Merge branch 'main' into no_default_as_host_vec

Author: LegNeato

.github/workflows/rust.yml

@@ -7,8 +7,6 @@ on:
   push:
     paths-ignore:
       - '**.md'
-    branches:
-      - master
 
 env:
   RUST_LOG: info
@@ -33,7 +31,7 @@ jobs:
         uses: actions/checkout@v2
 
       - name: Install CUDA 
-        uses: Jimver/[email protected].4
+        uses: Jimver/[email protected].21
         id: cuda-toolkit
         with:
           cuda: '11.2.2'
@@ -74,4 +72,4 @@ jobs:
       - name: Check documentation
         env:
           RUSTDOCFLAGS: -Dwarnings
-        run: cargo doc --workspace --all-features --document-private-items --no-deps --exclude "optix" --exclude "path_tracer" --exclude "denoiser" --exclude "add" --exclude "ex*"
+        run: cargo doc --workspace --all-features --document-private-items --no-deps --exclude "optix" --exclude "path_tracer" --exclude "denoiser" --exclude "add" --exclude "ex*"

Cargo.toml

@@ -1,4 +1,5 @@
 [workspace]
+resolver = "2"
 members = [
   "crates/*",
   "crates/optix/examples/ex*",

README.md

@@ -7,15 +7,22 @@
   </p>
 
   <h3>
-    <a href="guide">Guide</a>
+    <a href="https://rust-gpu.github.io/Rust-CUDA/guide/index.html">Guide</a>
     <span> | </span>
     <a href="guide/src/guide/getting_started.md">Getting Started</a>
     <span> | </span>
     <a href="guide/src/features.md">Features</a>
   </h3>
-<strong>⚠️ The project is still in early development, expect bugs, safety issues, and things that don't work ⚠️</strong>
+<strong>⚠️ The project is still in early development, expect bugs, safety issues, and things that don't work ⚠️</strong> 
 </div>
 
+<br/>
+
+> [!IMPORTANT]
+> This project is no longer dormant and is [being
+> rebooted](https://rust-gpu.github.io/blog/2025/01/27/rust-cuda-reboot).
+> Please contribute!
+
 ## Goal
 
 The Rust CUDA Project is a project aimed at making Rust a tier-1 language for extremely fast GPU computing
@@ -28,7 +35,7 @@ Historically, general purpose high performance GPU computing has been done using
 provides a way to use Fortran/C/C++ code for GPU computing in tandem with CPU code with a single source. It also provides
 many libraries, tools, forums, and documentation to supplement the single-source CPU/GPU code.
 
-CUDA is exclusively an NVIDIA-only toolkit. Many tools have been proposed for cross-platform GPU computing such as 
+CUDA is exclusively an NVIDIA-only toolkit. Many tools have been proposed for cross-platform GPU computing such as
 OpenCL, Vulkan Computing, and HIP. However, CUDA remains the most used toolkit for such tasks by far. This is why it is
 imperative to make Rust a viable option for use with the CUDA toolkit.
 
@@ -38,12 +45,12 @@ in recent years it has been shown time and time again that a specialized solutio
 of projects such as rust-gpu (for Rust -> SPIR-V).
 
 Our hope is that with this project we can push the Rust GPU computing industry forward and make Rust an excellent language
-for such tasks. Rust offers plenty of benefits such as `__restrict__` performance benefits for every kernel, An excellent module/crate system, 
+for such tasks. Rust offers plenty of benefits such as `__restrict__` performance benefits for every kernel, An excellent module/crate system,
 delimiting of unsafe areas of CPU/GPU code with `unsafe`, high level wrappers to low level CUDA libraries, etc.
 
 ## Structure
 
-The scope of the Rust CUDA Project is quite broad, it spans the entirety of the CUDA ecosystem, with libraries and tools to make it 
+The scope of the Rust CUDA Project is quite broad, it spans the entirety of the CUDA ecosystem, with libraries and tools to make it
 usable using Rust. Therefore, the project contains many crates for all corners of the CUDA ecosystem.
 
 The current line-up of libraries is the following:
@@ -52,7 +59,7 @@ The current line-up of libraries is the following:
   - Generates highly optimized PTX code which can be loaded by the CUDA Driver API to execute on the GPU.
   - For the near future it will be CUDA-only, but it may be used to target amdgpu in the future.
 - `cuda_std` for GPU-side functions and utilities, such as thread index queries, memory allocation, warp intrinsics, etc.
-  - *Not* a low level library, provides many utility functions to make it easier to write cleaner and more reliable GPU kernels.
+  - _Not_ a low level library, provides many utility functions to make it easier to write cleaner and more reliable GPU kernels.
   - Closely tied to `rustc_codegen_nvvm` which exposes GPU features through it internally.
 - [`cudnn`](https://github.com/Rust-GPU/Rust-CUDA/tree/master/crates/cudnn) for a collection of GPU-accelerated primitives for deep neural networks.
 - `cust` for CPU-side CUDA features such as launching GPU kernels, GPU memory allocation, device queries, etc.
@@ -67,12 +74,23 @@ In addition to many "glue" crates for things such as high level wrappers for cer
 ## Related Projects
 
