pyramids.cpp | 312 __m128i v_r1 = _mm_packs_epi32(_mm_loadu_si128((__m128i const *)(row1 + x)), local 317 __m128i v_2r1 = _mm_adds_epu16(v_r1, v_r1), v_4r1 = _mm_adds_epu16(v_2r1, v_2r1); 319 __m128i v_dst10 = _mm_slli_epi16(_mm_adds_epu16(v_r1, v_r2), 2); 323 v_r1 = _mm_packs_epi32(_mm_loadu_si128((__m128i const *)(row1 + x + 8)), 328 v_2r1 = _mm_adds_epu16(v_r1, v_r1), v_4r1 = _mm_adds_epu16(v_2r1, v_2r1); 330 __m128i v_dst11 = _mm_slli_epi16(_mm_adds_epu16(v_r1, v_r2), 2); 342 __m128i v_r1 = _mm_packs_epi32(_mm_loadu_si128((__m128i const *)(row1 + x)), local 347 __m128i v_2r1 = _mm_adds_epu16(v_r1, v_r1), v_4r1 = _mm_adds_epu16(v_2r1, v_2r1) 375 v_r1 = _mm_loadu_si128((__m128i const *)(row1 + x)), local 400 v_r1 = _mm_loadu_si128((__m128i const *)(row1 + x)), local 435 v_r1 = _mm_loadu_si128((__m128i const *)(row1 + x)), local 460 v_r1 = _mm_loadu_si128((__m128i const *)(row1 + x)), local 500 __m128 v_r1 = _mm_loadu_ps(row1 + x); local 533 uint16x8_t v_r1 = vcombine_u16(vqmovn_u32(vld1q_u32(row1 + x)), vqmovn_u32(vld1q_u32(row1 + x + 4))); local 638 float32x4_t v_r1 = vld1q_f32(row1 + x); local 674 uint16x8_t v_r1 = vcombine_u16(vqmovn_u32(vld1q_u32(row1 + x)), vqmovn_u32(vld1q_u32(row1 + x + 4))); local 698 uint16x8_t v_r1 = vcombine_u16(vqmovn_u32(vld1q_u32(row1 + x)), vqmovn_u32(vld1q_u32(row1 + x + 4))); local 724 uint32x4_t v_r0 = vld1q_u32(row0 + x), v_r1 = vld1q_u32(row1 + x), v_r2 = vld1q_u32(row2 + x); local 745 uint32x4_t v_r0 = vld1q_u32(row0 + x), v_r1 = vld1q_u32(row1 + x), v_r2 = vld1q_u32(row2 + x); local 770 int32x4_t v_r0 = vld1q_s32(row0 + x), v_r1 = vld1q_s32(row1 + x), v_r2 = vld1q_s32(row2 + x); local 791 int32x4_t v_r0 = vld1q_s32(row0 + x), v_r1 = vld1q_s32(row1 + x), v_r2 = vld1q_s32(row2 + x); local 817 float32x4_t v_r1 = vld1q_f32(row1 + x); local [all...] |