[PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Harry Geyer]

Patch allows usb devices with multiple interfaces to use this driver without
this driver assuming all interfaces are i2c-tiny-usb.

Signed-off-by: Harry Geyer <harry.geyer@devtank.co.uk>
---
 drivers/i2c/busses/i2c-tiny-usb.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/i2c/busses/i2c-tiny-usb.c b/drivers/i2c/busses/i2c-tiny-usb.c
index d1fa9ff5aeab..fc194c65e50e 100644
--- a/drivers/i2c/busses/i2c-tiny-usb.c
+++ b/drivers/i2c/busses/i2c-tiny-usb.c
@@ -222,6 +222,11 @@ static int i2c_tiny_usb_probe(struct usb_interface *interface,
 	int retval = -ENOMEM;
 	u16 version;
 
+	if (interface->intf_assoc &&
+		interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC) {
+		return -ENODEV;
+	}
+
 	dev_dbg(&interface->dev, "probing usb device\n");
 
 	/* allocate memory for our device state and initialize it */
-- 
2.34.1

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Andi Shyti]

Hi Harry,

it's OK, just few little adjustments...

On Wed, Jul 19, 2023 at 12:38:38PM +0100, Harry Geyer wrote:
> Patch allows usb devices with multiple interfaces to use this driver without
> this driver assuming all interfaces are i2c-tiny-usb.
> 
> Signed-off-by: Harry Geyer <harry.geyer@devtank.co.uk>

Please add:

Fixes: e8c76eed2ecd ("i2c: New i2c-tiny-usb bus driver")
Cc: Till Harbaum <till@harbaum.org>
Cc: <stable@vger.kernel.org> # v2.6.22+

> ---
>  drivers/i2c/busses/i2c-tiny-usb.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/drivers/i2c/busses/i2c-tiny-usb.c b/drivers/i2c/busses/i2c-tiny-usb.c
> index d1fa9ff5aeab..fc194c65e50e 100644
> --- a/drivers/i2c/busses/i2c-tiny-usb.c
> +++ b/drivers/i2c/busses/i2c-tiny-usb.c
> @@ -222,6 +222,11 @@ static int i2c_tiny_usb_probe(struct usb_interface *interface,
>  	int retval = -ENOMEM;
>  	u16 version;
>  
> +	if (interface->intf_assoc &&
> +		interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC) {

please fix the indentation here ('interface' and 'interface'
should go on the same line)

> +		return -ENODEV;
> +	}

please drop the brackets.

Andi

> +
>  	dev_dbg(&interface->dev, "probing usb device\n");
>  
>  	/* allocate memory for our device state and initialize it */
> -- 
> 2.34.1
> 

[PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Harry Geyer]

Patch allows usb devices with multiple interfaces to use this driver without
this driver assuming all interfaces are i2c-tiny-usb.

Signed-off-by: Harry Geyer <harry.geyer@devtank.co.uk>
---
 drivers/i2c/busses/i2c-tiny-usb.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/i2c/busses/i2c-tiny-usb.c b/drivers/i2c/busses/i2c-tiny-usb.c
index d1fa9ff5aeab..d6578e8908ac 100644
--- a/drivers/i2c/busses/i2c-tiny-usb.c
+++ b/drivers/i2c/busses/i2c-tiny-usb.c
@@ -222,6 +222,9 @@ static int i2c_tiny_usb_probe(struct usb_interface *interface,
 	int retval = -ENOMEM;
 	u16 version;
 
+	if (interface->intf_assoc && interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC)
+		return -ENODEV;
+
 	dev_dbg(&interface->dev, "probing usb device\n");
 
 	/* allocate memory for our device state and initialize it */
-- 
2.34.1

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Andi Shyti]

Hi Harry,

On Thu, Jul 27, 2023 at 09:33:54AM +0100, Harry Geyer wrote:
> Patch allows usb devices with multiple interfaces to use this driver without
> this driver assuming all interfaces are i2c-tiny-usb.
> 
> Signed-off-by: Harry Geyer <harry.geyer@devtank.co.uk>
> ---
>  drivers/i2c/busses/i2c-tiny-usb.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/i2c/busses/i2c-tiny-usb.c b/drivers/i2c/busses/i2c-tiny-usb.c
> index d1fa9ff5aeab..d6578e8908ac 100644
> --- a/drivers/i2c/busses/i2c-tiny-usb.c
> +++ b/drivers/i2c/busses/i2c-tiny-usb.c
> @@ -222,6 +222,9 @@ static int i2c_tiny_usb_probe(struct usb_interface *interface,
>  	int retval = -ENOMEM;
>  	u16 version;
>  
> +	if (interface->intf_assoc && interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC)

I'm sorry, I meant the same column, my bad here!
Would you please be so kind to resend it as


	if (interface->intf_assoc &&
	    interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC)

It's a nitpick, though.

On the other hand, checkpatch is OK with lines up to 100
characters and this one looks like it's withing that length. So,
up to your personal preference.

