/external/llvm-project/clang/test/CodeGen/X86/ |
D | avx512f-builtins-constrained-cmp.c | 20 return _mm512_cmp_ps_mask(a, b, _CMP_EQ_OQ); in test_mm512_cmp_ps_mask_eq_oq() 26 return _mm512_cmp_ps_mask(a, b, _CMP_LT_OS); in test_mm512_cmp_ps_mask_lt_os() 32 return _mm512_cmp_ps_mask(a, b, _CMP_LE_OS); in test_mm512_cmp_ps_mask_le_os() 38 return _mm512_cmp_ps_mask(a, b, _CMP_UNORD_Q); in test_mm512_cmp_ps_mask_unord_q() 44 return _mm512_cmp_ps_mask(a, b, _CMP_NEQ_UQ); in test_mm512_cmp_ps_mask_neq_uq() 50 return _mm512_cmp_ps_mask(a, b, _CMP_NLT_US); in test_mm512_cmp_ps_mask_nlt_us() 56 return _mm512_cmp_ps_mask(a, b, _CMP_NLE_US); in test_mm512_cmp_ps_mask_nle_us() 62 return _mm512_cmp_ps_mask(a, b, _CMP_ORD_Q); in test_mm512_cmp_ps_mask_ord_q() 68 return _mm512_cmp_ps_mask(a, b, _CMP_EQ_UQ); in test_mm512_cmp_ps_mask_eq_uq() 74 return _mm512_cmp_ps_mask(a, b, _CMP_NGE_US); in test_mm512_cmp_ps_mask_nge_us() [all …]
|
/external/XNNPACK/src/f32-prelu/gen/ |
D | avx512f-2x32.c | 60 …const __mmask16 vsign0x0123456789ABCDEF = _mm512_cmp_ps_mask(vi0x0123456789ABCDEF, vzero, _CMP_LT_… in xnn_f32_prelu_ukernel__avx512f_2x32() 62 …const __mmask16 vsign0xGHIJKLMNOPQRSTUV = _mm512_cmp_ps_mask(vi0xGHIJKLMNOPQRSTUV, vzero, _CMP_LT_… in xnn_f32_prelu_ukernel__avx512f_2x32() 64 …const __mmask16 vsign1x0123456789ABCDEF = _mm512_cmp_ps_mask(vi1x0123456789ABCDEF, vzero, _CMP_LT_… in xnn_f32_prelu_ukernel__avx512f_2x32() 66 …const __mmask16 vsign1xGHIJKLMNOPQRSTUV = _mm512_cmp_ps_mask(vi1xGHIJKLMNOPQRSTUV, vzero, _CMP_LT_… in xnn_f32_prelu_ukernel__avx512f_2x32() 85 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vi0, vzero, _CMP_LT_OQ); in xnn_f32_prelu_ukernel__avx512f_2x32() 87 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vi1, vzero, _CMP_LT_OQ); in xnn_f32_prelu_ukernel__avx512f_2x32() 108 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vi0, vzero, _CMP_LT_OQ); in xnn_f32_prelu_ukernel__avx512f_2x32() 110 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vi1, vzero, _CMP_LT_OQ); in xnn_f32_prelu_ukernel__avx512f_2x32()
|
D | avx512f-2x16.c | 57 …const __mmask16 vsign0x0123456789ABCDEF = _mm512_cmp_ps_mask(vi0x0123456789ABCDEF, vzero, _CMP_LT_… in xnn_f32_prelu_ukernel__avx512f_2x16() 59 …const __mmask16 vsign1x0123456789ABCDEF = _mm512_cmp_ps_mask(vi1x0123456789ABCDEF, vzero, _CMP_LT_… in xnn_f32_prelu_ukernel__avx512f_2x16() 80 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vi0, vzero, _CMP_LT_OQ); in xnn_f32_prelu_ukernel__avx512f_2x16() 82 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vi1, vzero, _CMP_LT_OQ); in xnn_f32_prelu_ukernel__avx512f_2x16()
|
/external/XNNPACK/src/f32-vlrelu/gen/ |
D | vlrelu-avx512f-x32.c | 36 …const __mmask16 vsign0123456789ABCDEF = _mm512_cmp_ps_mask(vacc0123456789ABCDEF, vzero, _CMP_LT_OQ… in xnn_f32_vlrelu_ukernel__avx512f_x32() 37 …const __mmask16 vsignGHIJKLMNOPQRSTUV = _mm512_cmp_ps_mask(vaccGHIJKLMNOPQRSTUV, vzero, _CMP_LT_OQ… in xnn_f32_vlrelu_ukernel__avx512f_x32() 49 const __mmask16 vsign = _mm512_cmp_ps_mask(vacc, vzero, _CMP_LT_OQ); in xnn_f32_vlrelu_ukernel__avx512f_x32()
|
/external/XNNPACK/src/f32-velu/gen/ |
D | velu-avx512f-rr1-lut16-p3-perm-x128.c | 160 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 162 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 164 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 166 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 168 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 170 const __mmask16 vsign5 = _mm512_cmp_ps_mask(vx5, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 172 const __mmask16 vsign6 = _mm512_cmp_ps_mask(vx6, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 174 const __mmask16 vsign7 = _mm512_cmp_ps_mask(vx7, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 200 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 233 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128()
|
