23   data.t_id_x = threadIdx.x;  in __launch_bounds__()
26   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
38 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()
50   data.t_id_x = threadIdx.x;  in __launch_bounds__()
53   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
65 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()
77   data.t_id_x = threadIdx.x;  in __launch_bounds__()
80   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
92 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()
107   data.t_id_x = threadIdx.x;  in __launch_bounds__()
110   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
122 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()
134   data.t_id_x = threadIdx.x;  in __launch_bounds__()
137   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
149 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()
161   data.t_id_x = threadIdx.x;  in __launch_bounds__()
164   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
176 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()
191   data.t_id_x = threadIdx.x;  in __launch_bounds__()
194   data.t_id   = threadIdx.x + threadIdx.y * blockDim.x + threadIdx.z * blockDim.y * blockDim.x;  in __launch_bounds__()
199 …for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * bl…  in __launch_bounds__()