Method add

tests/python/relax/test_vm_cuda_graph.py:54–61

(A: T.Buffer((16, 16), "float32"), B: T.Buffer((16, 16), "float32"))

    @T.prim_func(s_tir=True)
    def add(A: T.Buffer((16, 16), "float32"), B: T.Buffer((16, 16), "float32")):
        T.func_attr({"global_symbol": "add"})
        with T.sblock("root"):
            for i in T.thread_binding(16, thread="threadIdx.x"):
                for j in range(16):
                    with T.sblock("update"):
                        vi, vj = T.axis.remap("SS", [i, j])
                        B[vi, vj] = A[vi, vj] + T.float32(1)

    @R.function
    def cuda_graph_alloc() -> R.Tuple(R.Object, R.Object):