Chris Chiu <chiu@xxxxxxxxxxxx> writes:
> On Wed, Oct 16, 2019 at 8:33 PM <yhchuang@xxxxxxxxxxx> wrote:
>>
>> From: Tzu-En Huang <tehuang@xxxxxxxxxxx>
>>
>
>> +
>> + band = hal->current_band_type;
>> + if (band == RTW_BAND_2G) {
>> + band = NL80211_BAND_2GHZ;
>> + cfg_mask = mask->control[band].legacy;
>> + } else if (band == RTW_BAND_5G) {
>> + band = NL80211_BAND_5GHZ;
>> + cfg_mask = mask->control[band].legacy << 4;
>> + }
>> +
>> + if (!is_vht_enable) {
>> + if (ra_mask & RA_MASK_HT_RATES_1SS)
>> + cfg_mask |= mask->control[band].ht_mcs[0] << 12;
>> + if (ra_mask & RA_MASK_HT_RATES_2SS)
>> + cfg_mask |= mask->control[band].ht_mcs[1] << 20;
>> + } else {
>> + if (ra_mask & RA_MASK_VHT_RATES_1SS)
>> + cfg_mask |= mask->control[band].vht_mcs[0] << 12;
>> + if (ra_mask & RA_MASK_VHT_RATES_2SS)
>> + cfg_mask |= mask->control[band].vht_mcs[1] << 22;
>> + }
>> +
>> + ra_mask &= cfg_mask;
>> +
>> + return ra_mask;
>> +}
>> +
>
> I believe you can replace the 4, 12, 20, 22 with a more descriptive macro.
Good point. GENMASK() and FIELD_PREP() are my favourites. Or maybe
u64_encode_bits() is actually better than FIELD_PREP(), as cfg_mask is
u64?
--
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
