Function `offload`

pub const fn offload<F, T: Tuple, R>(f: F, workgroup_dim: [u32; 3], thread_dim: [u32; 3], dyn_cache: u32, device_id: i32, args: T) -> R

Generates the LLVM body of a wrapper function to offload a kernel f.

Type Parameters:

F: The kernel to offload. Must be a function item.
T: A tuple of arguments passed to f.
R: The return type of the kernel.

Arguments:

f: The kernel function to offload.
workgroup_dim: A 3D size specifying the number of workgroups to launch.
thread_dim: A 3D size specifying the number of threads per workgroup.
dyn_cache: The amount of dynamic shared memory to request for the kernel.
device_id: The device to offload to. Use -1 to select the default device.
args: A tuple of arguments forwarded to f.

Example usage (pseudocode):

fn kernel(x: *mut [f64; 128]) {
    core::intrinsics::offload(kernel_1, [256, 1, 1], [32, 1, 1], 0, -1, (x,))
}

#[cfg(target_os = "linux")]
extern "C" {
    pub fn kernel_1(array_b: *mut [f64; 128]);
}

#[cfg(not(target_os = "linux"))]
#[rustc_offload_kernel]
extern "gpu-kernel" fn kernel_1(x: *mut [f64; 128]) {
    unsafe { (*x)[0] = 21.0 };
}

For reference, see the Clang documentation on offloading: https://clang.llvm.org/docs/OffloadingDesign.html.