25   data.t_id_z = threadIdx.z;  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__()
52   data.t_id_z = threadIdx.z;  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__()
79   data.t_id_z = threadIdx.z;  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__()
109   data.t_id_z = threadIdx.z;  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__()
136   data.t_id_z = threadIdx.z;  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__()
163   data.t_id_z = threadIdx.z;  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__()
193   data.t_id_z = threadIdx.z;  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__()