Received-SPF: pass (google.com: domain of linux-wireless-owner@vger.kernel.org designates 2620:137:e000::3:2 as permitted sender) client-ip=2620:137:e000::3:2;
Message-ID: <2d3603c7-2377-4c0f-a46f-378214e41403@candelatech.com>
Date:   Thu, 26 Oct 2023 08:49:17 -0700
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.10.0
Subject: Re: Regression wireless-next: mt7921: update the channel usage when
 the regd domain changed
Content-Language: en-US
To:     =?UTF-8?B?RGVyZW4gV3UgKOatpuW+t+S7gSk=?= <Deren.Wu@mediatek.com>,
        "linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>
Cc:     =?UTF-8?B?TWluZ3llbiBIc2llaCAo6Kyd5piO6Ku6KQ==?= 
        <Mingyen.Hsieh@mediatek.com>
References: <2fb78387-d226-3193-8ca7-90040561b9ad@candelatech.com>
 <e0afaea7118b742d695fc0ef19e04ff0cb6a902f.camel@mediatek.com>
From:   Ben Greear <greearb@candelatech.com>
Organization: Candela Technologies
In-Reply-To: <e0afaea7118b742d695fc0ef19e04ff0cb6a902f.camel@mediatek.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 10/25/23 20:58, Deren Wu (武德仁) wrote:
> Hi Ben,
> 
> On Wed, 2023-10-25 at 09:05 -0700, Ben Greear wrote:
>> I noticed mt7922 radios would no longer enable 6e band in wireless-
>> next.
>> 
>> My test system has Intel radios and mt7922 radios in it.  I bisected
>> to the commit
>> below.  It looks like that patch can disable 6e based on regdom
>> change, but cannot
>> enable 6e again in case regdom changes back to something that
>> supports 6e.
>> 
>> Intel tri-band radios disable 6e by default, then enable it after
>> scanning
>> US regdom beacons.  Maybe that regdom changing by Intel helps break
>> the mt7922
>> with the below patch applied?
> 
> Thanks for the report. There is a capability check missed. Could you
> pleae help to check this patch? Let's make sure you run into the same
> problem.
> 
> diff --git a/drivers/net/wireless/mediatek/mt76/mt7921/init.c
> b/drivers/net/wireless/mediatek/mt76/mt7921/init.c
> index 55baac70860b..a738ea95110e 100644
> --- a/drivers/net/wireless/mediatek/mt76/mt7921/init.c
> +++ b/drivers/net/wireless/mediatek/mt76/mt7921/init.c
> @@ -69,6 +69,9 @@ mt7921_regd_channel_update(struct wiphy *wiphy,
> struct mt792x_dev *dev)
>          struct ieee80211_channel *ch;
>          int i, cfreq;
> 
> +       if (!(dev->phy.chip_cap & MT792x_CHIP_CAP_CLC_EVT_EN))
> +               dev->phy.clc_chan_conf = 0xff;
> +
>          np = mt76_find_power_limits_node(mdev);
> 
>          sband = wiphy->bands[NL80211_BAND_5GHZ];

This fixed the problem in a quick test.

I'll add this to our tree and we'll do more testing on it...

Thanks,
Ben

-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com