Re: [PATCH net-next 3/3] net: qrtr: implement shutdown()

[Chunkai Deng]

Thanks for pointing this out, you're right. That is actually what I had planned for the follow-up series. Since it is outside the scope of this one, I described it as "no way", which overstated it a bit. Thanks Chunkai Deng

On 10/26/2026 9:53 PM, Manivannan Sadhasivam wrote:

> On Mon, Oct 26, 2026 at 06:23:18PM +0800, Chunkai Deng wrote:
>> Yeah, this is a known limitation and the commit message calls it out.
>>
>> The sender sleeps on flow->resume_tx, which hangs off the node and is
>> keyed by destination node/port, so the socket has no way to reach it.
> It is fine to defer this fix as a follow-up patch, but I'm not buying the 'no
> way to reach flow control' from shutdown() callback claim. Why can't you just
> store the 'qtrt_tx_flow' pointer in qrtr_sock and dereference it in shutdown()
> callback to wake up qrtr_tx_wait()?
>
> Storing the pointer here is fine since all callers of qrtr_tx_wait() hold
> lock_sock().
>
> - Mani
>