[PATCH] Bluetooth: btusb: Handle download_firmware failure cases.

[Paul Menzel]

Dear Mark,


Some small nitpicks:

Am 06.12.21 um 10:25 schrieb mark-yw.chen at mediatek.com:
> From: mark-yw.chen <mark-yw.chen at mediatek.com>

Instead of the user name, maybe you can use Mark Chen (or the whole name)?

     $ git config --global user.name "Mark Chen"
     $ git commit --amend --reset-author="Mark Chen 
<mark-yw.chen at mediatek.com>"

Also could you please remove the trailing dot/period at the end of the 
git commit message summary?

> For Mediatek chipset, if there are no firmware bin or command_timeout,
> the process should be terminated in btusb_mtk_setup().

Otherwise what happens?

> Signed-off-by: mark-yw.chen <mark-yw.chen at mediatek.com>
> Change-Id: I99f1d7b72fa70643d9123e7e6cdc8d0b4369ce52

To what Gerrit instance does the Change-Id belong? Without that 
information (Reviewed-on tag?), it should be removed?

> ---
>   drivers/bluetooth/btmtk.c | 1 +
>   drivers/bluetooth/btusb.c | 4 ++++
>   2 files changed, 5 insertions(+)
> 
> diff --git a/drivers/bluetooth/btmtk.c b/drivers/bluetooth/btmtk.c
> index c2ee5c4b975a..526dfdf1fe01 100644
> --- a/drivers/bluetooth/btmtk.c
> +++ b/drivers/bluetooth/btmtk.c
> @@ -121,6 +121,7 @@ int btmtk_setup_firmware_79xx(struct hci_dev *hdev, const char *fwname,
>   				} else {
>   					bt_dev_err(hdev, "Failed wmt patch dwnld status (%d)",
>   						   status);
> +					err = -EIO;
>   					goto err_release_fw;
>   				}
>   			}
> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
> index ab169fc673ea..3ea04b1d0750 100644
> --- a/drivers/bluetooth/btusb.c
> +++ b/drivers/bluetooth/btusb.c
> @@ -2554,6 +2554,10 @@ static int btusb_mtk_setup(struct hci_dev *hdev)
>   			 dev_id & 0xffff, (fw_version & 0xff) + 1);
>   		err = btmtk_setup_firmware_79xx(hdev, fw_bin_name,
>   						btusb_mtk_hci_wmt_sync);
> +		if (err < 0) {
> +			bt_dev_err(hdev, "Failed to setup firmware (%d)", err);

The verb is spelled with a space: set up.

Also, this error message seems unrelated to the patch in question. Maybe 
add it in a separate commit?

> +			return err;
> +		}
>   
>   		/* It's Device EndPoint Reset Option Register */
>   		btusb_mtk_uhw_reg_write(data, MTK_EP_RST_OPT, MTK_EP_RST_IN_OUT_OPT);
> 


Kind regards,

Paul