| /external/XNNPACK/src/qs8-vcvt/gen/ |
| D | vcvt-ssse3-x16.c | 37 const __m128i vm0 = _mm_cmpgt_epi8(_mm_setzero_si128(), vx0); in xnn_qs8_vcvt_ukernel__ssse3_x16() local
|
| D | vcvt-ssse3-x32.c | 38 const __m128i vm0 = _mm_cmpgt_epi8(_mm_setzero_si128(), vx0); in xnn_qs8_vcvt_ukernel__ssse3_x32() local
|
| D | vcvt-sse2-x16.c | 36 const __m128i vm0 = _mm_cmpgt_epi8(_mm_setzero_si128(), vx0); in xnn_qs8_vcvt_ukernel__sse2_x16() local
|
| D | vcvt-sse2-x32.c | 37 const __m128i vm0 = _mm_cmpgt_epi8(_mm_setzero_si128(), vx0); in xnn_qs8_vcvt_ukernel__sse2_x32() local
|
| /external/XNNPACK/src/f32-argmaxpool/ |
| D | 9p8x-neon-c4.c | 144 const uint32x4_t vm0 = vcgtq_f32(vi0, vmax); in xnn_f32_argmaxpool_ukernel_9p8x__neon_c4() local
247 const uint32x4_t vm0 = vcgtq_f32(vi0, vmax); in xnn_f32_argmaxpool_ukernel_9p8x__neon_c4() local
302 const uint32x4_t vm0 = vcgtq_f32(vi0, vmax); in xnn_f32_argmaxpool_ukernel_9p8x__neon_c4() local
|
| D | 9p8x-wasmsimd-c4.c | 163 const v128_t vm0 = wasm_f32x4_gt(vi0, vmax); in xnn_f32_argmaxpool_ukernel_9p8x__wasmsimd_c4() local
278 const v128_t vm0 = wasm_f32x4_gt(vi0, vmax); in xnn_f32_argmaxpool_ukernel_9p8x__wasmsimd_c4() local
335 const v128_t vm0 = wasm_f32x4_gt(vi0, vmax); in xnn_f32_argmaxpool_ukernel_9p8x__wasmsimd_c4() local
|
| D | 9p8x-sse2-c4.c | 163 const __m128i vm0 = _mm_castps_si128(_mm_cmpgt_ps(vi0, vmax)); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() local
278 const __m128i vm0 = _mm_castps_si128(_mm_cmpgt_ps(vi0, vmax)); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() local
335 const __m128i vm0 = _mm_castps_si128(_mm_cmpgt_ps(vi0, vmax)); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() local
|
| /external/XNNPACK/src/qs8-vlrelu/gen/ |
| D | vlrelu-ssse3-x32.c | 39 const __m128i vm0 = _mm_cmpgt_epi8(_mm_setzero_si128(), vx0); in xnn_qs8_vlrelu_ukernel__ssse3_x32() local
|
| D | vlrelu-sse2-x32.c | 40 const __m128i vm0 = _mm_cmpgt_epi8(_mm_setzero_si128(), vx0); in xnn_qs8_vlrelu_ukernel__sse2_x32() local
|
| /external/XNNPACK/src/bf16-gemm/gen/ |
| D | 1x4c8-minmax-neonbf16-bfmlal.c | 76 const uint16x8_t vm0 = vceqq_u16(vreinterpretq_u16_bf16(vb0), vmovq_n_u16(0)); in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonbf16_bfmlal() local
|
| D | 1x4c8-minmax-neonbf16-bfdot.c | 71 const uint16x8_t vm0 = vceqq_u16(vreinterpretq_u16_bf16(vb0), vmovq_n_u16(0)); in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonbf16_bfdot() local
|
| D | 2x4c8-minmax-neonbf16-bfmlal.c | 96 const uint16x8_t vm0 = vceqq_u16(vreinterpretq_u16_bf16(vb0), vmovq_n_u16(0)); in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonbf16_bfmlal() local
|
| D | 2x4c8-minmax-neonbf16-bfdot.c | 87 const uint16x8_t vm0 = vceqq_u16(vreinterpretq_u16_bf16(vb0), vmovq_n_u16(0)); in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonbf16_bfdot() local
|
| D | 1x4c8-minmax-neonfma-zip.c | 91 const uint16x8_t vm0 = vceqq_u16(vb0, vmovq_n_u16(0)); in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonfma_zip() local
|
| D | 1x4c8-minmax-neonfma-shland.c | 91 const uint16x8_t vm0 = vceqq_u16(vb0, vmovq_n_u16(0)); in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonfma_shland() local
|
| /external/XNNPACK/src/f16-velu/gen/ |
| D | velu-neonfp16arith-rr1-p3-x16.c | 74 const uint16x8_t vm0 = vcltq_s16(vreinterpretq_s16_f16(vx0), vmovq_n_s16(0)); in xnn_f16_velu_ukernel__neonfp16arith_rr1_p3_x16() local
|
| /external/XNNPACK/src/f16-vsigmoid/gen/ |
| D | vsigmoid-neonfp16arith-rr2-p2-div-x16.c | 77 const uint16x8_t vm0 = vcltq_f16(vx0, vmovq_n_f16(0.0f)); in xnn_f16_vsigmoid_ukernel__neonfp16arith_rr2_p2_div_x16() local
|
| D | vsigmoid-neonfp16arith-rr2-p2-nr1fma-x16.c | 86 const uint16x8_t vm0 = vcltq_f16(vx0, vmovq_n_f16(0.0f)); in xnn_f16_vsigmoid_ukernel__neonfp16arith_rr2_p2_nr1fma_x16() local
|
| D | vsigmoid-neonfp16arith-rr2-p2-div-x24.c | 90 const uint16x8_t vm0 = vcltq_f16(vx0, vmovq_n_f16(0.0f)); in xnn_f16_vsigmoid_ukernel__neonfp16arith_rr2_p2_div_x24() local
|
| D | vsigmoid-neonfp16arith-rr2-p2-nr1recps-x16.c | 86 const uint16x8_t vm0 = vcltq_f16(vx0, vmovq_n_f16(0.0f)); in xnn_f16_vsigmoid_ukernel__neonfp16arith_rr2_p2_nr1recps_x16() local
|
| D | vsigmoid-neonfp16arith-rr2-p2-nr1recps-x24.c | 102 const uint16x8_t vm0 = vcltq_f16(vx0, vmovq_n_f16(0.0f)); in xnn_f16_vsigmoid_ukernel__neonfp16arith_rr2_p2_nr1recps_x24() local
|
| /external/XNNPACK/src/f16-raddstoreexpminusmax/gen/ |
| D | neonfp16arith-rr2-p2-x32-acc2.c | 88 const uint16x8_t vm0 = vcltq_f16(vx0, vdenorm_cutoff); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x32_acc2() local
|
| D | neonfp16arith-rr2-p2-x32.c | 87 const uint16x8_t vm0 = vcltq_f16(vx0, vdenorm_cutoff); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x32() local
|
| D | neonfp16arith-rr2-p2-x32-acc4.c | 90 const uint16x8_t vm0 = vcltq_f16(vx0, vdenorm_cutoff); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x32_acc4() local
|
| D | neonfp16arith-rr2-p2-x40-acc2.c | 97 const uint16x8_t vm0 = vcltq_f16(vx0, vdenorm_cutoff); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x40_acc2() local
|