Re: [PATCH v3 20/24] em28xx: Fix em28xx deplock

["Frank Schäfer" <fschaefer.oss@googlemail.com>]

Am 28.12.2013 13:16, schrieb Mauro Carvalho Chehab:
> From: Mauro Carvalho Chehab <m.chehab@samsung.com>
>
> When em28xx extensions are loaded/removed, there are two locks:
>
> a single static em28xx_devlist_mutex that registers each extension
> and the struct em28xx dev->lock.
>
> When extensions are registered, em28xx_devlist_mutex is taken first,
> and then dev->lock.
>
> Be sure that, when extensions are being removed, the same order
> will be used.
>
> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
> ---
>  drivers/media/usb/em28xx/em28xx-cards.c | 10 ++++++----
>  drivers/media/usb/em28xx/em28xx-core.c  |  2 ++
>  2 files changed, 8 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/media/usb/em28xx/em28xx-cards.c b/drivers/media/usb/em28xx/em28xx-cards.c
> index 4fe742429f2c..16383f46dae9 100644
> --- a/drivers/media/usb/em28xx/em28xx-cards.c
> +++ b/drivers/media/usb/em28xx/em28xx-cards.c
> @@ -3334,9 +3334,7 @@ static void em28xx_usb_disconnect(struct usb_interface *interface)
>  	dev->disconnected = 1;
>  
>  	if (dev->is_audio_only) {
> -		mutex_lock(&dev->lock);
>  		em28xx_close_extension(dev);
> -		mutex_unlock(&dev->lock);
>  		return;
>  	}
>  
> @@ -3355,19 +3353,23 @@ static void em28xx_usb_disconnect(struct usb_interface *interface)
>  		em28xx_uninit_usb_xfer(dev, EM28XX_ANALOG_MODE);
>  		em28xx_uninit_usb_xfer(dev, EM28XX_DIGITAL_MODE);
>  	}
> +	mutex_unlock(&dev->lock);
>  
>  	em28xx_close_extension(dev);
> +
>  	/* NOTE: must be called BEFORE the resources are released */
>  
> +	mutex_lock(&dev->lock);
>  	if (!dev->users)
>  		em28xx_release_resources(dev);
>  

> -	mutex_unlock(&dev->lock);
> -
>  	if (!dev->users) {
> +		mutex_unlock(&dev->lock);
>  		kfree(dev->alt_max_pkt_size_isoc);
>  		kfree(dev);
> +		return;
>  	}
> +	mutex_unlock(&dev->lock);
No functional change here, it just needlessly complicates the code.
I assume it's a leftover from experiments. ;)

>  }
>  
>  static struct usb_driver em28xx_usb_driver = {
> diff --git a/drivers/media/usb/em28xx/em28xx-core.c b/drivers/media/usb/em28xx/em28xx-core.c
> index 2ad84ff1fc4f..d6928d83fb2a 100644
> --- a/drivers/media/usb/em28xx/em28xx-core.c
> +++ b/drivers/media/usb/em28xx/em28xx-core.c
> @@ -1098,8 +1098,10 @@ void em28xx_close_extension(struct em28xx *dev)
>  
>  	mutex_lock(&em28xx_devlist_mutex);
>  	list_for_each_entry(ops, &em28xx_extension_devlist, next) {
> +		mutex_lock(&dev->lock);
>  		if (ops->fini)
>  			ops->fini(dev);
> +		mutex_unlock(&dev->lock);
>  	}

Why not move the locking/unlocking of dev->lock outside
list_for_each_entry() ?
No need to do this one time for each extension.

>  	list_del(&dev->devlist);
>  	mutex_unlock(&em28xx_devlist_mutex);
Apart from these 2 minor issues, the patch looks good and should fix the
warning.