Re: [PATCH] rust: io: convert ResourceSize into a transparent newtype

On Sun Jul 12, 2026 at 1:36 PM CEST, Lorenzo Delgado wrote:
> diff --git a/drivers/gpu/nova-core/firmware/gsp.rs b/drivers/gpu/nova-core/firmware/gsp.rs
> index 99a302bae567..d7593888e65b 100644
> --- a/drivers/gpu/nova-core/firmware/gsp.rs
> +++ b/drivers/gpu/nova-core/firmware/gsp.rs
> @@ -175,7 +175,7 @@ pub(crate) fn radix3_dma_handle(&self) -> DmaAddress {
>  fn map_into_lvl(sg_table: &SGTable<Owned<VVec<u8>>>, mut dst: VVec<u8>) -> Result<VVec<u8>> {
>      for sg_entry in sg_table.iter() {
>          // Number of pages we need to map.
> -        let num_pages = usize::from_safe_cast(sg_entry.dma_len()).div_ceil(GSP_PAGE_SIZE);
> +        let num_pages = usize::try_from(sg_entry.dma_len())?.div_ceil(GSP_PAGE_SIZE);

I think this is worse, as the conversion becomes fallible.

You could implement From<ResourceSize> for u64 and then keep using
usize::from_safe_cast() in nova-core.

Alternatively, we could also consider moving the FromSafeCast trait to
rust/kernel/num.rs and add FromSafeCast<ResourceSize> impls for usize.

I originally proposed the approach [1], since in nova-core we had a lot of
fallible conversion, while our Kconfig mandates that they can never actually
fail.

However, by making it commonly availble I do see a risk with the cfg-gated impls
silently breaking the build.

Now, technically that's the feature and 0-day etc. should usually catch it, but
on the other hand we also usually try to avoid cfg-gates in the kernel.

Miguel, I think you were in favor of this as well, what do you think?

[1] https://lore.kernel.org/rust-for-linux/[email protected]/