Received-SPF: pass (google.com: domain of linux-wireless+bounces-7079-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Message-ID: <824269da-c69d-668b-4e6b-d0453c4572c1@quicinc.com>
Date: Wed, 1 May 2024 21:04:04 +0530
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101
 Thunderbird/102.0
Subject: Re: [PATCH v2 0/8] wifi: ath12k: Introduce device group abstraction
To: Kalle Valo <kvalo@kernel.org>
CC: <ath12k@lists.infradead.org>, <linux-wireless@vger.kernel.org>
References: <20240425055759.1181165-1-quic_hprem@quicinc.com>
 <87le4voyka.fsf@kernel.org>
Content-Language: en-US
From: Harshitha Prem <quic_hprem@quicinc.com>
In-Reply-To: <87le4voyka.fsf@kernel.org>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 7bit


On 4/30/2024 5:24 PM, Kalle Valo wrote:
> 47] WARNING: possible circular locking dependency detected
> [  147.211599] 6.9.0-rc5-wt-ath+ #1403 Not tainted

Thank you, Kalle, for sharing the full logs. Unfortunately, I did not 
see this warning in my testbed with QCN9274.

Seems like during rmmod case, ath12k_core_soc_destroy() is invoked with 
ag->mutex lock which in-turn invokes ath12k_qmi_deinit_service() where 
cancel_work_sync for qmi events are called. So, if 
ATH12K_QMI_EVENT_FW_READY was in queue during rmmod, it would hit a 
deadlock as in core start it requires ag->mutex lock. But, this scenario 
is quite unlikely to occur where rmmod and FW_READY event are going to 
be in parallel. Nevertheless, it has to be addressed. I will check 
further on this warning and see how to address this.

Thanks,
Harshitha