Allow explicit data transfers to GPUs

[ZuseZ4]

View all comments

So far we had our offload intrinsics handle data movement automatically to/from the gpu.
That's convenient (and reasonably fast once our LLVM opts land). However, Rust generally also allows being explicit. That might give perf benefits (where our LLVM opts fail), and it could also be nice for modelling, when passing data around but still preventing CPU users from accesing it.

• Tracking:
  Tracking Issue for GPU-offload #131513

[ZuseZ4]

Vendoring llvm/llvm-project#198033 for now.

[oli-obk]

why is this not just an intrinsic call here?

View changes since the review

[ZuseZ4]

Partly just experimenting, partly because intrinsics recently changed a bit, they got updated for more explicit Place handling, about which I didn't want to think for my mvp. I'll update them to intrinsics after my deadline.

[oli-obk]

Yea I think these should just be intrinsics instead of catching lang item calls during codegen of call terminators.

View changes since the review

[ZuseZ4]

"Which lang items"? (my code who fails to catch an inlined terminator call :D)

[ZuseZ4]

ok, with an intrinsic it actually seems to work in release.
Not sure if we want one intrinsic with 2 arguments (mut/const, init/drop) or 4 intrinsics. Right now I have two.

[rust-log-analyzer]

The job tidy failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

fmt check
Diff in /checkout/compiler/rustc_codegen_llvm/src/builder/gpu_helper.rs:1:
-use crate::SimpleCx;
-use crate::builder::Builder;
-use crate::intrinsic::TransferType;
-use crate::llvm;
-use crate::llvm::{Type, Value};
 use rustc_abi::Align;
 use rustc_codegen_ssa::MemFlags;
 use rustc_codegen_ssa::common::TypeKind;
Diff in /checkout/compiler/rustc_codegen_llvm/src/builder/gpu_helper.rs:9:
 use rustc_codegen_ssa::traits::{BaseTypeCodegenMethods, BuilderMethods};
 use rustc_middle::bug;
 use rustc_middle::ty::offload_meta::{OffloadMetadata, OffloadSize};
+
+use crate::builder::Builder;
+use crate::intrinsic::TransferType;
+use crate::llvm::{Type, Value};
+use crate::{SimpleCx, llvm};
 
 pub(crate) fn scalar_width<'ll>(cx: &'ll SimpleCx<'_>, ty: &'ll Type) -> u64 {
     match cx.type_kind(ty) {
Diff in /checkout/library/core/src/offload/mod.rs:1:
 // offload module
 #[unstable(feature = "gpu_offload", issue = "131513")]
 pub use crate::macros::builtin::offload_kernel;
+use crate::marker::PhantomData;
 #[unstable(feature = "gpu_offload", issue = "131513")]
 pub use crate::offload;
-
-use crate::marker::PhantomData;
 
 // We store a raw pointer instead of a reference, since the real location of the data will be on a
 // GPU, at a different address. We only use the CPU pointer as a key to our runtime cpu-gpu pointer
fmt: checked 6902 files
Bootstrap failed while executing `test src/tools/tidy tidyselftest --extra-checks=py,cpp,js,spellcheck`
Build completed unsuccessfully in 0:00:48
  local time: Wed Jun 10 22:41:33 UTC 2026
  network time: Wed, 10 Jun 2026 22:41:33 GMT

[rust-bors]

☔ The latest upstream changes (presumably #158416) made this pull request unmergeable. Please resolve the merge conflicts by rebasing.