Re: [PATCH 1/2] nova-core: Add a library for bitfields in Rust structs

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

On Thu Sep 4, 2025 at 4:16 PM JST, Danilo Krummrich wrote:
> On Thu Sep 4, 2025 at 5:16 AM CEST, Alexandre Courbot wrote:
>> On Thu Sep 4, 2025 at 12:15 AM JST, Joel Fernandes wrote:
>> <snip>
>>>>> +use kernel::prelude::*;
>>>>> +
>>>>> +/// Macro for defining bitfield-packed structures in Rust.
>>>>> +/// The size of the underlying storage type is specified with #[repr(TYPE)].
>>>>> +///
>>>>> +/// # Example (just for illustration)
>>>>> +/// ```rust
>>>>> +/// bitstruct! {
>>>>> +///     #[repr(u64)]
>>>>> +///     pub struct PageTableEntry {
>>>>> +///         0:0       present     as bool,
>>>>> +///         1:1       writable    as bool,
>>>>> +///         11:9      available   as u8,
>>>>> +///         51:12     pfn         as u64,
>>>>> +///         62:52     available2  as u16,
>>>>> +///         63:63     nx          as bool,
>>>> 
>>>> A note on syntax: for nova-core, we may want to use the `H:L` notation,
>>>> as this is what OpenRM uses, but in the larger kernel we might want to
>>>> use inclusive ranges (`L..=H`) as it will look more natural in Rust
>>>> code (and is the notation the `bits` module already uses).
>>>
>>> Perhaps future add-on enhancement to have both syntax? I'd like to initially
>>> keep H:L and stabilize the code first, what do you think?
>>
>> Let's have the discussion with the other stakeholders (Daniel?). I think
>> in Nova we want to keep the `H:L` syntax, as it matches what the OpenRM
>> headers do (so Nova would have its own `register` macro that calls into
>> the common one, tweaking things as it needs). But in the kernel crate we
>> should use something intuitive for everyone.
>
> I don't care too much about whether it's gonna be H:L or L:H [1], but I do care
> about being consistent throughout the kernel. Let's not start the practice of
> twisting kernel APIs to NV_* specific APIs that differ from what people are used
> to work with in the kernel.
>
> [1] If it's gonna be H:L, I think we should also list things in reverse order,
>     i.e.:
>
> 	pub struct PageTableEntry {
> 	    63:63     nx          as bool,
> 	    62:52     available2  as u16,
> 	    51:12     pfn         as u64,
> 	    11:9      available   as u8,
> 	    1:1       writable    as bool,
> 	    0:0       present     as bool,
> 	}
>
> This is also what would be my preferred style for the kernel in general.

Sorry for the confusion. The discussion was whether to keep using the
`H:L` syntax of the current macro, or use Rust's inclusive ranges syntax
(i.e. `L..=H`), as the `genmask_*` macros currently do.