Re:Re: Re: [PATCH v6 11/13] serial: 8250_dw: add Ambarella CV75 quirks

[zl020895]

Hi Ilpo, Andy,

I'll go with the simple comment:
	/*
	 * Ambarella keeps variant IER bits (e.g. ETOI) set for correct RX
	 * timeout behaviour.
	 */
	up->ier |= d->pdata->ier_mask;

Best regards,
Long Zhao

At 2026-09-10 17:33:38, "Ilpo Järvinen" <ilpo.jarvinen@xxxxxxxxxxxxxxx> wrote:
>On Thu, 10 Sep 2026, Andy Shevchenko wrote:
>
>> On Thu, Sep 10, 2026 at 05:12:52PM +0800, zl020895 wrote:
>> 
>> > > Wondering if we should undo this in the port's .shutdown() callback.
>> > I saw the shutdown path already covers it: dw8250_shutdown() calls
>> > serial8250_do_shutdown(), which does
>> > 	up->ier = 0;
>> > 	serial_port_out(port, UART_IER, 0);
>> > so both the IER shadow and hardware IER are cleared, including ETOI.
>> 
>> Depending on the answer to Ilpo the enabling may need an updated comment
>> summarising the above.
>
>I'd prefer removing the entire "Force them ..." sentence, it feels really 
>basic detail about how ->ier works. I believe that would resolve your 
>concern if relates to what shutdown does for the flag (clears it while 
>the comment as is says nothing is supposed to clear it).
>
>> > At 2026-09-10 16:33:14, "Andy Shevchenko" <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
>> > >On Fri, Sep 04, 2026 at 02:38:18PM +0800, Long Zhao via B4 Relay wrote:
>> 
>> ...
>> 
>> > >> +	/*
>> > >> +	 * Ambarella keeps variant IER bits (e.g. ETOI) set for correct RX
>> > >> +	 * timeout behaviour. Force them into the 8250 IER shadow so later
>> > >> +	 * generic IER updates do not clear them.
>> > >> +	 */
>> > >
>> > >Wondering if we should undo this in the port's .shutdown() callback.
>> 
>> 
>
>-- 
> i.