Received-SPF: pass (google.com: domain of linux-wireless-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <25c05c43-fae5-8d25-8f8a-73e6cb0c36a7@quicinc.com>
Date:   Mon, 16 May 2022 15:35:39 -0700
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Thunderbird/91.9.0
Subject: Re: [RFC][PATCH] mac80211: fix VHT 160Mhz bandwidth when using
 extended NSS on PHY with NSS ratio
Content-Language: en-US
To:     <mbizon@freebox.fr>, Johannes Berg <johannes@sipsolutions.net>
CC:     linux-wireless <linux-wireless@vger.kernel.org>
References: <19cbe5dd1d7f0e194ed40c8db2cf2beffa3ae167.camel@freebox.fr>
From:   Jeff Johnson <quic_jjohnson@quicinc.com>
In-Reply-To: <19cbe5dd1d7f0e194ed40c8db2cf2beffa3ae167.camel@freebox.fr>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 5/16/2022 1:39 AM, Maxime Bizon wrote:
> 
> A PHY supporting 160Mhz with a reduced number of NSS may not have any

note correct abbreviation is: MHz
and it is suggested to use a space to separate the value from the units

> of the IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_MASK bits set.
> 
> For exemple, the QCA9984 is 4x4, but can only do 2x2 VHT160 or

s/exemple/example/

> 80+80. The corresponding VHT capabilities should be:
>   - IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_MASK == 0
>   - Extended NSS BW == 2
> 
> If we use that hardware in AP mode, and a 160Mhz 2x2 STA without NSS
> restriction connects to it, then that STA will set
> IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_MASK to 1.
> 
> The usual logic of using bitwise-and with our own capabilities does
> not work here, the current code clears the SUPP_CHAN_WIDTH bits and
> resolve STA bandwidth to 80Mhz instead of 160Mhz, fix this.
> 
> Signed-off-by: Maxime Bizon <mbizon@freebox.fr>
> ---
>   net/mac80211/vht.c | 6 ++++--
>   1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/net/mac80211/vht.c b/net/mac80211/vht.c
> index e856f9092137..7ef93adc4f24 100644
> --- a/net/mac80211/vht.c
> +++ b/net/mac80211/vht.c
> @@ -230,9 +230,11 @@ ieee80211_vht_cap_ie_to_sta_vht_cap(struct ieee80211_sub_if_data *sdata,
>   	       sizeof(struct ieee80211_vht_mcs_info));
>   
>   	/* copy EXT_NSS_BW Support value or remove the capability */
> -	if (ieee80211_hw_check(&sdata->local->hw, SUPPORTS_VHT_EXT_NSS_BW))
> +	if (ieee80211_hw_check(&sdata->local->hw, SUPPORTS_VHT_EXT_NSS_BW)) {
> +		vht_cap->cap |= cap_info &
> +			IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_MASK;
>   		vht_cap->cap |= (cap_info & IEEE80211_VHT_CAP_EXT_NSS_BW_MASK);
> -	else
> +	} else

per the coding style when you use braces on any leg of if/else, use them 
on all legs

>   		vht_cap->vht_mcs.tx_highest &=
>   			~cpu_to_le16(IEEE80211_VHT_EXT_NSS_BW_CAPABLE);
>