Hi Sean,
> Runtime pm ops is not aware the sdio_func status that is probably
> being disabled by btmtksdio_close. Thus, we are only able to access the
> sdio_func for the runtime pm operations only when the sdio_func is
> available.
>
> Fixes: 7f3c563c575e7 ("Bluetooth: btmtksdio: Add runtime PM support to SDIO based Bluetooth")
> Co-developed-by: Mark-yw Chen <mark-yw.chen@xxxxxxxxxxxx>
> Signed-off-by: Mark-yw Chen <mark-yw.chen@xxxxxxxxxxxx>
> Signed-off-by: Sean Wang <sean.wang@xxxxxxxxxxxx>
> ---
> drivers/bluetooth/btmtksdio.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c
> index 4f3412ad8fca..4c46c62e4623 100644
> --- a/drivers/bluetooth/btmtksdio.c
> +++ b/drivers/bluetooth/btmtksdio.c
> @@ -1037,6 +1037,9 @@ static int btmtksdio_runtime_suspend(struct device *dev)
> if (!bdev)
> return 0;
>
> + if (!test_bit(HCI_RUNNING, &bdev->hdev->flags))
> + return 0;
> +
> sdio_claim_host(bdev->func);
>
> sdio_writel(bdev->func, C_FW_OWN_REQ_SET, MTK_REG_CHLPCR, &err);
> @@ -1064,6 +1067,9 @@ static int btmtksdio_runtime_resume(struct device *dev)
> if (!bdev)
> return 0;
>
> + if (!test_bit(HCI_RUNNING, &bdev->hdev->flags))
> + return 0;
> +
> sdio_claim_host(bdev->func);
>
> sdio_writel(bdev->func, C_FW_OWN_REQ_CLR, MTK_REG_CHLPCR, &err);
I dislike looking at HCI_RUNNING since that check should be removed from a driver. Do you really need it? I mean, a driver should now if it is running or not.
Regards
Marcel
