/external/XNNPACK/src/f32-prelu/gen/ |
D | wasmsimd-minmax-1x8.c | 50 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x8() local 69 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x8() local 84 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x8() local
|
D | wasmsimd-bitselect-1x8.c | 50 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x8() local 69 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x8() local 84 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x8() local
|
D | neon-1x8.c | 47 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x8() local 64 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x8() local 77 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x8() local
|
D | wasmsimd-minmax-1x16.c | 54 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x16() local 81 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x16() local 96 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x16() local
|
D | wasmsimd-minmax-2x8.c | 59 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x8() local 89 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x8() local 111 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x8() local
|
D | sse-2x8.c | 59 __m128 vacc0x0123 = _mm_max_ps(_mm_setzero_ps(), vi0x0123); in xnn_f32_prelu_ukernel__sse_2x8() local 89 __m128 vacc0x0123 = _mm_max_ps(_mm_setzero_ps(), vi0x0123); in xnn_f32_prelu_ukernel__sse_2x8() local 111 __m128 vacc0x0123 = _mm_max_ps(_mm_setzero_ps(), vi0x0123); in xnn_f32_prelu_ukernel__sse_2x8() local
|
D | neon-1x16.c | 51 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x16() local 76 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x16() local 89 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x16() local
|
D | wasmsimd-bitselect-1x16.c | 54 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x16() local 81 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x16() local 96 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x16() local
|
D | wasmsimd-bitselect-2x8.c | 59 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x8() local 89 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x8() local 111 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x8() local
|
D | sse41-2x8.c | 63 const __m128 vacc0x0123 = _mm_blendv_ps(vi0x0123, vprod0x0123, vi0x0123); in xnn_f32_prelu_ukernel__sse41_2x8() local 87 __m128 vacc0x0123 = _mm_blendv_ps(vi0x0123, vprod0x0123, vi0x0123); in xnn_f32_prelu_ukernel__sse41_2x8() local 107 __m128 vacc0x0123 = _mm_blendv_ps(vi0x0123, vprod0x0123, vi0x0123); in xnn_f32_prelu_ukernel__sse41_2x8() local
|
D | neon-2x8.c | 55 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_2x8() local 82 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_2x8() local 101 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_2x8() local
|
D | wasmsimd-minmax-2x16.c | 65 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x16() local 111 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x16() local 133 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x16() local
|
D | wasmsimd-minmax-1x4.c | 48 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x4() local 63 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_1x4() local
|
D | neon-1x4.c | 45 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x4() local 58 float32x4_t vacc0x0123 = vmulq_f32(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__neon_1x4() local
|
D | wasmsimd-bitselect-1x4.c | 48 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x4() local 63 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_1x4() local
|
D | sse2-2x8.c | 67 …const __m128 vacc0x0123 = _mm_or_ps(_mm_and_ps(vprod0x0123, vmask0x0123), _mm_andnot_ps(vmask0x012… in xnn_f32_prelu_ukernel__sse2_2x8() local 93 …__m128 vacc0x0123 = _mm_or_ps(_mm_and_ps(vprod0x0123, vmask0x0123), _mm_andnot_ps(vmask0x0123, vi0… in xnn_f32_prelu_ukernel__sse2_2x8() local 115 …__m128 vacc0x0123 = _mm_or_ps(_mm_and_ps(vprod0x0123, vmask0x0123), _mm_andnot_ps(vmask0x0123, vi0… in xnn_f32_prelu_ukernel__sse2_2x8() local
|
D | sse-2x4.c | 56 __m128 vacc0x0123 = _mm_max_ps(_mm_setzero_ps(), vi0x0123); in xnn_f32_prelu_ukernel__sse_2x4() local 78 __m128 vacc0x0123 = _mm_max_ps(_mm_setzero_ps(), vi0x0123); in xnn_f32_prelu_ukernel__sse_2x4() local
|
D | wasmsimd-minmax-2x4.c | 56 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x4() local 78 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_2x4() local
|
D | wasmsimd-bitselect-2x16.c | 65 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x16() local 111 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x16() local 133 v128_t vacc0x0123 = wasm_f32x4_mul(vi0x0123, vw0123); in xnn_f32_prelu_ukernel__wasmsimd_bitselect_2x16() local
|
D | wasmsimd-minmax-4x8.c | 77 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_4x8() local 129 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_4x8() local 165 v128_t vacc0x0123 = wasm_i32x4_max(vi0x0123, vzero); in xnn_f32_prelu_ukernel__wasmsimd_minmax_4x8() local
|
/external/XNNPACK/src/f32-vmulcaddc/gen/ |
D | c8-minmax-sse-2x.c | 53 __m128 vacc0x0123 = _mm_loadu_ps(i0); in xnn_f32_vmulcaddc_minmax_ukernel_c8__sse_2x() local 95 __m128 vacc0x0123 = _mm_loadu_ps(i0); in xnn_f32_vmulcaddc_minmax_ukernel_c8__sse_2x() local 124 __m128 vacc0x0123 = _mm_loadu_ps(i0); in xnn_f32_vmulcaddc_minmax_ukernel_c8__sse_2x() local
|
D | c8-minmax-neon-2x.c | 53 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 += 4; in xnn_f32_vmulcaddc_minmax_ukernel_c8__neon_2x() local 89 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 += 4; in xnn_f32_vmulcaddc_minmax_ukernel_c8__neon_2x() local 112 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 = (const float*) ((uintptr_t) i0 + c); in xnn_f32_vmulcaddc_minmax_ukernel_c8__neon_2x() local
|
D | c8-minmax-neonfma-2x.c | 53 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 += 4; in xnn_f32_vmulcaddc_minmax_ukernel_c8__neonfma_2x() local 85 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 += 4; in xnn_f32_vmulcaddc_minmax_ukernel_c8__neonfma_2x() local 106 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 = (const float*) ((uintptr_t) i0 + c); in xnn_f32_vmulcaddc_minmax_ukernel_c8__neonfma_2x() local
|
D | c4-minmax-sse-2x.c | 52 __m128 vacc0x0123 = _mm_loadu_ps(i0); in xnn_f32_vmulcaddc_minmax_ukernel_c4__sse_2x() local 81 __m128 vacc0x0123 = _mm_loadu_ps(i0); in xnn_f32_vmulcaddc_minmax_ukernel_c4__sse_2x() local
|
D | c4-minmax-neon-2x.c | 52 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 += 4; in xnn_f32_vmulcaddc_minmax_ukernel_c4__neon_2x() local 75 float32x4_t vacc0x0123 = vld1q_f32(i0); i0 = (const float*) ((uintptr_t) i0 + c); in xnn_f32_vmulcaddc_minmax_ukernel_c4__neon_2x() local
|