D | velu-avx512f-rr1-lut16-p3-perm-x112.c | 147 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 149 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 151 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 153 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 155 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 157 const __mmask16 vsign5 = _mm512_cmp_ps_mask(vx5, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 159 const __mmask16 vsign6 = _mm512_cmp_ps_mask(vx6, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 183 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112() 216 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x112()
|
D | velu-avx512f-rr1-p6-x128.c | 170 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 172 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 174 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 176 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 178 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 180 const __mmask16 vsign5 = _mm512_cmp_ps_mask(vx5, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 182 const __mmask16 vsign6 = _mm512_cmp_ps_mask(vx6, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 184 const __mmask16 vsign7 = _mm512_cmp_ps_mask(vx7, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 210 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 244 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128()
|
D | velu-avx512f-rr1-lut16-p3-perm-x80.c | 121 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80() 123 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80() 125 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80() 127 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80() 129 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80() 149 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80() 182 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x80()
|
D | velu-avx512f-rr1-lut16-p3-perm-x96.c | 134 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 136 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 138 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 140 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 142 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 144 const __mmask16 vsign5 = _mm512_cmp_ps_mask(vx5, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 166 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96() 199 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x96()
|
D | velu-avx512f-rr1-p6-x96.c | 142 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 144 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 146 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 148 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 150 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 152 const __mmask16 vsign5 = _mm512_cmp_ps_mask(vx5, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 174 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96() 208 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x96()
|
D | velu-avx512f-rr1-p6-x112.c | 156 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 158 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 160 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 162 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 164 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 166 const __mmask16 vsign5 = _mm512_cmp_ps_mask(vx5, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 168 const __mmask16 vsign6 = _mm512_cmp_ps_mask(vx6, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 192 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112() 226 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x112()
|
D | velu-avx512f-rr1-p6-x80.c | 128 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80() 130 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80() 132 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80() 134 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80() 136 const __mmask16 vsign4 = _mm512_cmp_ps_mask(vx4, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80() 156 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80() 190 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x80()
|
D | velu-avx512f-rr1-p6-x64.c | 114 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x64() 116 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x64() 118 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x64() 120 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x64() 138 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x64() 172 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x64()
