Re: [PATCH 2/2] imx-usb-loader: Add support for i.MX8MP

[Sascha Hauer <sha@pengutronix.de>]

On Thu, Aug 05, 2021 at 05:22:51PM +0200, Uwe Kleine-König wrote:
> From: Uwe Kleine-Koenig <u.kleine-koenig@pengutronix.de>
> 
> The i.MX8MP uses a protocol similar to the MXS. The relevant differences
> are:
> 
>  - Maximal transfer size is 1020
>  - HID reports must be sent to EP1 instead of using a control transfer
>  - The FW_DOWNLOAD command must not be send.
> 
> Signed-off-by: Uwe Kleine-Koenig <u.kleine-koenig@pengutronix.de>
> ---
>  scripts/imx/imx-usb-loader.c | 65 +++++++++++++++++++++++-------------
>  1 file changed, 42 insertions(+), 23 deletions(-)
> 
> diff --git a/scripts/imx/imx-usb-loader.c b/scripts/imx/imx-usb-loader.c
> index 3e96c86f2f29..40a85ee26105 100644
> --- a/scripts/imx/imx-usb-loader.c
> +++ b/scripts/imx/imx-usb-loader.c
> @@ -71,6 +71,7 @@ struct mach_id {
>  #define DEV_IMX		0
>  #define DEV_MXS		1
>  	unsigned char dev_type;
> +	unsigned char hid_endpoint;
>  };
>  
>  struct usb_work {
> @@ -177,6 +178,13 @@ static const struct mach_id imx_ids[] = {
>  		.header_type = HDR_MX53,
>  		.mode = MODE_HID,
>  		.max_transfer = 1024,
> +	}, {
> +		.vid = 0x1fc9,
> +		.pid = 0x0146,
> +		.name = "i.MX8MP",
> +		.max_transfer = 1020,
> +		.dev_type = DEV_MXS,
> +		.hid_endpoint = 1,
>  	}, {
>  		.vid = 0x1fc9,
>  		.pid = 0x012b,
> @@ -522,15 +530,24 @@ static int transfer(int report, unsigned char *p, unsigned cnt, int *last_trans)
>  
>  		if (report < 3) {
>  			memcpy(&tmp[1], p, cnt);
> -			err = libusb_control_transfer(usb_dev_handle,
> -					CTRL_OUT,
> -					HID_SET_REPORT,
> -					(HID_REPORT_TYPE_OUTPUT << 8) | report,
> -					0,
> -					tmp, cnt + 1, 1000);
> -			*last_trans = (err > 0) ? err - 1 : 0;
> -			if (err > 0)
> -				err = 0;
> +			if (mach_id->hid_endpoint) {
> +				int trans;
> +				err = libusb_interrupt_transfer(usb_dev_handle,
> +						mach_id->hid_endpoint, tmp, cnt + 1, &trans, 1000);
> +				if (err == 0 && last_trans)

last_trans is assumed to be non NULL throughout this function, so there
shouldn't be a need to check it here.

> +					*last_trans = trans - 1;
> +			} else {
> +				err = libusb_control_transfer(usb_dev_handle,
> +						CTRL_OUT,
> +						HID_SET_REPORT,
> +						(HID_REPORT_TYPE_OUTPUT << 8) | report,
> +						0,
> +						tmp, cnt + 1, 1000);
> +				if (err > 0) {
> +					*last_trans = err - 1;
> +					err = 0;
> +				}

Previously *last_trans was set unconditionally, with this patch you only
set it when err > 0. I don't know if this makes a difference, but I
would expect this change in a separate patch.

Sascha

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox