/external/XNNPACK/src/u8-rmax/ |
D | sse2.c | 26 const __m128i vx = _mm_loadu_si128((const __m128i*) x); in xnn_u8_rmax_ukernel__sse2() local 34 const __m128i vx = _mm_loadu_si128((const __m128i*) x); in xnn_u8_rmax_ukernel__sse2() local 45 const uint8_t vx = *x++; in xnn_u8_rmax_ukernel__sse2() local
|
D | neon.c | 26 const uint8x16_t vx = vld1q_u8(x); x += 16; in xnn_u8_rmax_ukernel__neon() local 33 const uint8x16_t vx = vld1q_u8(x); in xnn_u8_rmax_ukernel__neon() local 44 const uint8x8_t vx = vld1_dup_u8(x); x += 1; in xnn_u8_rmax_ukernel__neon() local
|
/external/XNNPACK/src/x32-zip/ |
D | x2-sse2.c | 26 const __m128i vx = _mm_loadu_si128((const __m128i*) x); in xnn_x32_zip_x2_ukernel__sse2() local 39 const __m128i vx = _mm_loadl_epi64((const __m128i*) x); in xnn_x32_zip_x2_ukernel__sse2() local 48 const uint32_t vx = *x; in xnn_x32_zip_x2_ukernel__sse2() local
|
D | x2-wasmsimd.c | 26 const v128_t vx = wasm_v128_load(x); in xnn_x32_zip_x2_ukernel__wasmsimd() local 39 const double vx = *((const double*) x); in xnn_x32_zip_x2_ukernel__wasmsimd() local 48 const float vx = *x; in xnn_x32_zip_x2_ukernel__wasmsimd() local
|
D | x3-sse2.c | 28 const __m128 vx = _mm_loadu_ps(x); in xnn_x32_zip_x3_ukernel__sse2() local 60 const __m128 vx = _mm_castpd_ps(_mm_load_sd((const double*) x)); in xnn_x32_zip_x3_ukernel__sse2() local 81 const __m128 vx = _mm_load_ss(x); in xnn_x32_zip_x3_ukernel__sse2() local
|
/external/XNNPACK/src/x8-zip/ |
D | x2-sse2.c | 25 const __m128i vx = _mm_loadu_si128((const __m128i*) x); in xnn_x8_zip_x2_ukernel__sse2() local 38 const __m128i vx = _mm_loadu_si128((const __m128i*) ((uintptr_t) x + address_increment)); in xnn_x8_zip_x2_ukernel__sse2() local 48 const uint8_t vx = *x++; in xnn_x8_zip_x2_ukernel__sse2() local
|
/external/XNNPACK/src/f32-vunary/ |
D | sse.c.in | 60 const __m128 vx = _mm_loadu_ps(x); variable 67 const __m128 vx = _mm_loadu_ps(x); variable
|
D | wasmsimd.c.in | 60 const v128_t vx = wasm_v128_load(x); variable 69 const v128_t vx = wasm_v128_load(x); variable
|
D | neon.c.in | 51 const float32x4_t vx = vld1q_f32(x); x += 4; variable 56 const float32x4_t vx = vld1q_f32(x); variable
|
D | avx.c.in | 61 const __m256 vx = _mm256_loadu_ps(x); variable 72 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
D | scalar.c.in | 55 const float vx = *x++; variable 67 const float vx = *x++; variable
|
/external/XNNPACK/src/f32-vrnd/ |
D | neonv8.c.in | 46 const float32x4_t vx = vld1q_f32(x); x += 4; variable 51 const float32x4_t vx = vld1q_f32(x); variable
|
D | sse41.c.in | 50 const __m128 vx = _mm_loadu_ps(x); variable 59 const __m128 vx = _mm_loadu_ps(x); variable
|
D | avx512f.c.in | 51 const __m512 vx = _mm512_loadu_ps(x); variable 66 const __m512 vx = _mm512_maskz_loadu_ps(vmask, x); variable
|
D | avx.c.in | 52 const __m256 vx = _mm256_loadu_ps(x); variable 65 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
D | scalar-libm.c.in | 48 const float vx = *x++; variable 60 const float vx = *x++; variable
|
/external/XNNPACK/src/f32-vsqrt/ |
D | wasmsimd-sqrt.c.in | 43 const v128_t vx = wasm_v128_load(x); variable 51 const float vx = *x++; variable
|
D | sse-sqrt.c.in | 42 const __m128 vx = _mm_loadu_ps(x); variable 50 const __m128 vx = _mm_load_ss(x++); variable
|
D | neon-sqrt.c.in | 39 const float32x4_t vx = vld1q_f32(x); x += 4; variable 45 const float vx = *x++; variable
|
D | avx-sqrt.c.in | 44 const __m256 vx = _mm256_loadu_ps(x); variable 55 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
/external/XNNPACK/src/f32-vlrelu/ |
D | wasmsimd-minmax.c.in | 48 v128_t vx = wasm_v128_load(x); variable 57 v128_t vx = wasm_v128_load(x); variable
|
D | neon.c.in | 44 const float32x4_t vx = vld1q_f32(x); x += 4; variable 51 const float32x4_t vx = vld1q_f32(x); variable
|
D | avx.c.in | 48 const __m256 vx = _mm256_loadu_ps(x); variable 60 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
/external/XNNPACK/src/f32-hswish/ |
D | wasmsimd.c.in | 57 v128_t vx = wasm_v128_load(x); variable 70 v128_t vx = wasm_v128_load(x); variable
|
D | avx.c.in | 66 const __m256 vx = _mm256_loadu_ps(x); variable 84 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|