[PATCH 1/2] Adjust the SNR when comparing BSSes based on Tx power config

[Jouni Malinen]

On Fri, Sep 22, 2023 at 05:51:12PM +0000, Kaidong Wang wrote:
> The max transmit power of Standard Power (SP) Access Points (AP) in the
> 6 GHz Band is limited by effective isotropic radiated power (EIRP),
> while the max transmit power of Low Power Indoor (LPI) APs in the 6 GHz
> Band is limited by power spectral density (PSD). Therefore the max
> transmit power of LPI APs grows as the channel width increases.
> 
> Adjust the SNR of BSSes  when comparing BSSes based on Tx power config.

The commit message talks about 6 GHz band, but it was not obvious from
the changes whether this would have impact to other bands as well. If
this might change behavior for other bands, it would be good to note in
the commit message.

> diff --git a/src/drivers/driver.h b/src/drivers/driver.h
>   * @flags: information flags about the BSS/IBSS (WPA_SCAN_*)
>   * @bssid: BSSID
>   * @freq: frequency of the channel in MHz (e.g., 2412 = channel 1)
> + * @max_cw: the max channel width supported by both the AP and the STA
>   * @beacon_int: beacon interval in TUs (host byte order)
>   * @caps: capability information field in host byte order
>   * @qual: signal quality
> @@ -370,6 +371,7 @@ struct wpa_scan_res {
>  	unsigned int flags;
>  	u8 bssid[ETH_ALEN];
>  	int freq;
> +	enum chan_width max_cw;

This value is added here, but there is no driver_*.c changes in this
patchset to actually set that value. How is this expected to be filled
in? And how is his concept of "the AP and the STA" even relevant from
scan results? Scanning can be used to find BSSs when not associated and
it can also be used for IBSS and mesh BSS cases without any AP.

Or is this supposed to be some kind of internal-to-wpa_supplicant
variable that just happens to convenient live in this struct even though
it is not really going to be coming from the driver at all like
est_throughput? If so, it would be good to document that clearly (see
that est_throughput for an example) since this is quite confusing
without such explicit indication.

-- 
Jouni Malinen                                            PGP id EFC895FA