 Other projects related to using Rust on the GPU:
+
 - 2016: [glassful](https://github.com/kmcallister/glassful) Subset of Rust that compiles to GLSL.
 - 2017: [inspirv-rust](https://github.com/msiglreith/inspirv-rust) Experimental Rust MIR -> SPIR-V Compiler.
 - 2018: [nvptx](https://github.com/japaric-archived/nvptx) Rust to PTX compiler using the `nvptx` target for rustc (using the LLVM PTX backend).
-- 2020: [accel](https://github.com/termoshtt/accel) Higher level library that relied on the same mechanism that `nvptx` does.
+- 2020: [accel](https://github.com/termoshtt/accel) Higher-level library that relied on the same mechanism that `nvptx` does.
 - 2020: [rlsl](https://github.com/MaikKlein/rlsl) Experimental Rust -> SPIR-V compiler (predecessor to rust-gpu)
-- 2020: [rust-gpu](https://github.com/EmbarkStudios/rust-gpu) Rustc codegen backend to compile Rust to SPIR-V for use in shaders, similar mechanism as our project.
+- 2020: [rust-gpu](https://github.com/Rust-GPU/rust-gpu) `rustc` compiler backend to compile Rust to SPIR-V for use in shaders, similar mechanism as our project.
+
+## Usage
+```bash
+## setup your environment like:
+### export OPTIX_ROOT=/opt/NVIDIA-OptiX-SDK-9.0.0-linux64-x86_64
+### export OPTIX_ROOT_DIR=/opt/NVIDIA-OptiX-SDK-9.0.0-linux64-x86_64
+
+## build proj
+cargo build
+```
 
 ## License
 
@@ -86,4 +104,3 @@ at your discretion.
 ### Contribution
 
 Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.
-

crates/blastoff/Cargo.toml

@@ -6,11 +6,11 @@ authors = ["Riccardo D'Ambrosio <[email protected]>"]
 repository = "https://github.com/Rust-GPU/Rust-CUDA"
 
 [dependencies]
-bitflags = "1.3.2"
+bitflags = "2.8"
 cublas_sys = { version = "0.1", path = "../cublas_sys" }
 cust = { version = "0.3", path = "../cust", features = ["impl_num_complex"] }
-num-complex = "0.4.0"
-half = { version = "1.8.0", optional = true }
+num-complex = "0.4.6"
+half = { version = "2.4.1", optional = true }
 
 [package.metadata.docs.rs]
 rustdoc-args = ["--html-in-header", "katex-header.html", "--cfg", "docsrs"]

crates/blastoff/src/context.rs

@@ -140,7 +140,13 @@ impl CublasContext {
     ) -> Result<T> {
         unsafe {
             // cudaStream_t is the same as CUstream
-            sys::v2::cublasSetStream_v2(self.raw, mem::transmute(stream.as_inner())).to_result()?;
+            sys::v2::cublasSetStream_v2(
+                self.raw,
+                mem::transmute::<*mut cust::sys::CUstream_st, *mut cublas_sys::v2::CUstream_st>(
+                    stream.as_inner(),
+                ),
+            )
+            .to_result()?;
             let res = func(self)?;
             // reset the stream back to NULL just in case someone calls with_stream, then drops the stream, and tries to
             // execute a raw sys function with the context's handle.
@@ -227,10 +233,11 @@ impl CublasContext {
     /// ```
     pub fn set_math_mode(&self, math_mode: MathMode) -> Result<()> {
         unsafe {
-            Ok(
-                sys::v2::cublasSetMathMode(self.raw, mem::transmute(math_mode.bits()))
-                    .to_result()?,
+            Ok(sys::v2::cublasSetMathMode(
+                self.raw,
+                mem::transmute::<u32, cublas_sys::v2::cublasMath_t>(math_mode.bits()),
             )
+            .to_result()?)
         }
     }
 

crates/blastoff/src/level1.rs

@@ -24,8 +24,8 @@ fn check_stride<T: BlasDatatype>(x: &impl GpuBuffer<T>, n: usize, stride: Option
     );
 }
 
