master-pending: ODEBUG: free active (active state 0) object type: timer_list hint: ath10k_htt_rx_ring_refill_retry+0x0/0x10 [ath10k_core]

[Valo, Kalle]

Ben Greear <greearb at candelatech.com> writes:

> On 08/18/2016 08:09 AM, Valo, Kalle wrote:
>> Hi,
>>
>> I was testing master-pending branch and saw these warnings when I was
>> reloading ath10k. I have seen this two times out of two tests so seems
>> to be easy to reproduce. I don't have time to investigate this more at
>> this time but I'll post this anyway.
>
> I'm not sure I am seeing the same issue, but this patch below helps
> keep me from crashing when firmware dies. I'm not sure the spinlock
> change matters, but the txq_unref fixes the crash....

This didn't apply for me but I see that you submitted them as proper
patches later. I'll test those when I start debugging this NAPI issue
more.

-- 
Kalle Valo