derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
57 lines
2.2 KiB
C
57 lines
2.2 KiB
C
// This file was automatically generated. Do not edit.
|
|
|
|
// We use a special strategy to force the generation of the PTX. This is mainly
|
|
// a fight against dead-code-elimination in the NVVM layer.
|
|
//
|
|
// The reason we need this strategy is because certain older versions of ptxas
|
|
// segfault when a non-sensical sequence of PTX is generated. So instead, we try
|
|
// to force the instantiation and compilation to PTX of all the overloads of the
|
|
// PTX wrapping functions.
|
|
//
|
|
// We do this by writing a function pointer of each overload to the kernel
|
|
// parameter `fn_ptr`.
|
|
//
|
|
// Because `fn_ptr` is possibly visible outside this translation unit, the
|
|
// compiler must compile all the functions which are stored.
|
|
|
|
__global__ void test_barrier_cluster(void** fn_ptr)
|
|
{
|
|
#if __cccl_ptx_isa >= 780
|
|
NV_IF_TARGET(NV_PROVIDES_SM_90,
|
|
(
|
|
// barrier.cluster.arrive;
|
|
* fn_ptr++ = reinterpret_cast<void*>(static_cast<void (*)()>(cuda::ptx::barrier_cluster_arrive));));
|
|
#endif // __cccl_ptx_isa >= 780
|
|
|
|
#if __cccl_ptx_isa >= 780
|
|
NV_IF_TARGET(NV_PROVIDES_SM_90,
|
|
(
|
|
// barrier.cluster.wait;
|
|
* fn_ptr++ = reinterpret_cast<void*>(static_cast<void (*)()>(cuda::ptx::barrier_cluster_wait));));
|
|
#endif // __cccl_ptx_isa >= 780
|
|
|
|
#if __cccl_ptx_isa >= 800
|
|
NV_IF_TARGET(NV_PROVIDES_SM_90,
|
|
(
|
|
// barrier.cluster.arrive.release;
|
|
* fn_ptr++ = reinterpret_cast<void*>(
|
|
static_cast<void (*)(cuda::ptx::sem_release_t)>(cuda::ptx::barrier_cluster_arrive));));
|
|
#endif // __cccl_ptx_isa >= 800
|
|
|
|
#if __cccl_ptx_isa >= 800
|
|
NV_IF_TARGET(NV_PROVIDES_SM_90,
|
|
(
|
|
// barrier.cluster.arrive.relaxed;
|
|
* fn_ptr++ = reinterpret_cast<void*>(
|
|
static_cast<void (*)(cuda::ptx::sem_relaxed_t)>(cuda::ptx::barrier_cluster_arrive));));
|
|
#endif // __cccl_ptx_isa >= 800
|
|
|
|
#if __cccl_ptx_isa >= 800
|
|
NV_IF_TARGET(NV_PROVIDES_SM_90,
|
|
(
|
|
// barrier.cluster.wait.acquire;
|
|
* fn_ptr++ = reinterpret_cast<void*>(
|
|
static_cast<void (*)(cuda::ptx::sem_acquire_t)>(cuda::ptx::barrier_cluster_wait));));
|
|
#endif // __cccl_ptx_isa >= 800
|
|
}
|