open parenthesis
>
> Also I fixed the typo "rages" in debug.h. My changes:
Thanks for fixing it. In most of the cases, fixing above warnings would result in lines > 80 chars.
May be we need not be very strict about > 80 char limit?
Vasanth
More information about the ath10k
mailing list