Home | History | Annotate | Download | only in clang-include

Lines Matching defs:__a

44 _mm_add_ss(__m128 __a, __m128 __b)
46 __a[0] += __b[0];
47 return __a;
51 _mm_add_ps(__m128 __a, __m128 __b)
53 return __a + __b;
57 _mm_sub_ss(__m128 __a, __m128 __b)
59 __a[0] -= __b[0];
60 return __a;
64 _mm_sub_ps(__m128 __a, __m128 __b)
66 return __a - __b;
70 _mm_mul_ss(__m128 __a, __m128 __b)
72 __a[0] *= __b[0];
73 return __a;
77 _mm_mul_ps(__m128 __a, __m128 __b)
79 return __a * __b;
83 _mm_div_ss(__m128 __a, __m128 __b)
85 __a[0] /= __b[0];
86 return __a;
90 _mm_div_ps(__m128 __a, __m128 __b)
92 return __a / __b;
96 _mm_sqrt_ss(__m128 __a)
98 __m128 __c = __builtin_ia32_sqrtss(__a);
99 return (__m128) { __c[0], __a[1], __a[2], __a[3] };
103 _mm_sqrt_ps(__m128 __a)
105 return __builtin_ia32_sqrtps(__a);
109 _mm_rcp_ss(__m128 __a)
111 __m128 __c = __builtin_ia32_rcpss(__a);
112 return (__m128) { __c[0], __a[1], __a[2], __a[3] };
116 _mm_rcp_ps(__m128 __a)
118 return __builtin_ia32_rcpps(__a);
122 _mm_rsqrt_ss(__m128 __a)
124 __m128 __c = __builtin_ia32_rsqrtss(__a);
125 return (__m128) { __c[0], __a[1], __a[2], __a[3] };
129 _mm_rsqrt_ps(__m128 __a)
131 return __builtin_ia32_rsqrtps(__a);
135 _mm_min_ss(__m128 __a, __m128 __b)
137 return __builtin_ia32_minss(__a, __b);
141 _mm_min_ps(__m128 __a, __m128 __b)
143 return __builtin_ia32_minps(__a, __b);
147 _mm_max_ss(__m128 __a, __m128 __b)
149 return __builtin_ia32_maxss(__a, __b);
153 _mm_max_ps(__m128 __a, __m128 __b)
155 return __builtin_ia32_maxps(__a, __b);
159 _mm_and_ps(__m128 __a, __m128 __b)
161 return (__m128)((__v4si)__a & (__v4si)__b);
165 _mm_andnot_ps(__m128 __a, __m128 __b)
167 return (__m128)(~(__v4si)__a & (__v4si)__b);
171 _mm_or_ps(__m128 __a, __m128 __b)
173 return (__m128)((__v4si)__a | (__v4si)__b);
177 _mm_xor_ps(__m128 __a, __m128 __b)
179 return (__m128)((__v4si)__a ^ (__v4si)__b);
183 _mm_cmpeq_ss(__m128 __a, __m128 __b)
185 return (__m128)__builtin_ia32_cmpss(__a, __b, 0);
189 _mm_cmpeq_ps(__m128 __a, __m128 __b)
191 return (__m128)__builtin_ia32_cmpps(__a, __b, 0);
195 _mm_cmplt_ss(__m128 __a, __m128 __b)
197 return (__m128)__builtin_ia32_cmpss(__a, __b, 1);
201 _mm_cmplt_ps(__m128 __a, __m128 __b)
203 return (__m128)__builtin_ia32_cmpps(__a, __b, 1);
207 _mm_cmple_ss(__m128 __a, __m128 __b)
209 return (__m128)__builtin_ia32_cmpss(__a, __b, 2);
213 _mm_cmple_ps(__m128 __a, __m128 __b)
215 return (__m128)__builtin_ia32_cmpps(__a, __b, 2);
219 _mm_cmpgt_ss(__m128 __a, __m128 __b)
