Searched refs:vw01234567 (Results 1 – 5 of 5) sorted by relevance
/external/XNNPACK/src/f16-prelu/gen/ |
D | neonfp16arith-2x16.c | 48 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; in xnn_f16_prelu_ukernel__neonfp16arith_2x16() local 56 float16x8_t vacc0x001234567 = vmulq_f16(vi0x001234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x16() 60 float16x8_t vacc1x001234567 = vmulq_f16(vi1x001234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x16() 76 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; in xnn_f16_prelu_ukernel__neonfp16arith_2x16() local 83 float16x8_t vacc0x01234567 = vmulq_f16(vi0x01234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x16() 85 float16x8_t vacc1x01234567 = vmulq_f16(vi1x01234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x16() 95 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; in xnn_f16_prelu_ukernel__neonfp16arith_2x16() local 102 float16x8_t vacc0x01234567 = vmulq_f16(vi0x01234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x16() 104 float16x8_t vacc1x01234567 = vmulq_f16(vi1x01234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x16()
|
D | neonfp16arith-2x8.c | 48 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; in xnn_f16_prelu_ukernel__neonfp16arith_2x8() local 53 float16x8_t vacc0x001234567 = vmulq_f16(vi0x001234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x8() 55 float16x8_t vacc1x001234567 = vmulq_f16(vi1x001234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x8() 65 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; in xnn_f16_prelu_ukernel__neonfp16arith_2x8() local 72 float16x8_t vacc0x01234567 = vmulq_f16(vi0x01234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x8() 74 float16x8_t vacc1x01234567 = vmulq_f16(vi1x01234567, vw01234567); in xnn_f16_prelu_ukernel__neonfp16arith_2x8()
|
/external/XNNPACK/src/f32-prelu/gen/ |
D | avx-2x8.c | 49 const __m256 vw01234567 = _mm256_load_ps(w); in xnn_f32_prelu_ukernel__avx_2x8() local 57 const __m256 vprod0x01234567 = _mm256_mul_ps(vi0x01234567, vw01234567); in xnn_f32_prelu_ukernel__avx_2x8() 58 const __m256 vprod1x01234567 = _mm256_mul_ps(vi1x01234567, vw01234567); in xnn_f32_prelu_ukernel__avx_2x8()
|
D | avx-2x16.c | 49 const __m256 vw01234567 = _mm256_load_ps(w); in xnn_f32_prelu_ukernel__avx_2x16() local 60 const __m256 vprod0x01234567 = _mm256_mul_ps(vi0x01234567, vw01234567); in xnn_f32_prelu_ukernel__avx_2x16() 62 const __m256 vprod1x01234567 = _mm256_mul_ps(vi1x01234567, vw01234567); in xnn_f32_prelu_ukernel__avx_2x16()
|
/external/XNNPACK/src/f16-prelu/ |
D | neonfp16arith.c.in | 77 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; variable 84 float16x8_t vacc${M}x01234567 = vmulq_f16(vi${M}x01234567, vw01234567); 94 const float16x8_t vw01234567 = vld1q_f16(w); w += 8; variable 101 float16x8_t vacc${M}x01234567 = vmulq_f16(vi${M}x01234567, vw01234567);
|