162 	int i = 0;  in store_ymm()  local
163 	__asm__ volatile ("vmovaps  %%ymm0, %0" :"=m" (vec256array[i]));  in store_ymm()
164 	i++; __asm__ volatile ("vmovaps  %%ymm1, %0" :"=m" (vec256array[i]));  in store_ymm()
165 	i++; __asm__ volatile ("vmovaps  %%ymm2, %0" :"=m" (vec256array[i]));  in store_ymm()
166 	i++; __asm__ volatile ("vmovaps  %%ymm3, %0" :"=m" (vec256array[i]));  in store_ymm()
167 	i++; __asm__ volatile ("vmovaps  %%ymm4, %0" :"=m" (vec256array[i]));  in store_ymm()
168 	i++; __asm__ volatile ("vmovaps  %%ymm5, %0" :"=m" (vec256array[i]));  in store_ymm()
169 	i++; __asm__ volatile ("vmovaps  %%ymm6, %0" :"=m" (vec256array[i]));  in store_ymm()
170 	i++; __asm__ volatile ("vmovaps  %%ymm7, %0" :"=m" (vec256array[i]));  in store_ymm()
172 	i++; __asm__ volatile ("vmovaps  %%ymm8, %0" :"=m" (vec256array[i]));  in store_ymm()
173 	i++; __asm__ volatile ("vmovaps  %%ymm9, %0" :"=m" (vec256array[i]));  in store_ymm()
174 	i++; __asm__ volatile ("vmovaps  %%ymm10, %0" :"=m" (vec256array[i]));  in store_ymm()
175 	i++; __asm__ volatile ("vmovaps  %%ymm11, %0" :"=m" (vec256array[i]));  in store_ymm()
176 	i++; __asm__ volatile ("vmovaps  %%ymm12, %0" :"=m" (vec256array[i]));  in store_ymm()
177 	i++; __asm__ volatile ("vmovaps  %%ymm13, %0" :"=m" (vec256array[i]));  in store_ymm()
178 	i++; __asm__ volatile ("vmovaps  %%ymm14, %0" :"=m" (vec256array[i]));  in store_ymm()
179 	i++; __asm__ volatile ("vmovaps  %%ymm15, %0" :"=m" (vec256array[i]));  in store_ymm()
298 	int     i;  in copy_ymm_state_to_vector()  local
302 	for (i = 0; i < YMM_MAX; i++) {  in copy_ymm_state_to_vector()
303 		bcopy(&xmm[i], &vp[i], sizeof(*xmm));  in copy_ymm_state_to_vector()
304 		bcopy(&ymmh[i], (void *) ((uint64_t)&vp[i] + sizeof(*ymmh)), sizeof(*ymmh));  in copy_ymm_state_to_vector()
447 	int i = 0;  in store_zmm()  local
448 	__asm__ volatile ("vmovaps  %%zmm0, %0" :"=m" (vecarray[i]));  in store_zmm()
449 	i++; __asm__ volatile ("vmovaps  %%zmm1, %0" :"=m" (vecarray[i]));  in store_zmm()
450 	i++; __asm__ volatile ("vmovaps  %%zmm2, %0" :"=m" (vecarray[i]));  in store_zmm()
451 	i++; __asm__ volatile ("vmovaps  %%zmm3, %0" :"=m" (vecarray[i]));  in store_zmm()
452 	i++; __asm__ volatile ("vmovaps  %%zmm4, %0" :"=m" (vecarray[i]));  in store_zmm()
453 	i++; __asm__ volatile ("vmovaps  %%zmm5, %0" :"=m" (vecarray[i]));  in store_zmm()
454 	i++; __asm__ volatile ("vmovaps  %%zmm6, %0" :"=m" (vecarray[i]));  in store_zmm()
455 	i++; __asm__ volatile ("vmovaps  %%zmm7, %0" :"=m" (vecarray[i]));  in store_zmm()
457 	i++; __asm__ volatile ("vmovaps  %%zmm8, %0" :"=m" (vecarray[i]));  in store_zmm()
458 	i++; __asm__ volatile ("vmovaps  %%zmm9, %0" :"=m" (vecarray[i]));  in store_zmm()
459 	i++; __asm__ volatile ("vmovaps  %%zmm10, %0" :"=m" (vecarray[i]));  in store_zmm()
460 	i++; __asm__ volatile ("vmovaps  %%zmm11, %0" :"=m" (vecarray[i]));  in store_zmm()
