Lines Matching refs:CUDA
1 //===----- CGCUDANV.cpp - Interface to NVIDIA CUDA Runtime ----------------===//
9 // This provides a class for CUDA code generation targeting the NVIDIA CUDA
251 // CUDA 9.0+ uses new way to launch kernels. Parameters are packed in a local
385 /// Creates a function that sets up state on the host side for CUDA objects that
387 /// the host side of kernel functions and device global variables with the CUDA
444 // For HIP or CUDA 9.0+, device variable size is type of `size_t`.
515 /// For CUDA:
534 bool IsCUDA = CGM.getLangOpts().CUDA;
558 // Register GPU binary with the CUDA runtime, store returned handle in a
561 // handle so CUDA runtime can figure out what to call on the GPU side.
633 // For CUDA, create a string literal containing the fat binary loaded from
709 // Register binary with CUDA runtime. This is substantially different in
726 // Call __cudaRegisterFatBinaryEnd(Handle) if this CUDA version needs it.
763 // it during regular destructor phase worked in CUDA before 9.2 but results in
782 /// For CUDA: