Bug 217652
Summary: | ath11k: QCN9074: low data throughput | ||
---|---|---|---|
Product: | Drivers | Reporter: | Tanguy S. (tanguy) |
Component: | network-wireless | Assignee: | drivers_network-wireless (drivers_network-wireless) |
Status: | NEW --- | ||
Severity: | normal | CC: | kvalo, michal.smulski |
Priority: | P3 | ||
Hardware: | All | ||
OS: | Linux | ||
Kernel Version: | 6.4 | Subsystem: | |
Regression: | Yes | Bisected commit-id: | |
Attachments: |
HTT_PHY_STATS_TLV
HTT_RX_PDEV_STATS_TLV HTT_TX_PDEV_STATS_TLV hostapd_config dmesg_sta kernel config |
Description
Tanguy S.
2023-07-10 10:31:15 UTC
Created attachment 304581 [details]
HTT_RX_PDEV_STATS_TLV
Created attachment 304582 [details]
HTT_TX_PDEV_STATS_TLV
Created attachment 304583 [details]
hostapd_config
Testing 3m apart as suggested, leads to slightly better results on the TX side: MCS index 6 instead of 4. However, as for RX previously, 6-7 are the maximum indices I could see. > iw wlp6s0 link Connected to 00:15:61:27:cd:dc (on wlp6s0) SSID: SpaceX freq: 5520 RX: 2167311077 bytes (7283470 packets) TX: 2984256290 bytes (7782284 packets) signal: -45 dBm rx bitrate: 2594.2 MBit/s 160MHz HE-MCS 6 HE-NSS 4 HE-GI 0 HE-DCM 0 tx bitrate: 2594.2 MBit/s 160MHz HE-MCS 6 HE-NSS 4 HE-GI 0 HE-DCM 0 bss flags: short-slot-time dtim period: 4 beacon int: 100 Which produces the following throuhgput from STA: > iperf3 -P2 -R -c 10.0.0.1 [SUM] 0.00-10.04 sec 1.39 GBytes 1.19 Gbits/sec 2 sender [SUM] 0.00-10.00 sec 1.38 GBytes 1.18 Gbits/sec receiver > iperf3 -P2 -c 10.0.0.1 [SUM] 0.00-10.00 sec 1.38 GBytes 1.19 Gbits/sec 0 sender [SUM] 0.00-10.04 sec 1.38 GBytes 1.18 Gbits/sec receiver Attached full .config file and dmesg output from STA. Created attachment 304607 [details]
dmesg_sta
Created attachment 304608 [details]
kernel config
What kind of throughput are you expecting to see? 2.5 Gb/s would be ideal. Is that realistic ? I relaunched hostapd and re-associated without moving the setup and I am back to Connected to 00:15:61:27:cd:dc (on wlp6s0) SSID: SpaceX freq: 5500 RX: 3617195032 bytes (2428274 packets) TX: 1830499898 bytes (1272822 packets) signal: -46 dBm rx bitrate: 1729.3 MBit/s 160MHz HE-MCS 4 HE-NSS 4 HE-GI 0 HE-DCM 0 tx bitrate: 1729.3 MBit/s 160MHz HE-MCS 4 HE-NSS 4 HE-GI 0 HE-DCM 0 bss flags: short-slot-time dtim period: 4 beacon int: 100 I have an (ARM based) embedded system with QCN9074 module running Linux 5.15 with backported driver (5.15.92) in client mode. AP is a router with QCN9074 module (running closed sourced driver). On embedded client: #iw dev wlan0 link Connected to 90:09:d0:31:02:77 (on wlan0) SSID: xxxxx freq: 5745 RX: 1777648714 bytes (1475595 packets) TX: 3011532761 bytes (4820061 packets) signal: -95 dBm rx bitrate: 1921.5 MBit/s 80MHz HE-MCS 9 HE-NSS 4 HE-GI 0 HE-DCM 0 tx bitrate: 1729.6 MBit/s 80MHz HE-MCS 8 HE-NSS 4 HE-GI 0 HE-DCM 0 bss flags: short-slot-time dtim period: 4 beacon int: 100 The client and router (AP) are close to each other. 'Signal: -95 dBm' is wrong. It looks like a software bug. iperf3 is limited to the wired connection between router and a laptop (iperf3 -s) so my throughput is around 930Mbit in both directions. |