linux-kernel - Re: [PATCH V1] extcon/usb: Remove disable irq operation in system sleep.

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:   Thu, 3 Mar 2022 11:20:31 +0200
From:   Roger Quadros <rogerq@...nel.org>
To:     Bruce Chen <brucechen251@...il.com>, bruce.chen@...soc.com,
        myungjoo.ham@...sung.com, cw00.choi@...sung.com,
        linux-kernel@...r.kernel.org,
        Aswath Govindraju <a-govindraju@...com>,
        Vignesh Raghavendra <vigneshr@...com>
Cc:     orsonzhai@...il.com, gengcixi@...il.com, baolin.wang7@...il.com,
        zhang.lyra@...il.com
Subject: Re: [PATCH V1] extcon/usb: Remove disable irq operation in system
 sleep.

+TI folks Aswath & Vignesh.

Hi Bruce,

On 03/03/2022 09:36, Bruce Chen wrote:
> From: Bruce Chen <bruce.chen@...soc.com>
> 
> If disable vbus/id irq, it will lead to wakeup system fail
> in unisoc platform. In unisoc platform, Irq enable and irq
> wakeup are the same interrupt line. So remove disable vbus/id
> irq operation is a way to solve the issue> 
> Signed-off-by: Bruce Chen <bruce.chen@...soc.com>
> ---
>  drivers/extcon/extcon-usb-gpio.c | 15 ---------------
>  1 file changed, 15 deletions(-)
> 
> diff --git a/drivers/extcon/extcon-usb-gpio.c b/drivers/extcon/extcon-usb-gpio.c
> index f2b65d9..40d967a 100644
> --- a/drivers/extcon/extcon-usb-gpio.c
> +++ b/drivers/extcon/extcon-usb-gpio.c
> @@ -226,16 +226,6 @@ static int usb_extcon_suspend(struct device *dev)
>  		}
>  	}
>  
> -	/*
> -	 * We don't want to process any IRQs after this point
> -	 * as GPIOs used behind I2C subsystem might not be
> -	 * accessible until resume completes. So disable IRQ.
> -	 */

I think the reason this was added was that if USB cable is plugged/unplugged
while system was suspending it was causing some faults on TI DRA7 EVM
at usb_irq_handler()/usb_extcon_detect_cable() when it tries to read GPIO status
when I2C based GPIO controller was used but already suspended.
But looks like disabling the IRQ is not the right approach either so your patch
is right.

However I'm not sure if the above issue will still exist on DRA7 and other TI EVMS
using ID/VBUS GPIO over I2C device.

Let's wait for Aswath/Vignesh to confirm. If no issues I'm OK with this patch.

> -	if (info->id_gpiod)
> -		disable_irq(info->id_irq);
> -	if (info->vbus_gpiod)
> -		disable_irq(info->vbus_irq);
> -
>  	if (!device_may_wakeup(dev))
>  		pinctrl_pm_select_sleep_state(dev);
>  
> @@ -267,11 +257,6 @@ static int usb_extcon_resume(struct device *dev)
>  		}
>  	}
>  
> -	if (info->id_gpiod)
> -		enable_irq(info->id_irq);
> -	if (info->vbus_gpiod)
> -		enable_irq(info->vbus_irq);
> -
>  	queue_delayed_work(system_power_efficient_wq,
>  			   &info->wq_detcable, 0);
>  

cheers,
-roger