[PATCH] wifi: ath12k: assign device id from the WSI index
Jeff Johnson
jeff.johnson at oss.qualcomm.com
Mon Sep 28 11:07:44 PDT 2026
On 9/28/2026 8:43 AM, Benoit Masson wrote:
> Devices of a hardware group are added to ag->ab[] in PCI probe
> completion order (ab->device_id = ag->num_probed++). The wiphy radio
> array is built from ag->ab[] and the wiphy device is taken from its
> first entry, so the wiphy path and the radio index to band mapping
> change on every boot.
you don't describe why this is is a problem, and hence why it should be
considered for backporting to stable kernels.
>
> Store each ab in ag->ab[] at its wsi_info.index, which is derived from
> the device tree WSI ring. The group order is then deterministic.
> ag->num_probed remains the group device counter. The unassign path is
> unchanged, it already indexes ag->ab[] by ab->device_id.
>
> Verified on TP-Link Archer BE800 (three single-band radios in one
> group): the wiphy path and the radio index to band mapping are
> identical across cold boots.
>
> Fixes: 6f245ea0ec6c ("wifi: ath12k: introduce device group abstraction")
> Cc: stable at vger.kernel.org # 6.14+
> Signed-off-by: Benoit Masson <yahoo at perenite.com>
SOB doesn't match your From address. So you either need to change your SOB,
update your From address in your e-mail agent, or add a From: tag to you patch
so that From == SOB
> ---
> Changes in v2:
> - Set ab->device_id from wsi_info.index in ath12k_core_hw_group_assign()
> instead of reordering pdev_map[] in ath12k_mac_allocate(), as suggested
> by Rameshkumar. mac.c is no longer modified.
your e-mail subject doesn't reflect that this is a v2, and it was threaded
instead of being sent as a separate e-mail thread.
when you send a v3 that addresses the above please make sure to use a properly
versioned subject and send as a separate thread.
Thanks
/jeff
More information about the ath12k
mailing list