Re: [PATCH v3 02/11] gpu: nova-core: move GSP boot code out of `Gpu` constructor

["Alexandre Courbot" <acourbot@nvidia.com>]

On Wed Sep 10, 2025 at 8:18 PM JST, Alexandre Courbot wrote:
>>> here is what it looks like when I got it to compile:
>>
>> This looks great!
>>
>>>     pub(crate) fn new<'a>(
>>>         pdev: &'a pci::Device<device::Bound>,
>>>         devres_bar: Arc<Devres<Bar0>>,
>>>         bar: &'a Bar0,
>>>     ) -> impl PinInit<Self, Error> + 'a {
>>>         try_pin_init!(Self {
>>>             spec: Spec::new(bar).inspect(|spec| {
>>>                 dev_info!(
>>>                     pdev.as_ref(),
>>>                     "NVIDIA (Chipset: {}, Architecture: {:?}, Revision: {})\n",
>>>                     spec.chipset,
>>>                     spec.chipset.arch(),
>>>                     spec.revision
>>>                 );
>>>             })?,
>>
>> +             _: {
>> +                 gfw::wait_gfw_boot_completion(bar)
>> +                     .inspect_err(|_| dev_err!(pdev.as_ref(), "GFW boot did not complete"))?;
>> +             },
>>>
>>>             sysmem_flush: SysmemFlush::register(pdev.as_ref(), bar, spec.chipset)?,
>>>
>>>             gsp_falcon: Falcon::<Gsp>::new(
>>>                 pdev.as_ref(),
>>>                 spec.chipset,
>>>                 bar,
>>>                 spec.chipset > Chipset::GA100,
>>>             )
>>>             .inspect(|falcon| falcon.clear_swgen0_intr(bar))?,
>>>
>>>             sec2_falcon: Falcon::<Sec2>::new(pdev.as_ref(), spec.chipset, bar, true)?,
>>>
>> -             gsp: Self::start_gsp(pdev, bar, spec.chipset, gsp_falcon, sec2_falcon)?,
>> +             gsp <- Self::start_gsp(pdev, bar, spec.chipset, gsp_falcon, sec2_falcon),
>>>
>>>             bar: devres_bar,
>>>         })
>>>     }
>>>
>>> The wait for GFW initialization had to be moved to `probe`, but that's
>>> fine IMO.
>>
>> That's not necessary, you can keep it in Gpu::new() -- I don't see what's
>> preventing us from that. I inserted it in the code above.
>
> This one didn't work for me, but maybe you have a newer version of the
> internal references patch:
>
> error: no rules expected reserved identifier `_`
>     --> drivers/gpu/nova-core/gpu.rs:323:13
>      |
> 323  |             _: {
>      |             ^ no rules expected this token in macro call
>      |
> note: while trying to match `)`

Ah, just found out about this, neat!

https://lore.kernel.org/rust-for-linux/20250905140534.3328297-1-lossin@kernel.org/