Re: [PATCH v4 15/17] gpu: nova-core: service GSP events from the SWGEN0 interrupt

[Alexandre Courbot]

On Mon Sep 21, 2026 at 3:46 PM JST, Alexandre Courbot wrote:
<...>
>> @@ -380,12 +406,7 @@ pub(crate) fn new<'a>(
>>  
>>                  bar,
>>  
>> -                gsp_falcon: Falcon::new(
>> -                    dev,
>> -                    spec.chipset,
>> -                    bar
>> -                )
>> -                .inspect(|falcon| falcon.clear_swgen0_intr())?,
>> +                gsp_falcon: Falcon::new(dev, spec.chipset, bar)?,
>>  
>>                  sec2_falcon: Falcon::new(dev, spec.chipset, bar)?,
>>  
>> @@ -409,6 +430,30 @@ pub(crate) fn new<'a>(
>>                  })?,
>>              }),
>>  
>> +            _: {
>> +                irq::gsp::quiesce(bar, gsp_resources.spec.chipset, vectors_ref)?;

Another problem with this: `quiesce` calls `reset_tree` which then calls
`rearm_pci_irq`, which re-enables the TOP bit of the subtree. So we
effectively have TOP enabled before the handler is created, which
contradicts what the documentation says we should do.