1; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py 2; RUN: llc -march=amdgcn -mcpu=gfx1010 -stop-after=greedy < %s | FileCheck %s 3 4%llpc.array.element = type <{ i32, [12 x i8] }> 5%llpc.array.element.2 = type <{ i32, [12 x i8] }> 6%llpc.array.element.5 = type <{ i32, [12 x i8] }> 7 8define amdgpu_gs void @_amdgpu_gs_main(i32 inreg %primShaderTableAddrLow, <31 x i32> inreg %userData) { 9 ; CHECK-LABEL: name: _amdgpu_gs_main 10 ; CHECK: bb.0..expVert: 11 ; CHECK: liveins: $sgpr3, $sgpr4, $sgpr5, $sgpr8, $sgpr9, $sgpr10, $sgpr18, $sgpr19, $sgpr20, $sgpr21, $sgpr22, $sgpr23, $sgpr25, $sgpr27, $sgpr31 12 ; CHECK: undef %56.sub0:sgpr_64 = COPY $sgpr31 13 ; CHECK: SI_SPILL_S32_SAVE $sgpr27, %stack.2, implicit $exec, implicit $sgpr32 :: (store 4 into %stack.2, addrspace 5) 14 ; CHECK: [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr25 15 ; CHECK: [[COPY1:%[0-9]+]]:sgpr_32 = COPY $sgpr5 16 ; CHECK: [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr4 17 ; CHECK: [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr3 18 ; CHECK: [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr18 19 ; CHECK: undef %50.sub0:sgpr_64 = COPY $sgpr19 20 ; CHECK: [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr20 21 ; CHECK: [[COPY6:%[0-9]+]]:sgpr_32 = COPY $sgpr21 22 ; CHECK: [[COPY7:%[0-9]+]]:sgpr_32 = COPY $sgpr22 23 ; CHECK: [[COPY8:%[0-9]+]]:sgpr_32 = COPY $sgpr23 24 ; CHECK: [[COPY9:%[0-9]+]]:sgpr_32 = COPY $sgpr9 25 ; CHECK: [[COPY10:%[0-9]+]]:sgpr_32 = COPY $sgpr10 26 ; CHECK: [[COPY11:%[0-9]+]]:sgpr_32 = COPY $sgpr8 27 ; CHECK: undef %71.sub0_sub1:sgpr_128 = S_LOAD_DWORDX2_IMM %56, 232, 0, 0 :: (load 8 from %ir.40, addrspace 4) 28 ; CHECK: [[S_MOV_B32_:%[0-9]+]]:sreg_32 = S_MOV_B32 65535 29 ; CHECK: [[S_LSHL_B32_:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY3]], 4, implicit-def dead $scc 30 ; CHECK: [[S_LSHL_B32_1:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY2]], 4, implicit-def dead $scc 31 ; CHECK: [[S_LSHL_B32_2:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY1]], 4, implicit-def dead $scc 32 ; CHECK: [[S_ASHR_I32_:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_]], 31, implicit-def dead $scc 33 ; CHECK: [[S_ASHR_I32_1:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_1]], 31, implicit-def dead $scc 34 ; CHECK: [[S_AND_B32_:%[0-9]+]]:sreg_32_xm0_xexec = S_AND_B32 %71.sub1, [[S_MOV_B32_]], implicit-def dead $scc 35 ; CHECK: SI_SPILL_S32_SAVE [[S_AND_B32_]], %stack.0, implicit $exec, implicit $sgpr32 :: (store 4 into %stack.0, addrspace 5) 36 ; CHECK: [[S_ASHR_I32_2:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_2]], 31, implicit-def dead $scc 37 ; CHECK: undef %130.sub0:sreg_64 = S_ADD_U32 [[COPY4]], [[S_LSHL_B32_2]], implicit-def $scc 38 ; CHECK: %130.sub1:sreg_64 = S_ADDC_U32 undef %54:sreg_32, [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 39 ; CHECK: [[S_LOAD_DWORDX4_IMM:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %130, 16, 0, 0 :: (load 16 from %ir.84, addrspace 4) 40 ; CHECK: [[S_LOAD_DWORDX4_IMM1:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM undef %74:sreg_64, 0, 0, 0 :: (load 16 from `<4 x i32> addrspace(4)* undef`, addrspace 4) 41 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM undef %132:sgpr_128, 0, 0, 0 :: (dereferenceable invariant load 4) 42 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM1:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM [[S_LOAD_DWORDX4_IMM]], 0, 0, 0 :: (dereferenceable invariant load 4) 43 ; CHECK: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 0, implicit $exec 44 ; CHECK: %71.sub3:sgpr_128 = S_MOV_B32 553734060 45 ; CHECK: %71.sub2:sgpr_128 = S_MOV_B32 -1 46 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET undef %118:sgpr_128, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 47 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], undef %89:sgpr_128, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 48 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN1:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM1]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 49 ; CHECK: SI_SPILL_S128_SAVE %71, %stack.1, implicit $exec, implicit $sgpr32 :: (store 16 into %stack.1, align 4, addrspace 5) 50 ; CHECK: %71.sub1:sgpr_128 = S_MOV_B32 0 51 ; CHECK: [[S_SUB_I32_:%[0-9]+]]:sreg_32 = S_SUB_I32 [[S_BUFFER_LOAD_DWORD_IMM]], 29, implicit-def dead $scc 52 ; CHECK: [[S_SUB_I32_1:%[0-9]+]]:sreg_32 = S_SUB_I32 [[S_BUFFER_LOAD_DWORD_IMM]], 30, implicit-def dead $scc 53 ; CHECK: [[S_SUB_I32_2:%[0-9]+]]:sreg_32 = S_SUB_I32 [[S_BUFFER_LOAD_DWORD_IMM1]], 31, implicit-def dead $scc 54 ; CHECK: [[S_ADD_U32_:%[0-9]+]]:sreg_32 = S_ADD_U32 [[COPY4]], 64, implicit-def $scc 55 ; CHECK: [[S_ADDC_U32_:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %54:sreg_32, 0, implicit-def dead $scc, implicit $scc 56 ; CHECK: undef %149.