Re: [PATCH v2 3/3] rust: pci: provide access to PCI Vendor values

[John Hubbard <jhubbard@nvidia.com>]

On 8/18/25 9:06 AM, Danilo Krummrich wrote:
> On Mon Aug 18, 2025 at 3:33 AM CEST, John Hubbard wrote:
>> +            /// Create a `Vendor` from the raw vendor ID value, or `None` if the value doesn't
>> +            /// match any known vendor.
>> +            pub fn from_u32(value: u32) -> Option<Self> {
>> +                match value {
>> +                    $(x if x == Self::$variant.0 => Some(Self::$variant),)+
>> +                    _ => None,
>> +                }
>> +            }
> 
> Same here, I think this should be `impl TryFrom<u32> for Vendor`.

OK. 

> 
>> +
>> +            /// Get the raw 16-bit vendor ID value.
>> +            pub const fn as_u32(self) -> u32 {
>> +                self.0
>> +            }
>> +        }
>> +    };
>> +}
> 
>>  /// An adapter for the registration of PCI drivers.
>>  pub struct Adapter<T: Driver>(T);
>>  
>> @@ -335,9 +656,9 @@ pub const fn from_class(class: u32, class_mask: u32) -> Self {
>>      ///
>>      /// This is more targeted than [`DeviceId::from_class`]: in addition to matching by Vendor, it
>>      /// also matches the PCI Class (up to the entire 24 bits, depending on the mask).
>> -    pub const fn from_class_and_vendor(class: Class, class_mask: u32, vendor: u32) -> Self {
>> +    pub const fn from_class_and_vendor(class: Class, class_mask: u32, vendor: Vendor) -> Self {
>>          Self(bindings::pci_device_id {
>> -            vendor,
>> +            vendor: vendor.as_u32(),
>>              device: DeviceId::PCI_ANY_ID,
>>              subvendor: DeviceId::PCI_ANY_ID,
>>              subdevice: DeviceId::PCI_ANY_ID,
>> @@ -396,7 +717,7 @@ macro_rules! pci_device_table {
>>  ///     <MyDriver as pci::Driver>::IdInfo,
>>  ///     [
>>  ///         (
>> -///             pci::DeviceId::from_id(bindings::PCI_VENDOR_ID_REDHAT, bindings::PCI_ANY_ID as u32),
>> +///             pci::DeviceId::from_id(pci::Vendor::REDHAT.as_u32(), bindings::PCI_ANY_ID as u32),
> 
> We should change DeviceId::from_id() to consume a pci::Vendor value directly.

Interesting. OK.

Thanks for the review!

thanks,
-- 
John Hubbard