-/// Scalar and Vector-based operations such as `min`, `max`, `axpy`, `copy`, `dot`, `nrm2`, `rot`, `rotg`, `rotm`, `rotmg`, `scal`, and `swap`.
-
+/// Scalar and Vector-based operations such as `min`, `max`, `axpy`, `copy`, `dot`,
+/// `nrm2`, `rot`, `rotg`, `rotm`, `rotmg`, `scal`, and `swap`.
 impl CublasContext {
     /// Same as [`CublasContext::amin`] but with an explicit stride.
     ///

crates/blastoff/src/lib.rs

@@ -93,22 +93,17 @@ pub(crate) mod private {
 
 /// An optional operation to apply to a matrix before a matrix operation. This includes
 /// no operation, transpose, or conjugate transpose.
-#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+#[derive(Debug, Default, Clone, Copy, PartialEq, Eq, Hash)]
 pub enum MatrixOp {
     /// No operation, leave the matrix as is. This is the default.
+    #[default]
     None,
     /// Transpose the matrix in place.
     Transpose,
     /// Conjugate transpose the matrix in place.
     ConjugateTranspose,
 }
 
-impl Default for MatrixOp {
-    fn default() -> Self {
-        MatrixOp::None
-    }
-}
-
 impl MatrixOp {
     /// Returns the corresponding `cublasOperation_t` for this operation.
     pub fn to_raw(self) -> sys::v2::cublasOperation_t {

crates/cuda_builder/Cargo.toml

@@ -11,6 +11,6 @@ readme = "../../README.md"
 [dependencies]
 rustc_codegen_nvvm = { version = "0.3", path = "../rustc_codegen_nvvm" }
 nvvm = { path = "../nvvm", version = "0.1" }
-serde = { version = "1.0.130", features = ["derive"] }
-serde_json = "1.0.68"
+serde = { version = "1.0.217", features = ["derive"] }
+serde_json = "1.0.138"
 find_cuda_helper = { version = "0.2", path = "../find_cuda_helper" }

crates/cuda_builder/src/lib.rs

@@ -246,13 +246,13 @@ impl CudaBuilder {
 
     /// Emit LLVM IR, the exact same as rustc's `--emit=llvm-ir`.
     pub fn emit_llvm_ir(mut self, emit_llvm_ir: bool) -> Self {
-        self.emit = emit_llvm_ir.then(|| EmitOption::LlvmIr);
+        self.emit = emit_llvm_ir.then_some(EmitOption::LlvmIr);
         self
     }
 
     /// Emit LLVM Bitcode, the exact same as rustc's `--emit=llvm-bc`.
     pub fn emit_llvm_bitcode(mut self, emit_llvm_bitcode: bool) -> Self {
-        self.emit = emit_llvm_bitcode.then(|| EmitOption::Bitcode);
+        self.emit = emit_llvm_bitcode.then_some(EmitOption::Bitcode);
         self
     }
 
@@ -376,10 +376,13 @@ fn invoke_rustc(builder: &CudaBuilder) -> Result<PathBuf, CudaBuilderError> {
 
     let new_path = get_new_path_var();
 
-    let mut rustflags = vec![format!(
-        "-Zcodegen-backend={}",
-        rustc_codegen_nvvm.display(),
-    )];
+    let mut rustflags = vec![
+        format!("-Zcodegen-backend={}", rustc_codegen_nvvm.display()),
+        "-Zcrate-attr=feature(register_tool)".into(),
+        "-Zcrate-attr=register_tool(nvvm_internal)".into(),
+        "-Zcrate-attr=no_std".into(),
+        "-Zsaturating_float_casts=false".into(),
+    ];
 
     if let Some(emit) = &builder.emit {
         let string = match emit {
@@ -432,13 +435,12 @@ fn invoke_rustc(builder: &CudaBuilder) -> Result<PathBuf, CudaBuilderError> {
     }
 
     let mut cargo = Command::new("cargo");
-    cargo.args(&[
+    cargo.args([
         "build",
         "--lib",
         "--message-format=json-render-diagnostics",
         "-Zbuild-std=core,alloc",
-        "--target",
-        "nvptx64-nvidia-cuda",
+        "--target=nvptx64-nvidia-cuda",
     ]);
 
     cargo.args(&builder.build_args);
@@ -523,7 +525,7 @@ fn get_last_artifact(out: &str) -> Option<PathBuf> {
             }
         })
         .filter(|line| line.reason == "compiler-artifact")
-        .last()
+        .next_back()
         .expect("Did not find output file in rustc output");
 
     let mut filenames = last

crates/cuda_std/Cargo.toml

@@ -8,8 +8,8 @@ repository = "https://github.com/Rust-GPU/Rust-CUDA"
 readme = "../../README.md"
 
 [dependencies]
-vek = { version = "0.15.1", default-features = false, features = ["libm"] }
+vek = { version = "0.17.1", default-features = false, features = ["libm"] }
 cuda_std_macros = { version = "0.2", path = "../cuda_std_macros" }
-half = "1.7.1"
-bitflags = "1.3.2"
-paste = "1.0.5"
+half = "2.4.1"
+bitflags = "2.8"
+paste = "1.0.15"