Tom Dohrmann

> > ... (thus the discriminant must also be explicitly annotated) > > Hmm, when I originally wrote the code for this PR, I made `core::ffi::c_int` the default if no...

The alternative would be to not assume the discriminant size and derive it by duplicating the type, removing the fields, but keeping any explicit discriminant values, and then looking at...

I updated the desugaring to transform enums with fields into fieldless enums (as described in the [reference](https://doc.rust-lang.org/stable/reference/type-layout.html#representations)) and use that to get the size of the discriminant.

Thanks for the review @fu5ha !

> My only hesitation left is the fact we're generating a non-hygenic `BlahDiscriminant` type for all `#[repr(C)]` enums now, but I don't immediately see a way to prevent that. We...

> Wait, hm. For the `CheckedBitPattern` case, could we not get rid of the `BlahRawDiscriminant` and just inline the current type expression into the `BlahBits` type instead? I think I...

> Which kernel should be used on the CI server? I uploaded some binaries here: https://github.com/Freax13/linux/releases/tag/snp-guest-req-v1b-tag

> There's a lot of: > > ``` > test cli::platform::info::test::test_info_json ... ok > source: Custom { > kind: ResourceBusy, > code: 16, > code: 16, > code: 16, >...

> [ 13.125727] kvm_amd: SEV-SNP enabled (ASIDs 1 - 14) Try increasing the number of ASIDs for SEV-ES/SEV-SNP in the BIOS. I think we're running out of ASIDs when running...

> The maximum ASIDs was already set, but I changed the minimum to 1 (it was higher when I was trying to figure out how to get the server to...