69 template <typename T, int P, int Q, int NB>
73   if (n != NB) {  in read_B_g2s_1D_nosync()
79     for (i = 0; i < Q * NB - P; i += P) {  in read_B_g2s_1D_nosync()
93 template <typename T, int P, int Q, int NB>
94 static __device__ __inline__ void write_C_r2g_1D_nosync(const int tx, const int n, T rC[NB], T *dC)…  in write_C_r2g_1D_nosync()  argument
95   if (n != NB) {  in write_C_r2g_1D_nosync()
101     for (int i = 0; i < NB; i++) {  in write_C_r2g_1D_nosync()
112 template <typename T, int P, int Q, int NB>
113 static __device__ __inline__ void sum_C_r2g_1D_nosync(const int tx, const int n, T rC[NB], T *dC) {  in sum_C_r2g_1D_nosync()  argument
114   if (n != NB) {  in sum_C_r2g_1D_nosync()
120     for (int i = 0; i < NB; i++) {  in sum_C_r2g_1D_nosync()
132 template <typename T, int P, int Q, int NB>
133 static __device__ __inline__ void mul_rAsBrC_1D_nosync(T rA[Q], T *sB, T rC[NB]) {  in mul_rAsBrC_1D_nosync()  argument
137   for (int i = 0; i < NB; i++) {  in mul_rAsBrC_1D_nosync()
156 template <typename T, int P, int Q, int NB>
157 static __device__ __inline__ void addmul_rAsBrC_1D_nosync(T rA[Q], T *sB, T rC[NB]) {  in addmul_rAsBrC_1D_nosync()  argument
161   for (int i = 0; i < NB; i++) {  in addmul_rAsBrC_1D_nosync()