linux-kernel - Re: [PATCH v3 3/3] soundwire: qcom: add in-band wake up interrupt support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 1 Mar 2022 11:13:30 +0000
From:   Srinivas Kandagatla <srinivas.kandagatla@...aro.org>
To:     Pierre-Louis Bossart <pierre-louis.bossart@...ux.intel.com>,
        robh+dt@...nel.org, vkoul@...nel.org,
        yung-chuan.liao@...ux.intel.com
Cc:     devicetree@...r.kernel.org, alsa-devel@...a-project.org,
        linux-kernel@...r.kernel.org, quic_srivasam@...cinc.com
Subject: Re: [PATCH v3 3/3] soundwire: qcom: add in-band wake up interrupt
 support



On 28/02/2022 18:01, Pierre-Louis Bossart wrote:
> 
>> @@ -1424,6 +1464,11 @@ static int swrm_runtime_resume(struct device *dev)
>>   	struct qcom_swrm_ctrl *ctrl = dev_get_drvdata(dev);
>>   	int ret;
>>   
>> +	if (ctrl->wake_irq > 0) {
>> +		if (!irqd_irq_disabled(irq_get_irq_data(ctrl->wake_irq)))
>> +			disable_irq_nosync(ctrl->wake_irq);
>> +	}
>> +
>>   	clk_prepare_enable(ctrl->hclk);
> 
> This one is quite interesting. If you disable the IRQ mechanism but
> haven't yet resumed the clock, that leaves a time window where the
> peripheral could attempt to drive the line high. what happens in that case?


We did call pm_runtime_get_sync() from Wake IRQ handler, which means 
that resume should be finished as part of Wake IRQ handler. Any new 
Interrupt conditions/status generated by slave in the meantime will be 
cleared while handling SLAVE PEND interrupt.

> 
>>   
>>   	if (ctrl->clock_stop_not_supported) {
>> @@ -1491,6 +1536,11 @@ static int __maybe_unused swrm_runtime_suspend(struct device *dev)
>>   
>>   	usleep_range(300, 305);
>>   
>> +	if (ctrl->wake_irq > 0) {
>> +		if (irqd_irq_disabled(irq_get_irq_data(ctrl->wake_irq)))
>> +			enable_irq(ctrl->wake_irq);
>> +	}
>> +
> 
> and this one is similar, you could have a case where the peripheral
> signals a wake immediately after the ClockStopNow frame, but you may not
> yet have enabled the wake detection interrupt.
> 
> Would that imply that the wake is missed?
Its Possible it might be missed at that instance, however as the Slave 
interrupt source condition/status (Ex: button Press) is still not 
cleared it should generate a Wake interrupt as soon as its enabled.

--srini
> 
> 
> 
>>   	return 0;
>>   }
>>