| /external/pytorch/aten/src/ATen/native/quantized/cpu/qnnpack/src/u8rmax/ |
| D | sse2.c | 22 const __m128i vx = _mm_loadu_si128((const __m128i*)x); in pytorch_u8rmax_ukernel__sse2() local
30 const __m128i vx = _mm_loadu_si128((const __m128i*)x); in pytorch_u8rmax_ukernel__sse2() local
42 const uint8_t vx = *x++; in pytorch_u8rmax_ukernel__sse2() local
|
| D | neon.c | 22 const uint8x16_t vx = vld1q_u8(x); in pytorch_u8rmax_ukernel__neon() local
30 const uint8x16_t vx = vld1q_u8(x); in pytorch_u8rmax_ukernel__neon() local
42 const uint8x8_t vx = vld1_dup_u8(x); in pytorch_u8rmax_ukernel__neon() local
|
| /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
|
| /external/XNNPACK/src/qs8-f32-vcvt/ |
| D | scalar.c.in | 50 int32_t vx = *x++; variable
61 int32_t vx = *x; variable
71 int32_t vx = *x++; variable
|
| /external/pytorch/aten/src/ATen/native/quantized/cpu/qnnpack/src/x8zip/ |
| D | x2-sse2.c | 20 const __m128i vx = _mm_loadu_si128((const __m128i*)x); in pytorch_qnnp_x8zip_x2__sse2() local
33 const __m128i vx = in pytorch_qnnp_x8zip_x2__sse2() local
45 const uint8_t vx = *x++; in pytorch_qnnp_x8zip_x2__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-vrnd/ |
| D | wasmsimd.c.in | 46 const v128_t vx = wasm_v128_load(x); x += 4; variable
51 const v128_t vx = wasm_v128_load(x); variable
|
| 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 | 50 const __m256 vx = _mm256_loadu_ps(x); variable
63 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
| /external/XNNPACK/src/f32-vunary/ |
| 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 | sse.c.in | 60 const __m128 vx = _mm_loadu_ps(x); variable
67 const __m128 vx = _mm_loadu_ps(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 | 59 const __m256 vx = _mm256_loadu_ps(x); variable
70 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
| /external/XNNPACK/src/qs8-f32-vcvt/gen/ |
| D | vcvt-avx512skx-x64.c | 36 __m512i vx = _mm512_cvtepi8_epi32(_mm_loadu_si128((const __m128i*) (x + 48))); in xnn_qs8_f32_vcvt_ukernel__avx512skx_x64() local
61 __m512i vx = _mm512_cvtepi8_epi32(_mm_loadu_si128((const __m128i*) x)); in xnn_qs8_f32_vcvt_ukernel__avx512skx_x64() local
78 __m512i vx = _mm512_cvtepi8_epi32(_mm_maskz_loadu_epi8(vmask, x)); in xnn_qs8_f32_vcvt_ukernel__avx512skx_x64() local
|
| /external/XNNPACK/src/qu8-f32-vcvt/gen/ |
| D | vcvt-avx512skx-x64.c | 36 __m512i vx = _mm512_cvtepu8_epi32(_mm_loadu_si128((const __m128i*) (x + 48))); in xnn_qu8_f32_vcvt_ukernel__avx512skx_x64() local
61 __m512i vx = _mm512_cvtepu8_epi32(_mm_loadu_si128((const __m128i*) x)); in xnn_qu8_f32_vcvt_ukernel__avx512skx_x64() local
78 __m512i vx = _mm512_cvtepu8_epi32(_mm_maskz_loadu_epi8(vmask, x)); in xnn_qu8_f32_vcvt_ukernel__avx512skx_x64() local
|
| /external/XNNPACK/src/f32-vsqrt/ |
| D | sse-sqrt.c.in | 42 const __m128 vx = _mm_loadu_ps(x); variable
49 const __m128 vx = _mm_loadu_ps(x); variable
|
| D | wasmsimd-sqrt.c.in | 43 const v128_t vx = wasm_v128_load(x); variable
50 const v128_t vx = wasm_v128_load(x); variable
|
| D | neon-sqrt.c.in | 39 const float32x4_t vx = vld1q_f32(x); x += 4; variable
44 const float32x4_t vx = vld1q_f32(x); variable
|
| D | avx-sqrt.c.in | 42 const __m256 vx = _mm256_loadu_ps(x); variable
53 const __m256 vx = _mm256_maskload_ps(x, vmask); variable
|
| /external/XNNPACK/src/f32-qs8-vcvt/ |
| D | scalar-imagic.c.in | 62 float vx = *x++; variable
77 float vx = *x; variable
91 float vx = *x++; variable
|