linux-kernel - Re: [PATCH] serial: port: Don't suspend if the port is still busy

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] [day] [month] [year] [list]

Message-ID: <ea950cd4-b778-12e3-4ccd-301b213da450@huawei.com>
Date: Tue, 6 Feb 2024 16:02:34 +0800
From: Yicong Yang <yangyicong@...wei.com>
To: Tony Lindgren <tony@...mide.com>
CC: <yangyicong@...ilicon.com>, <gregkh@...uxfoundation.org>,
	<jirislaby@...nel.org>, <linux-kernel@...r.kernel.org>,
	<linux-serial@...r.kernel.org>, <john.ogness@...utronix.de>,
	<andriy.shevchenko@...ux.intel.com>, <tglx@...utronix.de>,
	<linuxarm@...wei.com>, <prime.zeng@...ilicon.com>,
	<jonathan.cameron@...wei.com>, <fanghao11@...wei.com>
Subject: Re: [PATCH] serial: port: Don't suspend if the port is still busy

On 2024/2/5 17:07, Tony Lindgren wrote:
> * Yicong Yang <yangyicong@...wei.com> [240205 08:55]:
>> On 2024/2/5 14:51, Tony Lindgren wrote:
>>> Can you please confirm if this still happens also with commit 6f699743aebf
>>> ("serial: core: Fix runtime PM handling for pending tx")? It adds a check
>>> for -EINPROGRESS.
>>
>> Tested nagetive on latest v6.8-rc3. Paste the current code snippet below in __uart_start():
> 
> OK thanks for confirming it.
> 
>> In our issue case, the dev->power.runtime_status == RPM_RESUMING as analyzed in
>> commit. So we cannot pass the pm_runtime_active() check and the chars will still
>> be pending.
> 
> OK
> 
>> Do you mean something like below?
>>
>> static int serial_port_runtime_suspend(struct device *dev)
>> {
>> 	struct serial_port_device *port_dev = to_serial_base_port_device(dev);
>> 	struct uart_port *port;
>> 	unsigned long flags;
>> 	int ret = 0;
>>
>> 	port = port_dev->port;
>>
>> 	if (port->flags & UPF_DEAD)
>> 		return ret;
>>
>> 	spin_lock_irqsave(&port->lock, flags);
>> 	if (__serial_port_busy(port)) {
>> 		port->ops->start_tx(port);
>> 		ret = -EBUSY;
>> 	}
>> 	spin_unlock_irqrestore(&port->lock, flags);
>>
>> 	return ret;
>> }
> 
> Yes the above should work.
> 
>> If so will the port fail to suspend after flushing the pending chars? Considering
>> underlay driver like amba-pl011 doesn't implement runtime power management, does
>> anyone will get the port into suspend routine later? I'm not quite sure about it.
> 
> Hmm yeah you may need to also call pm_runtime_mark_last_busy() to
> ensure the port gets idled later on. Not sure if PM runtime core does that for
> you on returning -EBUSY, worth checking it :)

In this if the runtime_suspend() callback return -EBUSY, rpm core will try to
repeat to try to suspend the device. So this shall be ok.

So I respin a v2 as suggested:
https://lore.kernel.org/all/20240206073322.5560-1-yangyicong@huawei.com/

> 
> The PM runtime hierarchy will block the serial port controller driver from
> suspending, so the port drivers won't runtime suspend.
> 
>> In the patch's implementation the pending chars will be flushed in runtime_resume()
>> callback and rpm_resume() will try to call rpm_idle() later.
> 
> On serial_port_runtime_suspend() the serial port controller will be active, so
> you can call start_tx() directly.
> 
> Regards,
> 
> Tony
> 
> .
>