Received-SPF: pass (google.com: domain of linux-bluetooth+bounces-5019-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Message-ID: <ecee3a54-1a09-40fa-afdb-057ca02cb574@leemhuis.info>
Date: Wed, 29 May 2024 11:06:51 +0200
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: Bluetooth kernel BUG with Intel AX211 (regression in 6.1.83)
To: Mike <user.service2016@gmail.com>
Cc: "stable@vger.kernel.org" <stable@vger.kernel.org>,
 Marcel Holtmann <marcel@holtmann.org>,
 Johan Hedberg <johan.hedberg@gmail.com>,
 Luiz Augusto von Dentz <luiz.dentz@gmail.com>,
 linux-bluetooth@vger.kernel.org, Paul Menzel <pmenzel@molgen.mpg.de>,
 Greg KH <gregkh@linuxfoundation.org>, Sasha Levin <sashal@kernel.org>,
 =?UTF-8?Q?Jeremy_Lain=C3=A9?= <jeremy.laine@m4x.org>,
 Linux regressions mailing list <regressions@lists.linux.dev>
References: <c947e600-e126-43ea-9530-0389206bef5e@gmail.com>
From: Thorsten Leemhuis <regressions@leemhuis.info>
Content-Language: en-US, de-DE
In-Reply-To: <c947e600-e126-43ea-9530-0389206bef5e@gmail.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

On 28.05.24 22:54, Mike wrote:
> On 29.04.24 20:46, Linux regression tracking (Thorsten Leemhuis) wrote:>
>> Well, did you try what I suggested earlier (see above) and check if a
>> revert of 6083089ab00631617f9eac678df3ab050a9d837a ontop of latest 6.1.y
>> helps?
>
> I hope you don't mind if I jump into the conversation trying to help.

On the contrary, thx for providing the needed information!

> I tried reverting 6083089ab00631617f9eac678df3ab050a9d837a
> on top of6.1.91 and it looks much better: it's been 10 days, and the BT
> and the system are stable.
> Previously, I encountered the mentioned "kernel BUG" at each boot,

Might be a good idea to share it, the developers might want to confirm
it's really the same bug.

> and I was unable to stop/kill the bluetoothd process.
> Let me know if/how I can help further.

Jeremy Lainé already confirmed that 6.9-rc5[1] worked fine and that
another fix for the culprit did not help[2]. Therefore we just have
three options left:

1. test another fix for the culprit I found on lore -- but note, this is
just a shot in the dark
https://lore.kernel.org/all/20240411151929.403263-1-kovalev@altlinux.org/

2. revert 6083089ab00631617f9eac678df3ab050a9d837a in 6.1.y if that is
still possible, does not create a even bigger regression, or leads to
some security vulnerability

3. motivate the BT developers to look into this (some other patch the
culprit depends on might be missing), even if this strictly speaking is
a problem they are free to ignore.

Maybe give "1." a try; then we'll ask Greg for "2.", unless this
discussion or something else leads to "3."

Ciao, Thorsten

[1]
https://lore.kernel.org/all/CADRbXaA2yFjMo=_8_ZTubPbrrmWH9yx+aG5pUadnk395koonXg@mail.gmail.com/
[2]
https://lore.kernel.org/all/CADRbXaBkkGmqnibGvcAF2YH5CjLRJ2bnnix1xKozKdw_Hv3qNg@mail.gmail.com/