Re: [PATCH net-next v2] declance: Remove IRQF_ONESHOT

[Sebastian Andrzej Siewior <bigeasy@linutronix.de>]

On 2026-05-06 10:25:02 [+0100], Maciej W. Rozycki wrote:
> On Tue, 5 May 2026, Sebastian Andrzej Siewior wrote:
> 
> > I'm not if sure if you may need to change the primary handler if the
> > interrupt flow is EOI and cascading based on what you wrote. If you have
> > access to the HW then you it should be easy to test given the
> > `threadirqs' argument should expose problems.
> 
>  The interrupt is exceedingly rare, I've only seen it actually fire maybe 
> a dozen times across all my systems in 25+ years.  It happens when there 
> is a memory read error on DMA, such as an uncorrected ECC or parity error 
> (depending on the system variant), or a bus timeout.

I assumed you have other interrupts on that hw, cascaded/ operating the
same way. But otherwise…

>  It should be possible to orchestrate it, such as by making the LANCE DMA 
> pointer register refer an unpopulated location in the system address map; 
> memory ECC errors can be induced too by the DRAM controller's diagnostic 
> feature.  It seems enough hassle though I'd rather get things right by the 
> spec.

Oh, yeah. That should do it.

>  Thanks for the hint as to the `threadirqs' facility though, it may come 
> up helpful sometime.

:)

>   Maciej

Sebastian