Random fixes (#504)

This is a collection of random changes coming from the workload I'm working on. The most important change is better support for `.params`: PTX uses .param namespace both for some local variables and kernel args. This is a problem for us because those are different address spaces on AMDGPU. So far we've made an effort to convert to local and const namespaces whenever possible, but this commit tries to handle more patterns, which are impossible to track precisely, by converting to generic space.
This commit is contained in:
Andrzej Janik 2025-09-12 22:52:33 +02:00 committed by GitHub
commit 6c811a55d2
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
22 changed files with 282 additions and 64 deletions

View file

@ -313,7 +313,7 @@ fn join(
pub fn test_cuda(_attr: TokenStream, item: TokenStream) -> TokenStream {
let fn_ = parse_macro_input!(item as syn::ItemFn);
let cuda_fn = format_ident!("{}{}", fn_.sig.ident, "_nvidia");
let zluda_fn = format_ident!("{}{}", fn_.sig.ident, "_amdgpu");
let zluda_fn = format_ident!("{}{}", fn_.sig.ident, "_zluda");
let fn_name = fn_.sig.ident.clone();
quote! {
#[test]