linux-kernel - Re: [PATCH v13] USB: serial: cp210x: add support for GPIOs on CP2108

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]

Date:   Wed, 16 Jun 2021 17:44:43 +0200
From:   Johan Hovold <johan@...nel.org>
To:     tu pham <thanhtung1909@...il.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        linux-usb@...r.kernel.org, linux-kernel@...r.kernel.org,
        Hung.Nguyen@...abs.com, Tung.Pham@...abs.com,
        Pho Tran <pho.tran@...abs.com>
Subject: Re: [PATCH v13] USB: serial: cp210x: add support for GPIOs on CP2108

On Thu, Jun 10, 2021 at 03:28:44PM +0200, Johan Hovold wrote:
> From: Pho Tran <pho.tran@...abs.com>
> 
> Similar to some other CP210x device types, CP2108 has a number of GPIO
> pins that can be exposed through gpiolib.
> 
> CP2108 has four serial interfaces but only one set of GPIO pins, which
> is modelled as a single gpio chip and registered as a child of the first
> interface.
> 
> CP2108 has 16 GPIOs so the width of the state variables needs to be
> extended to 16 bits and this is also reflected in the control requests.
> 
> Like CP2104, CP2108 have GPIO pins with configurable alternate
> functions and pins unavailable for GPIO use are determined and reported
> to gpiolib at probe.
> 
> Signed-off-by: Pho Tran <pho.tran@...abs.com>
> Co-developed-by: Tung Pham <tung.pham@...abs.com>
> Signed-off-by: Tung Pham <tung.pham@...abs.com>
> [ johan: rewrite gpio get() and set(); misc cleanups; amend commit
>          message ]
> Signed-off-by: Johan Hovold <johan@...nel.org>
> ---
> 
> Tested using CP2108 and CP2102N.
> 
> Changes in v13
>  - rewrite cp210x_gpio_get() using a shared 16 bit mask and
>    le16_to_cpus()
>  - rewrite cp210x_gpio_set() using shared 16-bit mask and state
>    variables
>  - drop pointless no-op shift and mask operations during initialisation
>  - reorder defines
>  - reword some comments
>  - fix some style issues
>  - amend commit message
> 
> v12 can be found here:
>  - https://lore.kernel.org/r/20210426091244.19994-1-tupham@silabs.com
> 
> 
>  drivers/usb/serial/cp210x.c | 189 ++++++++++++++++++++++++++++++++----
>  1 file changed, 170 insertions(+), 19 deletions(-)

> @@ -1414,52 +1468,84 @@ static int cp210x_gpio_get(struct gpio_chip *gc, unsigned int gpio)
>  {
>  	struct usb_serial *serial = gpiochip_get_data(gc);
>  	struct cp210x_serial_private *priv = usb_get_serial_data(serial);
> -	u8 req_type = REQTYPE_DEVICE_TO_HOST;
> +	u8 req_type;
> +	u16 mask;
>  	int result;
> -	u8 buf;
> -
> -	if (priv->partnum == CP210X_PARTNUM_CP2105)
> -		req_type = REQTYPE_INTERFACE_TO_HOST;
> +	int len;
>  
>  	result = usb_autopm_get_interface(serial->interface);
>  	if (result)
>  		return result;
>  
> -	result = cp210x_read_vendor_block(serial, req_type,
> -					  CP210X_READ_LATCH, &buf, sizeof(buf));
> +	switch (priv->partnum) {
> +	case CP210X_PARTNUM_CP2105:
> +		req_type = REQTYPE_INTERFACE_TO_HOST;
> +		len = 1;
> +		break;
> +	case CP210X_PARTNUM_CP2108:
> +		req_type = REQTYPE_INTERFACE_TO_HOST;
> +		len = 2;
> +		break;
> +	default:
> +		req_type = REQTYPE_DEVICE_TO_HOST;
> +		len = 1;
> +		break;
> +	}
> +
> +	mask = 0;
> +	result = cp210x_read_vendor_block(serial, req_type, CP210X_READ_LATCH,
> +					  &mask, len);
> +
>  	usb_autopm_put_interface(serial->interface);
> +
>  	if (result < 0)
>  		return result;
>  
> -	return !!(buf & BIT(gpio));
> +	le16_to_cpus((__le16 *)&mask);

Now applied after removing the (__le16 *) cast here.

> +
> +	return !!(mask & BIT(gpio));
>  }

Johan