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], [[S_LSHL_B32_]], implicit-def $scc 57 ; CHECK: %149.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 58 ; CHECK: undef %156.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], [[S_LSHL_B32_1]], implicit-def $scc 59 ; CHECK: %156.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc 60 ; CHECK: undef %163.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], [[S_LSHL_B32_2]], implicit-def $scc 61 ; CHECK: [[S_LOAD_DWORDX4_IMM2:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %149, 0, 0, 0 :: (load 16 from %ir.91, addrspace 4) 62 ; CHECK: [[S_LOAD_DWORDX4_IMM3:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %156, 0, 0, 0 :: (load 16 from %ir.97, addrspace 4) 63 ; CHECK: %163.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 64 ; CHECK: [[S_ASHR_I32_3:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 undef %171:sreg_32, 31, implicit-def dead $scc 65 ; CHECK: undef %176.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], undef %171:sreg_32, implicit-def $scc 66 ; CHECK: [[S_LOAD_DWORDX4_IMM4:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %163, 0, 0, 0 :: (load 16 from %ir.103, addrspace 4) 67 ; CHECK: %176.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_3]], implicit-def dead $scc, implicit $scc 68 ; CHECK: undef %183.sub0:sreg_64 = S_ADD_U32 %50.sub0, [[S_LSHL_B32_]], implicit-def $scc 69 ; CHECK: %183.sub1:sreg_64 = S_ADDC_U32 undef %51:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 70 ; CHECK: undef %190.sub0:sreg_64 = S_ADD_U32 %50.sub0, [[S_LSHL_B32_1]], implicit-def $scc 71 ; CHECK: %190.sub1:sreg_64 = S_ADDC_U32 undef %51:sreg_32, [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc 72 ; CHECK: undef %200.sub0:sreg_64 = S_ADD_U32 %50.sub0, undef %171:sreg_32, implicit-def $scc 73 ; CHECK: %200.sub1:sreg_64 = S_ADDC_U32 undef %51:sreg_32, [[S_ASHR_I32_3]], implicit-def dead $scc, implicit $scc 74 ; CHECK: [[S_ADD_U32_1:%[0-9]+]]:sreg_32 = S_ADD_U32 %50.sub0, 224, implicit-def $scc 75 ; CHECK: [[S_ADDC_U32_1:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %51:sreg_32, 0, implicit-def dead $scc, implicit $scc 76 ; CHECK: undef %210.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_1]], [[S_LSHL_B32_]], implicit-def $scc 77 ; CHECK: %210.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_1]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 78 ; CHECK: undef %217.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_1]], [[S_LSHL_B32_1]], implicit-def $scc 79 ; CHECK: %217.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_1]], [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc 80 ; CHECK: undef %224.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_1]], [[S_LSHL_B32_2]], implicit-def $scc 81 ; CHECK: %224.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_1]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 82 ; CHECK: [[S_ADD_U32_2:%[0-9]+]]:sreg_32 = S_ADD_U32 %50.sub0, 576, implicit-def $scc 83 ; CHECK: [[S_ADDC_U32_2:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %51:sreg_32, 0, implicit-def dead $scc, implicit $scc 84 ; CHECK: undef %241.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_2]], [[S_LSHL_B32_]], implicit-def $scc 85 ; CHECK: %241.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_2]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 86 ; CHECK: undef %253.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_2]], [[S_LSHL_B32_2]], implicit-def $scc 87 ; CHECK: %253.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_2]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 88 ; CHECK: undef %261.