221 return (__m128)__builtin_ia32_cmpss(__b, __a, 1);
225 _mm_cmpgt_ps(__m128 __a, __m128 __b)
227 return (__m128)__builtin_ia32_cmpps(__b, __a, 1);
231 _mm_cmpge_ss(__m128 __a, __m128 __b)
233 return (__m128)__builtin_ia32_cmpss(__b, __a, 2);
237 _mm_cmpge_ps(__m128 __a, __m128 __b)
239 return (__m128)__builtin_ia32_cmpps(__b, __a, 2);
243 _mm_cmpneq_ss(__m128 __a, __m128 __b)
245 return (__m128)__builtin_ia32_cmpss(__a, __b, 4);
249 _mm_cmpneq_ps(__m128 __a, __m128 __b)
251 return (__m128)__builtin_ia32_cmpps(__a, __b, 4);
255 _mm_cmpnlt_ss(__m128 __a, __m128 __b)
257 return (__m128)__builtin_ia32_cmpss(__a, __b, 5);
261 _mm_cmpnlt_ps(__m128 __a, __m128 __b)
263 return (__m128)__builtin_ia32_cmpps(__a, __b, 5);
267 _mm_cmpnle_ss(__m128 __a, __m128 __b)
269 return (__m128)__builtin_ia32_cmpss(__a, __b, 6);
273 _mm_cmpnle_ps(__m128 __a, __m128 __b)
275 return (__m128)__builtin_ia32_cmpps(__a, __b, 6);
279 _mm_cmpngt_ss(__m128 __a, __m128 __b)
281 return (__m128)__builtin_ia32_cmpss(__b, __a, 5);
285 _mm_cmpngt_ps(__m128 __a, __m128 __b)
287 return (__m128)__builtin_ia32_cmpps(__b, __a, 5);
291 _mm_cmpnge_ss(__m128 __a, __m128 __b)
293 return (__m128)__builtin_ia32_cmpss(__b, __a, 6);
297 _mm_cmpnge_ps(__m128 __a, __m128 __b)
299 return (__m128)__builtin_ia32_cmpps(__b, __a, 6);
303 _mm_cmpord_ss(__m128 __a, __m128 __b)
305 return (__m128)__builtin_ia32_cmpss(__a, __b, 7);
309 _mm_cmpord_ps(__m128 __a, __m128 __b)
311 return (__m128)__builtin_ia32_cmpps(__a, __b, 7);
315 _mm_cmpunord_ss(__m128 __a, __m128 __b)
317 return (__m128)__builtin_ia32_cmpss(__a, __b, 3);
321 _mm_cmpunord_ps(__m128 __a, __m128 __b)
323 return (__m128)__builtin_ia32_cmpps(__a, __b, 3);
327 _mm_comieq_ss(__m128 __a, __m128 __b)
329 return __builtin_ia32_comieq(__a, __b);
333 _mm_comilt_ss(__m128 __a, __m128 __b)
335 return __builtin_ia32_comilt(__a, __b);
339 _mm_comile_ss(__m128 __a, __m128 __b)
341 return __builtin_ia32_comile(__a, __b);
345 _mm_comigt_ss(__m128 __a, __m128 __b)
347 return __builtin_ia32_comigt(__a, __b);
351 _mm_comige_ss(__m128 __a, __m128 __b)
353 return __builtin_ia32_comige(__a, __b);
357 _mm_comineq_ss(__m128 __a, __m128 __b)
359 return __builtin_ia32_comineq(__a, __b);
363 _mm_ucomieq_ss(__m128 __a, __m128 __b)
365 return __builtin_ia32_ucomieq(__a, __b);
369 _mm_ucomilt_ss(__m128 __a, __m128 __b)
371 return __builtin_ia32_ucomilt(__a, __b);
375 _mm_ucomile_ss(__m128 __a, __m128 __b)
377 return __builtin_ia32_ucomile(__a, __b);
381 _mm_ucomigt_ss(__m128 __a, __m128 __b)
383 return __builtin_ia32_ucomigt(__a, __b);
