TDX: Pull TLB flush ringbuf struct into support/, add tests #1387
Conversation
smalis-msft
added
the
backport_2505
| @@ -1817,6 +1817,18 @@ pub mod hypercall { | |||
| #[repr(transparent)] | |||
| pub struct HvGvaRange(pub u64); | |||
|
|
|||
| impl From<u64> for HvGvaRange { | |||
There was a problem hiding this comment.
So that the new support crate can return a T rather than a raw u64
There was a problem hiding this comment.
This seems error prone for other users of HvGvaRange.
There was a problem hiding this comment.
They could always do .0 or HvGvaRange(foo) before, this just works with generics.
There was a problem hiding this comment.
I just worry that things like gpa.into() are going to creep in, which will work as long as gpa doesn't have any low bits set... but will cause random failures if there are low bits set.
There was a problem hiding this comment.
That's fair, but From + Into definitely feels like the right API to be exposing from the ringbuf itself so that you don't lose the type of whatever you're storing, right? Unless you'd rather the callers had to do the translations, but that just feels annoying.
|
Looks like loom doesn't build for Windows-aarch64. |
|
Yeah, this |
The added tests use loom to ensure that all possible order of operations work correctly. This required tweaking the orderings we use. Can be reviewed commit-by-commit.