linux-kernel - Re: [PATCH] Input: matrix_keypad

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]

Message-ID: <20180203134450.GC707@gmail.com>
Date:   Sat, 3 Feb 2018 14:44:50 +0100
From:   Marcus Folkesson <marcus.folkesson@...il.com>
To:     Zhang Bo <zbsdta@....com>
Cc:     dmitry.torokhov@...il.com, DRivshin@...worx.com, robh@...nel.org,
        linux-input@...r.kernel.org, linux-kernel@...r.kernel.org,
        zhang.bo19@....com.cn
Subject: Re: [PATCH] Input: matrix_keypad - fix keypad does not response

Hi!
On Sat, Feb 03, 2018 at 08:00:46PM +0800, Zhang Bo wrote:
> From: zhangbo <zbsdta@....com>
> 
> If matrix_keypad_stop() is calling and the keypad interrupt is triggered,
> disable_row_irqs() is called by both matrix_keypad_interrupt() and
> matrix_keypad_stop() at the same time. then disable_row_irqs() is called
> twice, and the device enter suspend state before keypad->work is executed.
> At this condition the device will start keypad and enable irq once after
> resume. and then irqs are disabled yet because irqs are disabled twice and
> only enable once.
> 
> Signed-off-by: zhangbo <zbsdta@....com>

Please use your full real name in the signed-off-by tag.

> ---
>  drivers/input/keyboard/matrix_keypad.c | 18 ++++++++++++------
>  1 file changed, 12 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c
> index 1f316d66e6f7..13fe51824637 100644
> --- a/drivers/input/keyboard/matrix_keypad.c
> +++ b/drivers/input/keyboard/matrix_keypad.c
> @@ -169,7 +169,8 @@ static void matrix_keypad_scan(struct work_struct *work)
>  	/* Enable IRQs again */
>  	spin_lock_irq(&keypad->lock);
>  	keypad->scan_pending = false;
> -	enable_row_irqs(keypad);
> +	if (keypad->stopped == false)
> +		enable_row_irqs(keypad);
>  	spin_unlock_irq(&keypad->lock);
>  }
>  
> @@ -202,14 +203,16 @@ static int matrix_keypad_start(struct input_dev *dev)
>  {
>  	struct matrix_keypad *keypad = input_get_drvdata(dev);
>  
> +	spin_lock_irq(&keypad->lock);
>  	keypad->stopped = false;
> -	mb();
>  
>  	/*
>  	 * Schedule an immediate key scan to capture current key state;
>  	 * columns will be activated and IRQs be enabled after the scan.
>  	 */
> -	schedule_delayed_work(&keypad->work, 0);
> +	if (keypad->scan_pending == false)
> +		schedule_delayed_work(&keypad->work, 0);
> +	spin_unlock_irq(&keypad->lock);
>  
>  	return 0;
>  }
> @@ -218,14 +221,17 @@ static void matrix_keypad_stop(struct input_dev *dev)
>  {
>  	struct matrix_keypad *keypad = input_get_drvdata(dev);
>  
> +	spin_lock_irq(&keypad->lock);
>  	keypad->stopped = true;
> -	mb();
> -	flush_work(&keypad->work.work);
>  	/*
>  	 * matrix_keypad_scan() will leave IRQs enabled;
>  	 * we should disable them now.
>  	 */
> -	disable_row_irqs(keypad);
> +	if (keypad->scan_pending == false)
> +		disable_row_irqs(keypad);
> +	spin_unlock_irq(&keypad->lock);
> +
> +	flush_work(&keypad->work.work);
>  }


Hum, I think we should use spin_lock_irqsave/spin_lock_irqrestore
instead to be on the safe side. 
I don't see how we could guarantee that irqs is allways enabled when
calling spin_lock_irq().


>  
>  #ifdef CONFIG_PM_SLEEP
> -- 
> 2.14.3


Best regards
Marcus Folkesson
> 

Download attachment "signature.asc" of type "application/pgp-signature" (834 bytes)