ps_wake_refcount warning in ath10k_pci_hif_stop

[Michal Kazior]

On 4 August 2015 at 20:42, Ben Greear <greearb at candelatech.com> wrote:
> I'm porting forward to the 4.2-rc5+ kernel.
>
> I'm seeing this splat hit often.  I'm using my CT firmware and
> have a fair amount of ath10k patches, so it could be me.
>
>         WARN_ON(ar_pci->ps_wake_refcount > 0);
>
>
> Any idea what can cause this sort of thing?

You must be having unbalanced calls to
ath10k_pci_wake/ath10k_pci_sleep (i.e. more wake() than sleep() in
this case).


Michał