Hi,
On 4-Nov-24 9:50 AM, Stanislaw Gruszka wrote:
> Do not mark interface as ready to suspend when we are still waiting
> for response messages from the device.
>
> Fixes: acd6199f195d ("usb: Add support for Intel LJCA device")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Stanislaw Gruszka <stanislaw.gruszka@xxxxxxxxxxxxxxx>
> ---
> drivers/usb/misc/usb-ljca.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/usb/misc/usb-ljca.c b/drivers/usb/misc/usb-ljca.c
> index 01ceafc4ab78..dcb3c5d248ac 100644
> --- a/drivers/usb/misc/usb-ljca.c
> +++ b/drivers/usb/misc/usb-ljca.c
> @@ -332,9 +332,6 @@ static int ljca_send(struct ljca_adapter *adap, u8 type, u8 cmd,
>
> ret = usb_bulk_msg(adap->usb_dev, adap->tx_pipe, header,
> msg_len, &transferred, LJCA_WRITE_TIMEOUT_MS);
> -
> - usb_autopm_put_interface(adap->intf);
> -
> if (ret < 0)
> goto out;
> if (transferred != msg_len) {
> @@ -353,6 +350,8 @@ static int ljca_send(struct ljca_adapter *adap, u8 type, u8 cmd,
> ret = adap->actual_length;
>
> out:
> + usb_autopm_put_interface(adap->intf);
> +
I'm afraid that this now does a double pm_runtime_put() on
usb_autopm_get_interface() failures. usb_autopm_get_interface() uses
pm_runtime_resume_and_get() which already does a pm_runtime_put()
on failure.
So you need to add a second out label which skips the new
location of the usb_autopm_put_interface(adap->intf); call
and jump to this second label on usb_autopm_get_interface()
failures.
With that fixed this is:
Reviewed-by: Hans de Goede <hdegoede@xxxxxxxxxx>
I have also given this a test run on a "ThinkPad X1 Yoga Gen 8" and
everything there works at least as well as before:
Tested-by: Hans de Goede <hdegoede@xxxxxxxxxx> # ThinkPad X1 Yoga Gen 8, ov2740
Regards,
Hans
> spin_lock_irqsave(&adap->lock, flags);
> adap->ex_buf = NULL;
> adap->ex_buf_len = 0;