If you resend it, you can add my r-b.

Thank you and again, sorry,
Andi

PS Next time, please add a changelog and a versioning to
   following patches.

> +		return -ENODEV;
> +
>  	dev_dbg(&interface->dev, "probing usb device\n");
>  
>  	/* allocate memory for our device state and initialize it */
> -- 
> 2.34.1
> 

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Harry Geyer]

Hi Andi,

On 27/07/2023 12:30, Andi Shyti wrote:
> On the other hand, checkpatch is OK with lines up to 100
> characters and this one looks like it's withing that length. So,
> up to your personal preference.
I am happy to resend the patch.
> If you resend it, you can add my r-b.
Not sure what you mean, I am sorry. Please, can you explain what you 
want me to do.

Thank you,

Harry.

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Andi Shyti]

On Thu, Jul 27, 2023 at 02:54:42PM +0100, Harry Geyer wrote:
> Hi Andi,
> 
> On 27/07/2023 12:30, Andi Shyti wrote:
> > On the other hand, checkpatch is OK with lines up to 100
> > characters and this one looks like it's withing that length. So,
> > up to your personal preference.
> I am happy to resend the patch.

Thanks!

> > If you resend it, you can add my r-b.
> Not sure what you mean, I am sorry. Please, can you explain what you want me
> to do.

I was meaning that you could add my Reviewed-by:... you can
also ignore it, I can r-b it later.

Thanks,
Andi

PS r-b = Reviewed-by, if this was the reason of confusion :)

[PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Harry Geyer]

Patch allows usb devices with multiple interfaces to use this driver without
this driver assuming all interfaces are i2c-tiny-usb.

Signed-off-by: Harry Geyer <harry.geyer@devtank.co.uk>
---
 drivers/i2c/busses/i2c-tiny-usb.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/i2c/busses/i2c-tiny-usb.c b/drivers/i2c/busses/i2c-tiny-usb.c
index d1fa9ff5aeab..1bffe36c40ad 100644
--- a/drivers/i2c/busses/i2c-tiny-usb.c
+++ b/drivers/i2c/busses/i2c-tiny-usb.c
@@ -222,6 +222,10 @@ static int i2c_tiny_usb_probe(struct usb_interface *interface,
 	int retval = -ENOMEM;
 	u16 version;
 
+	if (interface->intf_assoc &&
+	    interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC)
+		return -ENODEV;
+
 	dev_dbg(&interface->dev, "probing usb device\n");
 
 	/* allocate memory for our device state and initialize it */
-- 
2.34.1

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Andi Shyti]

Hi Harry,

On Thu, Jul 27, 2023 at 05:22:55PM +0100, Harry Geyer wrote:
> Patch allows usb devices with multiple interfaces to use this driver without
> this driver assuming all interfaces are i2c-tiny-usb.
> 
> Signed-off-by: Harry Geyer <harry.geyer@devtank.co.uk>
> ---
>  drivers/i2c/busses/i2c-tiny-usb.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/i2c/busses/i2c-tiny-usb.c b/drivers/i2c/busses/i2c-tiny-usb.c
> index d1fa9ff5aeab..1bffe36c40ad 100644
> --- a/drivers/i2c/busses/i2c-tiny-usb.c
> +++ b/drivers/i2c/busses/i2c-tiny-usb.c
> @@ -222,6 +222,10 @@ static int i2c_tiny_usb_probe(struct usb_interface *interface,
>  	int retval = -ENOMEM;
>  	u16 version;
>  
> +	if (interface->intf_assoc &&
> +	    interface->intf_assoc->bFunctionClass != USB_CLASS_VENDOR_SPEC)
> +		return -ENODEV;
> +

Thanks for resending it!

Reviewed-by: Andi Shyti <andi.shyti@kernel.org> 

Andi

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Andi Shyti]

Hi

On Thu, 27 Jul 2023 17:22:55 +0100, Harry Geyer wrote:
> Patch allows usb devices with multiple interfaces to use this driver without
> this driver assuming all interfaces are i2c-tiny-usb.
> 
> 

Applied to i2c/andi-for-current on

https://git.kernel.org/pub/scm/linux/kernel/git/andi.shyti/linux.git

Please note that this patch may still undergo further evaluation
and the final decision will be made in collaboration with
Wolfram.

Thank you,
Andi

Patches applied
===============
[1/1] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver
      commit: c6607c0c7118ab96dcbc57df67fe316e871666a1

Re: [PATCH] i2c: tiny-usb: check usb base class before assuming the interface on device is for this driver

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 407 bytes --]


> Please note that this patch may still undergo further evaluation
> and the final decision will be made in collaboration with
> Wolfram.

I chose to apply it to for-next. I don't know how many clones are out
there and if they all have the same FunctionClass. So, I'd like to give
this some testing in -next and during the next release cycle.

Applied to for-next, thanks! Thanks for backing me up, Andi!


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]