Re: [PATCH 5/5] input/rmi4/rmi_smbus.c: don't increment rmiaddr in rmi_smb_read_block()

[Lucas Stach <l.stach@pengutronix.de>]

Hi Hans,

On Di, 2019-11-19 at 11:51 +0100, Hans Verkuil wrote:
> This increment of rmi_smbus causes garbage to be returned.
> The first read of SMB_MAX_COUNT bytes is fine, but after that
> it is nonsense. Trial-and-error showed that by dropping the
> increment of rmiaddr everything is fine and the F54 function
> properly works.
> 
> Even going back to the original code when F54 was added, I
> could not make it work without this patch. So I do not understand
> how this ever worked.

My guess is that F54 has mostly been tested with touchscreens that are
connected to a regular i2c bus, not smbus. i2c has a separate transport
implementation in rmi4. Most of the other functions are probably
reading much smaller block data than F54.

Regards,
Lucas

> My guess is that the same change is needed in rmi_smb_write_block,
> but I wouldn't know how to test this.
> 
> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
> ---
>  drivers/input/rmi4/rmi_smbus.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/input/rmi4/rmi_smbus.c b/drivers/input/rmi4/rmi_smbus.c
> index 2407ea43de59..79ecea5edacc 100644
> --- a/drivers/input/rmi4/rmi_smbus.c
> +++ b/drivers/input/rmi4/rmi_smbus.c
> @@ -215,7 +215,6 @@ static int rmi_smb_read_block(struct rmi_transport_dev *xport, u16 rmiaddr,
>  		/* prepare to read next block of bytes */
>  		cur_len -= SMB_MAX_COUNT;
>  		databuff += SMB_MAX_COUNT;
> -		rmiaddr += SMB_MAX_COUNT;
>  	}
>  
>  	retval = 0;