|
D | velu-avx512f-rr1-lut16-p3-perm-x64.c | 108 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64() 110 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64() 112 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64() 114 const __mmask16 vsign3 = _mm512_cmp_ps_mask(vx3, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64() 132 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64() 165 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x64()
|
D | velu-avx512f-rr1-lut16-p3-perm-x48.c | 95 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x48() 97 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x48() 99 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x48() 115 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x48() 148 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x48()
|
D | velu-avx512f-rr1-p6-x48.c | 100 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x48() 102 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x48() 104 const __mmask16 vsign2 = _mm512_cmp_ps_mask(vx2, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x48() 120 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x48() 154 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x48()
|
D | velu-avx512f-rr1-lut16-p3-perm-x32.c | 82 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x32() 84 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x32() 98 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x32() 131 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x32()
|
D | velu-avx512f-rr1-p6-x32.c | 86 const __mmask16 vsign0 = _mm512_cmp_ps_mask(vx0, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x32() 88 const __mmask16 vsign1 = _mm512_cmp_ps_mask(vx1, vzero, _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x32() 102 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x32() 136 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x32()
|
D | velu-avx512f-rr1-lut16-p3-perm-x16.c | 47 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x16() 80 const __mmask16 vsign = _mm512_cmp_ps_mask(vx, _mm512_setzero_ps(), _CMP_NLT_US); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x16()
|
/external/eigen/Eigen/src/Core/arch/AVX512/ |
D | MathFunctions.h | 68 _mm512_cmp_ps_mask(x, _mm512_setzero_ps(), _CMP_NGE_UQ); 70 _mm512_cmp_ps_mask(x, _mm512_setzero_ps(), _CMP_EQ_UQ); 90 __mmask16 mask = _mm512_cmp_ps_mask(x, p16f_cephes_SQRTHF, _CMP_LT_OQ); 268 __mmask16 non_zero_mask = _mm512_cmp_ps_mask(_x, p16f_flt_min, _CMP_GE_OQ); 335 __mmask16 le_zero_mask = _mm512_cmp_ps_mask(_x, p16f_flt_min, _CMP_LT_OQ); 340 __mmask16 neg_mask = _mm512_cmp_ps_mask(_x, _mm512_setzero_ps(), _CMP_LT_OQ);
|
/external/XNNPACK/src/math/ |
D | exp-avx512f-rr2-p5-scalef.c | 49 const __mmask16 vinvof = _mm512_cmp_ps_mask(vx, vinf_cutoff, _CMP_NGT_UQ); in xnn_math_f32_exp__avx512f_rr2_p5_scalef() 50 const __mmask16 vinvuf = _mm512_cmp_ps_mask(vx, vzero_cutoff, _CMP_NLT_UQ); in xnn_math_f32_exp__avx512f_rr2_p5_scalef()
|
D | exp-avx512f-rr2-lut32-p2-perm2.c | 61 const __mmask16 vinvof = _mm512_cmp_ps_mask(vx, vinf_cutoff, _CMP_NGT_UQ); in xnn_math_f32_exp__avx512f_rr2_lut32_p2_perm2() 62 const __mmask16 vinvuf = _mm512_cmp_ps_mask(vx, vzero_cutoff, _CMP_NLT_UQ); in xnn_math_f32_exp__avx512f_rr2_lut32_p2_perm2()
|
D | exp-avx512f-rr2-p5.c | 88 … __m512 vf = _mm512_maskz_fmadd_ps(_mm512_cmp_ps_mask(vx, vzero_cutoff, _CMP_NLT_US), vt, vp, vso); in xnn_math_f32_exp__avx512f_rr2_p5() 91 vf = _mm512_mask_mul_ps(vplus_inf, _mm512_cmp_ps_mask(vx, vinf_cutoff, _CMP_NGT_US), vsn, vf); in xnn_math_f32_exp__avx512f_rr2_p5()
|
D | exp-avx512f-rr2-lut16-p3-perm.c | 56 const __mmask16 vinvof = _mm512_cmp_ps_mask(vx, vinf_cutoff, _CMP_NGT_UQ); in xnn_math_f32_exp__avx512f_rr2_lut16_p3_perm() 57 const __mmask16 vinvuf = _mm512_cmp_ps_mask(vx, vzero_cutoff, _CMP_NLT_UQ); in xnn_math_f32_exp__avx512f_rr2_lut16_p3_perm()
|
/external/XNNPACK/src/f32-vlrelu/ |
D | avx512f.c.in | 37 … const __mmask16 vsign${ABC[N:N+16]} = _mm512_cmp_ps_mask(vacc${ABC[N:N+16]}, vzero, _CMP_LT_OQ); 51 const __mmask16 vsign = _mm512_cmp_ps_mask(vacc, vzero, _CMP_LT_OQ);
|