Re: [PATCH wireless-next v4 3/3] wifi: mac80211_hwsim: add incumbent signal interference detection support

[Amith A <amith.a@oss.qualcomm.com>]

On 2/5/2026 5:54 PM, Johannes Berg wrote:
> On Thu, 2026-02-05 at 16:51 +0530, Amith A wrote:
>> +static void hwsim_6ghz_chanctx_iter(struct ieee80211_hw *hw,
>> +				    struct ieee80211_chanctx_conf *conf,
>> +				    void *data)
>> +{
>> +	struct ieee80211_chanctx_conf **out = data;
>> +
>> +	if (conf->def.chan && conf->def.chan->band == NL80211_BAND_6GHZ)
>> +		*out = conf;
>> +}
>> +
>> +static int hwsim_write_simulate_incumbent_signal(void *dat, u64 val)
>> +{
>> +	struct mac80211_hwsim_data *data = dat;
>> +	struct ieee80211_chanctx_conf *chanctx_conf = NULL;
>> +
>> +	ieee80211_iter_chan_contexts_atomic(data->hw,
>> +					    hwsim_6ghz_chanctx_iter,
>> +					    &chanctx_conf);
>> +
>> +	if (!chanctx_conf)
>> +		return -EINVAL;
> Wouldn't it make more sense require userspace to write also the
> frequency, rather than randomly picking the first 6 GHz chandef?
Agree. I’m currently using DEFINE_DEBUGFS_ATTRIBUTE, which only supports
a single u64 write. To accept both frequency and bitmap, is it okay if
define a custom fops with a .write that parses two arguments?
>
>> +++ b/drivers/net/wireless/virtual/mac80211_hwsim.h
>> @@ -343,4 +343,6 @@ enum hwsim_rate_info_attributes {
>>  	HWSIM_RATE_INFO_ATTR_MAX = NUM_HWSIM_RATE_INFO_ATTRS - 1
>>  };
>>  
>> +/* no longer used - previously used for debugfs iterator */
> Hmm?
Will remove this in next version.
>
> johannes