/external/swiftshader/third_party/llvm-7.0/llvm/test/CodeGen/AMDGPU/ |
D | store-hi16.ll | 10 ; NO-D16-HI-NEXT: v_lshrrev_b32_e32 v2, 16, v2 30 ; NO-D16-HI-NEXT: v_lshrrev_b32_e32 v2, 16, v2 50 ; NO-D16-HI-NEXT: v_lshrrev_b32_e32 v2, 16, v2 69 ; NO-D16-HI-NEXT: v_lshrrev_b32_e32 v2, 16, v2 89 ; NO-D16-HI-NEXT: v_lshrrev_b32_e32 v2, 16, v2 109 ; GFX803-DAG: v_lshrrev_b32_e32 v2, 16, v2 112 ; GFX906-NEXT: v_lshrrev_b32_e32 v2, 16, v2 133 ; GFX803-DAG: v_lshrrev_b32_e32 v2, 16, v2 136 ; GFX906-NEXT: v_lshrrev_b32_e32 v2, 16, v2 156 ; GFX803-DAG: v_lshrrev_b32_e32 v2, 16, v2 [all …]
|
D | llvm.fma.f16.ll | 108 ; SI: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 109 ; SI: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]] 110 ; SI: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 124 ; VI: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 125 ; VI: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 126 ; VI: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]] 159 ; GCN-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 160 ; GCN-DAG: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]] 202 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 204 ; SI-DAG: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]] [all …]
|
D | sdwa-peephole.ll | 6 ; NOSDWA: v_lshrrev_b32_e32 v[[DST:[0-9]+]], 16, v{{[0-9]+}} 22 ; NOSDWA: v_lshrrev_b32_e32 v[[DST:[0-9]+]], 16, v{{[0-9]+}} 37 ; NOSDWA: v_lshrrev_b32_e32 v[[DST0:[0-9]+]], 16, v{{[0-9]+}} 38 ; NOSDWA: v_lshrrev_b32_e32 v[[DST1:[0-9]+]], 16, v{{[0-9]+}} 70 ; NOSDWA: v_lshrrev_b32_e32 v[[DST0:[0-9]+]], 16, v{{[0-9]+}} 71 ; NOSDWA: v_lshrrev_b32_e32 v[[DST1:[0-9]+]], 16, v{{[0-9]+}} 93 ; NOSDWA: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v{{[0-9]+}} 94 ; NOSDWA: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v{{[0-9]+}} 120 ; NOSDWA: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v{{[0-9]+}} 121 ; NOSDWA: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v{{[0-9]+}} [all …]
|
D | partial-shift-shrink.ll | 8 ; GCN-NEXT: v_lshrrev_b32_e32 v0, 16, v0 28 ; GCN-NEXT: v_lshrrev_b32_e32 v0, 15, v0 40 ; GCN-NEXT: v_lshrrev_b32_e32 v0, 16, v0 80 ; GCN-DAG: v_lshrrev_b32_e32 v0, 16, v0 107 ; GCN-NEXT: v_lshrrev_b32_e32 v0, v1, v0
|
D | srl.ll | 8 ; SI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 9 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 24 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 25 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 44 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 45 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 46 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 47 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}}
|
D | llvm.fmuladd.f16.ll | 112 ; SI: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 113 ; SI: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]] 114 ; SI: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 131 ; VI-FLUSH: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]] 138 ; VI-DENORM-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 139 ; VI-DENORM-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 140 ; VI-DENORM-DAG: v_lshrrev_b32_e32 v[[C_F16_1:[0-9]+]], 16, v[[C_V2_F16]]
|
D | shift-and-i128-ubfe.ll | 10 ; GCN: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 33 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 55 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 78 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 98 ; GCN-DAG: v_lshrrev_b32_e32 v[[ELT1PART:[0-9]+]], 2, v{{[[0-9]+}}
|
D | llvm.amdgcn.image.d16.dim.ll | 62 ; UNPACKED: v_lshrrev_b32_e32 74 ; UNPACKED: v_lshrrev_b32_e32 76 ; UNPACKED: v_lshrrev_b32_e32 88 ; UNPACKED: v_lshrrev_b32_e32 90 ; UNPACKED: v_lshrrev_b32_e32
|
D | llvm.amdgcn.buffer.load.format.d16.ll | 18 ; PACKED: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v[[FULL]] 31 ; PACKED: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v[[HI]]
|
