WCN7850 on Surface Pro 11: no board-2.bin entry, firmware reports no board ID

[Jeff Johnson]

On 9/18/2026 5:13 PM, Dale Whinham wrote:
> Hi,
> 
> The WCN7850 in the Microsoft Surface Pro 11 (X1E80100) finds no entry in
> ath12k/WCN7850/hw2.0/board-2.bin and falls through to the board.bin
> API-1 path. I am looking for some advice on what the correct fix is.
> 
> The device reports:
> 
>   chip_id 0x2 chip_family 0x4 board_id 0xff soc_id 0x40170200
>   PCI 17cb:1107, subsystem 17cb:1107
> 
> A debug print in ath12k_qmi_request_target_cap() shows board_info is
> absent from the capability response:
> 
>   cap_resp board_info_valid 0 chip_info_valid 1 soc_info_valid 1
> 
> chip_info and soc_info are present, so the response itself is fine. The
> 0xff is ATH12K_BOARD_ID_DEFAULT, substituted by the driver.
> 
> The names ath12k then looks up are:
> 
>   bus=pci,vendor=17cb,device=1107,subsystem-vendor=17cb,subsystem-device=1107,qmi-chip-id=2,qmi-board-id=255
>   bus=pci,qmi-chip-id=2,qmi-board-id=255
> 
> Neither exists in board-2.bin (linux-firmware-atheros 20260810-2).
> The only subsystem-device=1107 entries there are qmi-board-id=44 and
> qmi-board-id=82.
> 
> Using a different entry from the same board-2.bin as board.bin does work
> (i.e. extract using ath12k-bdencoder and rename). I have been running
> the subsystem-device=3378, qmi-board-id=255 entry, which associates and
> passes traffic normally. That may be another board's data rather than a fix,
> but it does suggest the firmware accepts entries from this
> linux-firmware release, and that what is missing is an entry under a
> name this device derives.
> 
> I also checked what Windows provides:
> 
>   - Windows loads bdwlan.elf on this machine, via the catch-all
>     PCI\VEN_17CB&DEV_1107 section of qcwlanhmt8380.inf. The same file
>     appears in that driver for several other 8380 devices, so it is not
>     Surface-specific.
> 
>   - That file fails against the upstream firmware with -110 ("qmi failed
>     to load bdf file"). It works if the Windows firmware and PHY ucode
>     are used with it, so the data is valid for the part but looks tied
>     to the firmware version:
> 
>     Windows:  WLAN.HMT.1.1.c5-00310-QCAHMTSWPL_V1.0_V2.0_SILICONZ-1
>     upstream: WLAN.HMT.1.1.c7-00108-QCAHMTSWPL_V1.0_V2.0_SILICONZ_UPSTREAM-3
> 
> What is the right way to give this device board data? Happy to provide
> further info or test patches.
> 
> Thanks,
> Dale

Can you create a bug report per:
https://wireless.docs.kernel.org/en/latest/en/users/drivers/ath12k/bugreport.html

Then I can assign that.

/jeff