Re: [PATCH v3 4/7] drm: nova: Add GETPARAM parameter to read the GPU chipid

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] Conditional compilation error on 32-bit architectures due to mismatched match arm types.
--

--- Patch [4]: [PATCH v3 4/7] drm: nova: Add GETPARAM parameter to read the GPU chipid ---
commit c560fdc21fe8a1d71109ee27e0b91d35e7283f2a
Author: Alistair Popple <apopple@nvidia.com>

drm: nova: Add GETPARAM parameter to read the GPU chipid

One of the first things a user needs to know about a GPU is its chipid,
so add a GETPARAM parameter to return that.

> diff --git a/drivers/gpu/drm/nova/file.rs b/drivers/gpu/drm/nova/file.rs
> index 1156df51c533a..855f238777349 100644
> --- a/drivers/gpu/drm/nova/file.rs
> +++ b/drivers/gpu/drm/nova/file.rs
[ ... ]
> @@ -43,6 +43,7 @@ pub(crate) fn get_param(
>  
>          let value = match getparam.param as u32 {
>              uapi::NOVA_GETPARAM_VRAM_BAR_SIZE => pdev.resource_len(1)?,
> +            uapi::NOVA_GETPARAM_GPU_CHIPID => reg_data.api.chipset() as u64,

[Severity: Low]
Does this cause a conditional compilation error on 32-bit architectures?

In File::get_param(), all match arms must evaluate to exactly the same
type. The first arm evaluates to pdev.resource_len(1)? which returns a
bindings::resource_size_t. The newly added second arm evaluates to u64
via the explicit cast.

On 32-bit kernel configurations without CONFIG_PHYS_ADDR_T_64BIT enabled,
resource_size_t resolves to u32. This mismatch between u32 and u64 across
the match arms will result in a hard build error.

>              _ => return Err(EINVAL),
>          };
>

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260723063046.1265791-1-apopple@nvidia.com?part=4