Retain typed Rust callbacks through prepared MIR by tpn · Pull Request #1575 · NVIDIA/cuda-rust

tpn · 2026-10-20T18:42:01Z

Native C++ functors need retained Rust callbacks and their captured state. Callback admission now runs in rustc-free prepared MIR before provider compilation. The thin compiler bridge carries genuine function, signature and source facts; lowering retains the typed thunk and passes its Rust-owned environment through the native call.

The synchronous ABI supports i32 binary callbacks and immutable numeric scalar captures. Borrowed or pointer captures and unsupported transitive CUDA calls remain rejected. This update integrates Rust 1.99.1 and preserves the parent's borrowed-wrapper identity and storage-mixing rules.

Owning host tests and strict checks pass. All 18 original admission outcomes match, with six recorded diagnostic differences preserved. Three ordinary-Git variants pass 324 GPU reductions and seven sanitizer runs. Their 22 mapped executable sections, typed ABI, launch facts and resources match the accepted callback implementation; whole artifacts differ. Fresh HTTPS builds reproduce those 22 executable sections, ABI, facts and resources; the GPU runs used the ordinary-Git binaries.

Builds on #1574. Review the 22-file callback change above #1574. GitHub's Files tab remains cumulative against main: narrow existing-number stacks require maintainer-created upstream parent branches or prerequisite landing. This PR remains draft.

With `lto = "fat"` or `lto = "thin"`, the build of a binary that uses a library with kernels stopped with an error: "failed to get bitcode from object file for LTO (Can't find section .llvmbc)". rustc keeps a list of the Rust objects in each rlib. This list contains the objects of all modules. With LTO, rustc reads bitcode from these objects and removes them from the link. The artifact objects have no bitcode. If a session writes an rlib, the backend now gives each artifact object to rustc as the DWARF object of its module. This object is in the rlib, but it is not in the list. LTO does not read this object, and the linker gets it from the rlib. The system linker does not get the DWARF objects. If the session also writes a binary or a dynamic library, the backend gives the artifact objects to this linker as link arguments. With `-Zlink-native-libraries=no`, rustc does not give link arguments to the linker. Then the artifact objects of this session are Rust objects, as before. The book said that `join_codegen()` and `link()` only call the LLVM backend. Now it tells what they add to the outputs. CI builds `cuda_module_in_lib` with fat LTO. The library is an rlib, and then an rlib and a cdylib. After each build, CI makes sure that the binary has the PTX bundles of the library and of the binary. It also makes sure that the cdylib has the PTX bundle of the library. Fixes NVIDIA#1495 Signed-off-by: Mikhail Borisov <mikhail.boriso2014@gmail.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Use Cargo target names for standalone emit-ltoir artifacts. Bind the compiler and reader to the same intermediate directory when invoked from a Cargo workspace member. Signed-off-by: Trent Nelson <trent@trent.me>

Accept ordered external LTO files when materializing embedded cubins. Include their content in the codegen fingerprint and verify owned bytes before linking through the existing finalizer. Preserve inputs that overlap generated output paths and diagnose unresolved device symbols. Add a standalone CUDA FFI example and a native regression script. Signed-off-by: Trent Nelson <trent@trent.me>

Reuse existing example dependency pins and classify external LTO as an FFI fixture. Run its linking regression in native sweeps and verify relocatable debug output in the existing full-debug route. Signed-off-by: Trent Nelson <trent@trent.me>

Preserve exact geometry and fixed dynamic shared-memory requirements through artifact loading so preparation can select and validate the compiled allocation. Keep artifact writers and readers on one in-tree implementation while preserving legacy bundles without launch facts. Signed-off-by: Trent Nelson <trent@trent.me>

Combine the reviewed compiler prerequisites for native library calls. Keep the external LTO fixture on the shared in-tree artifact implementation. Signed-off-by: Trent Nelson <trent@trent.me>

Use the reviewed Cargo target and output-directory correction in the combined compiler prerequisite tree. Signed-off-by: Trent Nelson <trent@trent.me>

Resolve concrete registration functions and scalar argument types during device compilation, then invoke the native provider and link its owned LTO output through the ordinary artifact finalizer. Signed-off-by: Trent Nelson <trent@trent.me>