On 11/27/2023 10:05 AM, Nicolas Escande wrote:
> The is a layout mismatch between the bitfield representing scan_flags in
> struct scan_req_params & the bits as defined in the WMI_SCAN_XXX macros.
> Lets fix it by making the struct match the #defines.
>
> I tried to correct it by making the struct match the #define and it
> worked for WMI_SCAN_FLAG_FORCE_ACTIVE_ON_DFS / scan_f_force_active_dfs_chn
> so I'm assuming this is the right thing to do.
>
> Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.7.0.1-01744-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: Nicolas Escande <nico.escande@xxxxxxxxx>
> ---
> drivers/net/wireless/ath/ath11k/wmi.h | 10 +++++-----
> 1 file changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath11k/wmi.h b/drivers/net/wireless/ath/ath11k/wmi.h
> index 100bb816b592..0b4e6c2f7860 100644
> --- a/drivers/net/wireless/ath/ath11k/wmi.h
> +++ b/drivers/net/wireless/ath/ath11k/wmi.h
> @@ -3348,17 +3348,17 @@ struct scan_req_params {
> scan_f_filter_prb_req:1,
> scan_f_bypass_dfs_chn:1,
> scan_f_continue_on_err:1,
> + scan_f_promisc_mode:1,
> + scan_f_force_active_dfs_chn:1,
> + scan_f_add_tpc_ie_in_probe:1,
> + scan_f_add_ds_ie_in_probe:1,
> + scan_f_add_spoofed_mac_in_probe:1,
> scan_f_offchan_mgmt_tx:1,
> scan_f_offchan_data_tx:1,
> - scan_f_promisc_mode:1,
> scan_f_capture_phy_err:1,
> scan_f_strict_passive_pch:1,
> scan_f_half_rate:1,
> scan_f_quarter_rate:1,
> - scan_f_force_active_dfs_chn:1,
> - scan_f_add_tpc_ie_in_probe:1,
> - scan_f_add_ds_ie_in_probe:1,
> - scan_f_add_spoofed_mac_in_probe:1,
> scan_f_add_rand_seq_in_probe:1,
> scan_f_en_ie_whitelist_in_probe:1,
> scan_f_forced:1,
You are convoluting two different data structures.
struct scan_req_params is used to represent a scan request within the
host driver. This does not use the WMI_SCAN_XXX macros.
struct wmi_start_scan_cmd is used to represent the scan request command
sent to firmware. This struct uses the WMI_SCAN_XXX macros to fill some
members of this struct in ath11k_wmi_copy_scan_event_cntrl_flags().
So your change has no effect on the driver operation and incorrectly
tries to foist the firmware definition upon the host internal
representation.
So NAK to this patch.
/jeff