387 _mm_ucomige_ss(__m128 __a, __m128 __b)
389 return __builtin_ia32_ucomige(__a, __b);
393 _mm_ucomineq_ss(__m128 __a, __m128 __b)
395 return __builtin_ia32_ucomineq(__a, __b);
399 _mm_cvtss_si32(__m128 __a)
401 return __builtin_ia32_cvtss2si(__a);
405 _mm_cvt_ss2si(__m128 __a)
407 return _mm_cvtss_si32(__a);
413 _mm_cvtss_si64(__m128 __a)
415 return __builtin_ia32_cvtss2si64(__a);
421 _mm_cvtps_pi32(__m128 __a)
423 return (__m64)__builtin_ia32_cvtps2pi(__a);
427 _mm_cvt_ps2pi(__m128 __a)
429 return _mm_cvtps_pi32(__a);
433 _mm_cvttss_si32(__m128 __a)
435 return __a[0];
439 _mm_cvtt_ss2si(__m128 __a)
441 return _mm_cvttss_si32(__a);
445 _mm_cvttss_si64(__m128 __a)
447 return __a[0];
451 _mm_cvttps_pi32(__m128 __a)
453 return (__m64)__builtin_ia32_cvttps2pi(__a);
457 _mm_cvtt_ps2pi(__m128 __a)
459 return _mm_cvttps_pi32(__a);
463 _mm_cvtsi32_ss(__m128 __a, int __b)
465 __a[0] = __b;
466 return __a;
470 _mm_cvt_si2ss(__m128 __a, int __b)
472 return _mm_cvtsi32_ss(__a, __b);
478 _mm_cvtsi64_ss(__m128 __a, long long __b)
480 __a[0] = __b;
481 return __a;
487 _mm_cvtpi32_ps(__m128 __a, __m64 __b)
489 return __builtin_ia32_cvtpi2ps(__a, (__v2si)__b);
493 _mm_cvt_pi2ps(__m128 __a, __m64 __b)
495 return _mm_cvtpi32_ps(__a, __b);
499 _mm_cvtss_f32(__m128 __a)
501 return __a[0];
505 _mm_loadh_pi(__m128 __a, const __m64 *__p)
513 return __builtin_shufflevector(__a, __bb, 0, 1, 4, 5);
517 _mm_loadl_pi(__m128 __a, const __m64 *__p)
525 return __builtin_shufflevector(__a, __bb, 4, 5, 2, 3);
568 __m128 __a = _mm_load_ps(__p);
569 return __builtin_shufflevector(__a, __a, 3, 2, 1, 0);
610 _mm_storeh_pi(__m64 *__p, __m128 __a)
612 __builtin_ia32_storehps((__v2si *)__p, __a);
616 _mm_storel_pi(__m64 *__p, __m128 __a)
618 __builtin_ia32_storelps((__v2si *)__p, __a);
622 _mm_store_ss(float *__p, __m128 __a)
627 ((struct __mm_store_ss_struct*)__p)->__u = __a[0];
631 _mm_storeu_ps(float *__p, __m128 __a)
633 __builtin_ia32_storeups(__p, __a);
637 _mm_store1_ps(float *__p, __m128 __a)
639 __a = __builtin_shufflevector(__a, __a, 0, 0, 0, 0);
640 _mm_storeu_ps(__p, __a);
644 _mm_store_ps1(float *__p, __m128 __a)
646 return _mm_store1_ps(__p, __a);
650 _mm_store_ps(float *__p, __m128 __a)
652 *(__m128 *)__p = __a;
656 _mm_storer_ps(float *__p, __m128 __a)
658 __a = __builtin_shufflevector(__a, __a, 3, 2, 1, 0);
659 _mm_store_ps(__p, __a);
673 _mm_stream_pi(__m64 *__p, __m64 __a)
675 __builtin_ia32_movntq(__p, __a);
679 _mm_stream_ps(float *__p, __m128 __a)
681 __builtin_ia32_movntps(__p, __a);
