Re: [PATCH] Bluetooth: Always allow SCO packets for user channel

[Hsin-chen Chuang <chharry@google.com>]

Hi Luiz,

On Fri, Feb 14, 2025 at 7:17 PM Hsin-chen Chuang <chharry@google.com> wrote:
>
> From: Hsin-chen Chuang <chharry@chromium.org>
>
> The SCO packets from Bluetooth raw socket are now rejected because
> hci_conn_num is left 0. This patch allows such the usecase to enable
> the userspace SCO support.
>
> Fixes: b16b327edb4d ("Bluetooth: btusb: add sysfs attribute to control USB alt setting")
> Signed-off-by: Hsin-chen Chuang <chharry@chromium.org>
> ---
>
>  drivers/bluetooth/btusb.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
> index 1caf7a071a73..de3fa725d210 100644
> --- a/drivers/bluetooth/btusb.c
> +++ b/drivers/bluetooth/btusb.c
> @@ -2130,7 +2130,8 @@ static int btusb_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
>                 return submit_or_queue_tx_urb(hdev, urb);
>
>         case HCI_SCODATA_PKT:
> -               if (hci_conn_num(hdev, SCO_LINK) < 1)
> +               if (!hci_dev_test_flag(hdev, HCI_USER_CHANNEL) &&
> +                   hci_conn_num(hdev, SCO_LINK) < 1)
>                         return -ENODEV;
>
>                 urb = alloc_isoc_urb(hdev, skb);
> @@ -2604,7 +2605,8 @@ static int btusb_send_frame_intel(struct hci_dev *hdev, struct sk_buff *skb)
>                 return submit_or_queue_tx_urb(hdev, urb);
>
>         case HCI_SCODATA_PKT:
> -               if (hci_conn_num(hdev, SCO_LINK) < 1)
> +               if (!hci_dev_test_flag(hdev, HCI_USER_CHANNEL) &&
> +                   hci_conn_num(hdev, SCO_LINK) < 1)
>                         return -ENODEV;
>
>                 urb = alloc_isoc_urb(hdev, skb);
> --
> 2.48.1.601.g30ceb7b040-goog
>

Friendly ping for review, thanks.

-- 
Best Regards,
Hsin-chen