Update rustc-dev-guide

Author: Sa4dUs

library/core/src/intrinsics/mod.rs

@@ -3329,6 +3329,9 @@ pub const fn autodiff<F, G, T: crate::marker::Tuple, R>(f: F, df: G, args: T) ->
 ///     unsafe { (*x)[0] = 21.0 };
 /// }
 /// ```
+///
+/// For reference, see the Clang documentation on offloading:
+/// <https://clang.llvm.org/docs/OffloadingDesign.html>.
 #[rustc_nounwind]
 #[rustc_intrinsic]
 pub const fn offload<F, T: crate::marker::Tuple, R>(f: F, args: T) -> R;

src/doc/rustc-dev-guide/src/offload/usage.md

@@ -5,13 +5,16 @@ We currently work on launching the following Rust kernel on the GPU. To follow a
 
 ```rust
 #![feature(abi_gpu_kernel)]
+#![feature(rustc_attrs)]
+#![feature(core_intrinsics)]
 #![no_std]
 
 #[cfg(target_os = "linux")]
 extern crate libc;
 #[cfg(target_os = "linux")]
 use libc::c_char;
 
+#[cfg(target_os = "linux")]
 use core::mem;
 
 #[panic_handler]
@@ -38,7 +41,7 @@ fn main() {
     }
 
     unsafe {
-        kernel_1(array_c);
+        kernel(array_c);
     }
     core::hint::black_box(&array_c);
     unsafe {
@@ -52,6 +55,11 @@ fn main() {
     }
 }
 
+#[inline(never)]
+unsafe fn kernel(x: *mut [f64; 256]) {
+    core::intrinsics::offload(kernel_1, (x,))
+}
+
 #[cfg(target_os = "linux")]
 unsafe extern "C" {
     pub fn kernel_1(array_b: *mut [f64; 256]);
@@ -60,6 +68,7 @@ unsafe extern "C" {
 #[cfg(not(target_os = "linux"))]
 #[unsafe(no_mangle)]
 #[inline(never)]
+#[rustc_offload_kernel]
 pub extern "gpu-kernel" fn kernel_1(x: *mut [f64; 256]) {
     unsafe { (*x)[0] = 21.0 };
 }
@@ -76,7 +85,7 @@ rustc +offload --edition 2024 src/lib.rs -g --crate-type cdylib -C opt-level=3 -
 
 Now we generate the device code. Replace the target-cpu with the right code for your gpu.
 ```
-RUSTFLAGS="-Ctarget-cpu=gfx90a --emit=llvm-bc,llvm-ir" cargo +offload build -Zunstable-options -r -v --target amdgcn-amd-amdhsa -Zbuild-std=core
+RUSTFLAGS="-Ctarget-cpu=gfx90a --emit=llvm-bc,llvm-ir -Zoffload=Enable -Zunstable-options" cargo +offload build -Zunstable-options -r -v --target amdgcn-amd-amdhsa -Zbuild-std=core
 ```
 
 Now find the `<libname>.ll` under target/amdgcn-amd-amdhsa folder and copy it to a device.ll file (or adjust the file names below).