691 _mm_extract_pi16(__m64 __a, int __n)
693 __v4hi __b = (__v4hi)__a;
698 _mm_insert_pi16(__m64 __a, int __d, int __n)
700 __v4hi __b = (__v4hi)__a;
706 _mm_max_pi16(__m64 __a, __m64 __b)
708 return (__m64)__builtin_ia32_pmaxsw((__v4hi)__a, (__v4hi)__b);
712 _mm_max_pu8(__m64 __a, __m64 __b)
714 return (__m64)__builtin_ia32_pmaxub((__v8qi)__a, (__v8qi)__b);
718 _mm_min_pi16(__m64 __a, __m64 __b)
720 return (__m64)__builtin_ia32_pminsw((__v4hi)__a, (__v4hi)__b);
724 _mm_min_pu8(__m64 __a, __m64 __b)
726 return (__m64)__builtin_ia32_pminub((__v8qi)__a, (__v8qi)__b);
730 _mm_movemask_pi8(__m64 __a)
732 return __builtin_ia32_pmovmskb((__v8qi)__a);
736 _mm_mulhi_pu16(__m64 __a, __m64 __b)
738 return (__m64)__builtin_ia32_pmulhuw((__v4hi)__a, (__v4hi)__b);
742 __m64 __a = (a); \
743 (__m64)__builtin_ia32_pshufw((__v4hi)__a, (n)); })
752 _mm_avg_pu8(__m64 __a, __m64 __b)
754 return (__m64)__builtin_ia32_pavgb((__v8qi)__a, (__v8qi)__b);
758 _mm_avg_pu16(__m64 __a, __m64 __b)
760 return (__m64)__builtin_ia32_pavgw((__v4hi)__a, (__v4hi)__b);
764 _mm_sad_pu8(__m64 __a, __m64 __b)
766 return (__m64)__builtin_ia32_psadbw((__v8qi)__a, (__v8qi)__b);
782 __m128 __a = (a); \
784 (__m128)__builtin_shufflevector((__v4sf)__a, (__v4sf)__b, \
790 _mm_unpackhi_ps(__m128 __a, __m128 __b)
792 return __builtin_shufflevector(__a, __b, 2, 6, 3, 7);
796 _mm_unpacklo_ps(__m128 __a, __m128 __b)
798 return __builtin_shufflevector(__a, __b, 0, 4, 1, 5);
802 _mm_move_ss(__m128 __a, __m128 __b)
804 return __builtin_shufflevector(__a, __b, 4, 1, 2, 3);
808 _mm_movehl_ps(__m128 __a, __m128 __b)
810 return __builtin_shufflevector(__a, __b, 6, 7, 2, 3);
814 _mm_movelh_ps(__m128 __a, __m128 __b)
816 return __builtin_shufflevector(__a, __b, 0, 1, 4, 5);
820 _mm_cvtpi16_ps(__m64 __a)
826 __b = _mm_cmpgt_pi16(__b, __a);
827 __c = _mm_unpackhi_pi16(__a, __b);
831 __c = _mm_unpacklo_pi16(__a, __b);
838 _mm_cvtpu16_ps(__m64 __a)
844 __c = _mm_unpackhi_pi16(__a, __b);
848 __c = _mm_unpacklo_pi16(__a, __b);
855 _mm_cvtpi8_ps(__m64 __a)
860 __b = _mm_cmpgt_pi8(__b, __a);
861 __b = _mm_unpacklo_pi8(__a, __b);
867 _mm_cvtpu8_ps(__m64 __a)
872 __b = _mm_unpacklo_pi8(__a, __b);
878 _mm_cvtpi32x2_ps(__m64 __a, __m64 __b)
886 return _mm_cvtpi32_ps(__c, __a);
890 _mm_cvtps_pi16(__m128 __a)
894 __b = _mm_cvtps_pi32(__a);
895 __a = _mm_movehl_ps(__a, __a);
896 __c = _mm_cvtps_pi32(__a);
902 _mm_cvtps_pi8(__m128 __a)
906 __b = _mm_cvtps_pi16(__a);
913 _mm_movemask_ps(__m128 __a)
915 return __builtin_ia32_movmskps(__a);