D | llvm.amdgcn.tbuffer.load.d16.ll | 18 ; PACKED: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v[[FULL]] 31 ; PACKED: v_lshrrev_b32_e32 v{{[0-9]+}}, 16, v[[HI]]
|
D | fsub.f16.ll | 68 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 70 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 113 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 147 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]]
|
D | fadd.f16.ll | 69 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 71 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 106 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 137 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]]
|
D | bfe-combine.ll | 10 ; CI: v_lshrrev_b32_e32 v[[SHR:[0-9]+]], 6, v{{[0-9]+}} 33 ; CI: v_lshrrev_b32_e32 v[[SHR:[0-9]+]], 1, v{{[0-9]+}}
|
D | udiv.ll | 64 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 4, [[VAL]] 78 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 25, [[MULHI]] 92 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 24, [[MULHI]] 180 ; SI-NEXT: v_lshrrev_b32_e32 v0, 1, v0
|
D | reduction.ll | 179 ; GFX9: v_lshrrev_b32_e32 181 ; GFX9-NEXT: v_lshrrev_b32_e32 183 ; GFX9-NEXT: v_lshrrev_b32_e32 262 ; GFX9: v_lshrrev_b32_e32 264 ; GFX9-NEXT: v_lshrrev_b32_e32 266 ; GFX9-NEXT: v_lshrrev_b32_e32 268 ; GFX9-NEXT: v_lshrrev_b32_e32 270 ; GFX9-NEXT: v_lshrrev_b32_e32 272 ; GFX9-NEXT: v_lshrrev_b32_e32 274 ; GFX9-NEXT: v_lshrrev_b32_e32
|
D | fmul.f16.ll | 69 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 71 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 106 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 139 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]]
|
D | fneg.f16.ll | 107 ; CI: v_lshrrev_b32_e32 108 ; CI: v_lshrrev_b32_e32 154 ; CIVI: v_lshrrev_b32_e32 [[ELT1:v[0-9]+]], 16, [[NEG]]
|
D | llvm.minnum.f16.ll | 72 ; SI: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 73 ; SI: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 109 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 141 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]]
|
D | llvm.maxnum.f16.ll | 72 ; SI: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 73 ; SI: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] 110 ; SI-DAG: v_lshrrev_b32_e32 v[[B_F16_1:[0-9]+]], 16, v[[B_V2_F16]] 142 ; SI-DAG: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]]
|
D | shift-and-i64-ubfe.ll | 9 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 27 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 141 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 1, [[VAL]] 245 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 313 ; GCN: v_lshrrev_b32_e32 [[SHR:v[0-9]+]], 20, v[[VALLO]] 355 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHR:[0-9]+]], 2, [[VAL]]
|
/external/llvm/test/CodeGen/AMDGPU/ |
D | srl.ll | 8 ; SI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 9 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 24 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 25 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 44 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 45 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 46 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}} 47 ; VI: v_lshrrev_b32_e32 v{{[0-9]+, v[0-9]+, v[0-9]+}}
|
D | shift-and-i128-ubfe.ll | 10 ; GCN: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 32 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 54 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 76 ; GCN-DAG: v_lshrrev_b32_e32 v[[SHIFT:[0-9]+]], 31, [[VAL]] 96 ; GCN-DAG: v_lshrrev_b32_e32 v[[ELT1PART:[0-9]+]], 2, v{{[[0-9]+}}
|
D | udiv.ll | 57 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 4, [[VAL]] 71 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 25, [[MULHI]] 85 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 24, [[MULHI]]
|
D | scalar_to_vector.ll | 7 ; SI: v_lshrrev_b32_e32 [[SHR:v[0-9]+]], 16, [[VAL]] 22 ; SI: v_lshrrev_b32_e32 [[RESULT:v[0-9]+]], 16, [[VAL]]
|
D | mulhu.ll | 6 ;CHECK-NEXT: v_lshrrev_b32_e32 v0, 1, v0
|