TL-WN725N does not come "up", weird issue with management frame data

[jhwgh1968]

This issue is may be related to either #423 or #404, but hopefully I can facilitate progress by bringing debug logs.

I'm on Arch Linux with kernel 5.19, by the way.

I first tried the current master, compiled myself. The problem is that neither ip link nor wpa_supplicant can actually raise the interface.

Attached are logs with debug=9 on the driver. The part of the log that sticks out to me is this:

[  517.335393] -88eu_drv - dev_open
[  517.335394] R8188EU: -88eu_drv - drv_open, bup =1
[  517.569721] R8188EU:  [0x01000000,5]
[  517.569741] usb_read_port_complete!!!
[  517.569768] R8188EU:  [0x00000004,8]
[  517.569772] recvbuf2recvframe: rxdesc=offsset 0:0x84040042, 4:0x80f00000, 8:0x000103af, C:0x00000000
[  517.569803] R8188EU:  [0x00000004,8]
[  517.569807] +validate_recv_mgnt_frame
[  517.569810] R8188EU:  [0x00000010,8]
[  517.569814] +mgt_dispatcher: type(0x0) subtype(0x40)
[  517.569819] R8188EU:  [0x00000004,8]
[  517.569821] recv_func: validate_recv_frame fail! drop pkt
[  517.569835] R8188EU:  [0x00000004,8]
[  517.569838] rtw_recv_entry: recv_func return fail!!!
[  517.569840] R8188EU:  [0x00000004,5]
[  517.569842] recvbuf2recvframe: rtw_recv_entry(precvframe) != _SUCCESS
[  518.336474] R8188EU:  [0x80000000,8]
[  518.336480] Blinktimes (0): turn on
[  519.349972] R8188EU:  [0x80000000,8]
[  519.349981] Blinktimes (0): turn off
[  519.376181] R8188EU: ==>rtw_ps_processor .fw_state(0)
[  519.376187] R8188EU: ==>ips_enter cnts:1
[  519.376189] R8188EU: INFO nolinked power save enter
[  519.376191] R8188EU: ===> rtw_ips_pwr_down...................
[  519.376720] R8188EU:  [0x80000000,8]
[  519.376725] Led 2
[  519.376727] R8188EU:  [0x80000000,8]
[  519.376730] LedStrategy:1, LedAction 7
[  519.376732] R8188EU: ====> rtw_ips_dev_unload...
[  519.400849] R8188EU:  [0x00100000,5]
[  519.400855] +usb_intf_stop

It appears to fail to decode some critical management frame, and turns the device off again.

---

I also tried the latest 5.2 driver, and it didn't work either. To my surprise, however there was no module parameter to enable debug in the same way.

Trying to manually set config options like CONFIG_RTW_DEBUG during compliation resulted in compiler errors related to macros and missing variables with debug in the name, so it appears the 5.2 current branch has broken debug.

---

My questions are:

1. Is this master issue something that could be be investigated?
2. Is there a way to enable debug debug mode on the 5.2 branch?

Thanks.
log_master.txt.gz