461 	i++; __asm__ volatile ("vmovaps  %%zmm12, %0" :"=m" (vecarray[i]));  in store_zmm()
462 	i++; __asm__ volatile ("vmovaps  %%zmm13, %0" :"=m" (vecarray[i]));  in store_zmm()
463 	i++; __asm__ volatile ("vmovaps  %%zmm14, %0" :"=m" (vecarray[i]));  in store_zmm()
464 	i++; __asm__ volatile ("vmovaps  %%zmm15, %0" :"=m" (vecarray[i]));  in store_zmm()
465 	i++; __asm__ volatile ("vmovaps  %%zmm16, %0" :"=m" (vecarray[i]));  in store_zmm()
466 	i++; __asm__ volatile ("vmovaps  %%zmm17, %0" :"=m" (vecarray[i]));  in store_zmm()
467 	i++; __asm__ volatile ("vmovaps  %%zmm18, %0" :"=m" (vecarray[i]));  in store_zmm()
468 	i++; __asm__ volatile ("vmovaps  %%zmm19, %0" :"=m" (vecarray[i]));  in store_zmm()
469 	i++; __asm__ volatile ("vmovaps  %%zmm20, %0" :"=m" (vecarray[i]));  in store_zmm()
470 	i++; __asm__ volatile ("vmovaps  %%zmm21, %0" :"=m" (vecarray[i]));  in store_zmm()
471 	i++; __asm__ volatile ("vmovaps  %%zmm22, %0" :"=m" (vecarray[i]));  in store_zmm()
472 	i++; __asm__ volatile ("vmovaps  %%zmm23, %0" :"=m" (vecarray[i]));  in store_zmm()
473 	i++; __asm__ volatile ("vmovaps  %%zmm24, %0" :"=m" (vecarray[i]));  in store_zmm()
474 	i++; __asm__ volatile ("vmovaps  %%zmm25, %0" :"=m" (vecarray[i]));  in store_zmm()
475 	i++; __asm__ volatile ("vmovaps  %%zmm26, %0" :"=m" (vecarray[i]));  in store_zmm()
476 	i++; __asm__ volatile ("vmovaps  %%zmm27, %0" :"=m" (vecarray[i]));  in store_zmm()
477 	i++; __asm__ volatile ("vmovaps  %%zmm28, %0" :"=m" (vecarray[i]));  in store_zmm()
478 	i++; __asm__ volatile ("vmovaps  %%zmm29, %0" :"=m" (vecarray[i]));  in store_zmm()
479 	i++; __asm__ volatile ("vmovaps  %%zmm30, %0" :"=m" (vecarray[i]));  in store_zmm()
480 	i++; __asm__ volatile ("vmovaps  %%zmm31, %0" :"=m" (vecarray[i]));  in store_zmm()
759 	for (int i = 0; i < KARRAY_MAX; i++) {  in assert_opmask_eq()  local
760 		if (a[i] != b[i]) {  in assert_opmask_eq()
788 	for (int i = 0; i < KARRAY_MAX; i++) {  in copy_state_to_opmask()  local
789 		bcopy(&k[i], &op[i], sizeof(*op));  in copy_state_to_opmask()
796 	int     i;  in copy_zmm_state_to_vector()  local
803 	for (i = 0; i < ZMM_MAX / 2; i++) {  in copy_zmm_state_to_vector()
804 		bcopy(&xmm[i], &vp[i], sizeof(*xmm));  in copy_zmm_state_to_vector()
805 		bcopy(&ymmh[i], (void *) ((uint64_t)&vp[i] + sizeof(*ymmh)), sizeof(*ymmh));  in copy_zmm_state_to_vector()
806 		bcopy(&zmmh[i], (void *) ((uint64_t)&vp[i] + sizeof(*zmmh)), sizeof(*zmmh));  in copy_zmm_state_to_vector()
807 		bcopy(&zmm[i], &vp[(ZMM_MAX / 2) + i], sizeof(*zmm));  in copy_zmm_state_to_vector()
810 	for (i = 0; i < ZMM_MAX; i++) {  in copy_zmm_state_to_vector()
811 		bcopy(&xmm[i], &vp[i], sizeof(*xmm));  in copy_zmm_state_to_vector()
812 		bcopy(&ymmh[i], (void *) ((uint64_t)&vp[i] + sizeof(*ymmh)), sizeof(*ymmh));  in copy_zmm_state_to_vector()
813 		bcopy(&zmmh[i], (void *) ((uint64_t)&vp[i] + sizeof(*zmmh)), sizeof(*zmmh));  in copy_zmm_state_to_vector()