Lines Matching refs:vi

67       // vi${M}x3012 = ( vi${M}2, vi${M}1, vi{M}0, vi{M}3 )
68       __m128 vi${M}x3012 = _mm_setzero_ps();
71       __m128 vi${M}x4567 = _mm_loadu_ps(i${M});
77         // vi${M}x89AB = ( vi${M}B, vi${M}A, vi${M}9, vi${M}8 )
78         const __m128 vi${M}x89AB = _mm_loadu_ps(i${M});
82         // vi${M}x7456 = ( vi${M}6, vi${M}5, vi${M}4, vi${M}7 )
83 …      const __m128 vi${M}x7456 = _mm_shuffle_ps(vi${M}x4567, vi${M}x4567, _MM_SHUFFLE(2, 1, 0, 3));
88             __m128 vo${M}p0 = _mm_add_ps(vbias, _mm_mul_ps(vi${M+K}x4567, vk${K}1));
90             __m128 vo${M}p${K} = _mm_mul_ps(vi${M+K}x4567, vk${K}1);
92 …vo${M}p${K % ACCUMULATORS} = _mm_add_ps(vo${M}p${K % ACCUMULATORS}, _mm_mul_ps(vi${M+K}x4567, vk${…
95         // vi${M}x3456 = ( vi${M}6, vi${M}5, vi${M}4, vi${M}3 )
96         const __m128 vi${M}x3456 = _mm_move_ss(vi${M}x7456, vi${M}x3012);
101             __m128 vo${M}p${K+3} = _mm_mul_ps(vi${M+K}x3456, vk${K}0);
103 …3) % ACCUMULATORS} = _mm_add_ps(vo${M}p${(K+3) % ACCUMULATORS}, _mm_mul_ps(vi${M+K}x3456, vk${K}0)…
106         vi${M}x3012 = vi${M}x7456;
109         // vi${M}x8567 = ( vi${M}7, vi${M}6, vi${M}5, vi${M}8 )
110         const __m128 vi${M}x8567 = _mm_move_ss(vi${M}x4567, vi${M}x89AB);
113         // vi${M}x5678 = ( vi${M}8, vi${M}7, vi${M}6, vi${M}5 )
114 …      const __m128 vi${M}x5678 = _mm_shuffle_ps(vi${M}x8567, vi${M}x8567, _MM_SHUFFLE(0, 3, 2, 1));
118 …vo${M}p${(K+6) % ACCUMULATORS} = _mm_add_ps(vo${M}p${(K+6) % ACCUMULATORS}, _mm_mul_ps(vi${M+K}x56…
121         vi${M}x4567 = vi${M}x89AB;
147         vi${M}x4567 = _mm_and_ps(vmask, vi${M}x4567);
150         // vi${M}x7456 = ( vi${M}6, vi${M}5, vi${M}4, vi${M}7 )
151 …      const __m128 vi${M}x7456 = _mm_shuffle_ps(vi${M}x4567, vi${M}x4567, _MM_SHUFFLE(2, 1, 0, 3));
156             __m128 vo${M}p0 = _mm_add_ps(vbias, _mm_mul_ps(vi${M+K}x4567, vk${K}1));
158             __m128 vo${M}p${K} = _mm_mul_ps(vi${M+K}x4567, vk${K}1);
160 …vo${M}p${K % ACCUMULATORS} = _mm_add_ps(vo${M}p${K % ACCUMULATORS}, _mm_mul_ps(vi${M+K}x4567, vk${…
163         // vi${M}x3456 = ( vi${M}6, vi${M}5, vi${M}4, vi${M}3 )
164         const __m128 vi${M}x3456 = _mm_move_ss(vi${M}x7456, vi${M}x3012);
169             __m128 vo${M}p${K+3} = _mm_mul_ps(vi${M+K}x3456, vk${K}0);
171 …3) % ACCUMULATORS} = _mm_add_ps(vo${M}p${(K+3) % ACCUMULATORS}, _mm_mul_ps(vi${M+K}x3456, vk${K}0)…
175         // vi${M}x8567 = ( vi${M}7, vi${M}6, vi${M}5, 0.0 )
176         const __m128 vi${M}x8567 = _mm_move_ss(vi${M}x4567, vzero);
179         // vi${M}x5678 = ( vi${M}8, vi${M}7, vi${M}6, vi${M}5 )
180 …      const __m128 vi${M}x5678 = _mm_shuffle_ps(vi${M}x8567, vi${M}x8567, _MM_SHUFFLE(0, 3, 2, 1));
184 …vo${M}p${(K+6) % ACCUMULATORS} = _mm_add_ps(vo${M}p${(K+6) % ACCUMULATORS}, _mm_mul_ps(vi${M+K}x56…