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_2]], undef %171:sreg_32, implicit-def $scc 89 ; CHECK: %261.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_2]], [[S_ASHR_I32_3]], implicit-def dead $scc, implicit $scc 90 ; CHECK: undef %273.sub0:sreg_64 = S_ADD_U32 [[COPY5]], [[S_LSHL_B32_]], implicit-def $scc 91 ; CHECK: %273.sub1:sreg_64 = S_ADDC_U32 undef %48:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 92 ; CHECK: undef %286.sub0:sreg_64 = S_ADD_U32 [[COPY6]], [[S_LSHL_B32_1]], implicit-def $scc 93 ; CHECK: %286.sub1:sreg_64 = S_ADDC_U32 undef %45:sreg_32, [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc 94 ; CHECK: undef %293.sub0:sreg_64 = S_ADD_U32 [[COPY6]], [[S_LSHL_B32_2]], implicit-def $scc 95 ; CHECK: %293.sub1:sreg_64 = S_ADDC_U32 undef %45:sreg_32, [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 96 ; CHECK: [[S_ADD_I32_:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_LSHL_B32_]], 16, implicit-def dead $scc 97 ; CHECK: [[S_ADD_I32_1:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_LSHL_B32_2]], 16, implicit-def dead $scc 98 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR %71, [[S_ADD_I32_]], 0, 0 :: (dereferenceable invariant load 4) 99 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR1:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR %71, undef %314:sreg_32, 0, 0 :: (dereferenceable invariant load 4) 100 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR2:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR %71, [[S_ADD_I32_1]], 0, 0 :: (dereferenceable invariant load 4) 101 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM2:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %71, 16, 0, 0 :: (dereferenceable invariant load 4) 102 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR3:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR undef %369:sgpr_128, undef %370:sreg_32, 0, 0 :: (dereferenceable invariant load 4) 103 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM3:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM undef %380:sgpr_128, 16, 0, 0 :: (dereferenceable invariant load 4) 104 ; CHECK: [[S_LOAD_DWORDX4_IMM5:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %176, 0, 0, 0 :: (load 16 from %ir.111, addrspace 4) 105 ; CHECK: [[S_LOAD_DWORDX4_IMM6:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %183, 0, 0, 0 :: (load 16 from %ir.117, addrspace 4) 106 ; CHECK: [[S_LOAD_DWORDX4_IMM7:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %190, 0, 0, 0 :: (load 16 from %ir.123, addrspace 4) 107 ; CHECK: [[S_LOAD_DWORDX4_IMM8:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %200, 0, 0, 0 :: (load 16 from %ir.131, addrspace 4) 108 ; CHECK: [[S_LOAD_DWORDX4_IMM9:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %210, 0, 0, 0 :: (load 16 from %ir.138, addrspace 4) 109 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN2:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM2]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 110 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN3:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM3]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 111 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN4:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM4]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 112 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR4:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR undef %364:sgpr_128, [[S_ADD_I32_]], 0, 0 :: (dereferenceable invariant load 4) 113 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR5:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR undef %375:sgpr_128, [[S_ADD_I32_1]], 0, 0 :: (dereferenceable invariant load 4) 114 ; CHECK: [[S_ADD_I32_2:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR]], -98, implicit-def dead $scc 115 ; CHECK: [[S_ADD_I32_3:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR1]], -114, implicit-def dead $scc 116 ; CHECK: [[S_ADD_I32_4:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR2]], -130, implicit-def dead $scc 117 ; CHECK: [[S_ADD_I32_5:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM2]], -178, implicit-def dead $scc 118 ; CHECK: undef %327.sub0:sreg_64 = S_ADD_U32 [[COPY7]], [[S_LSHL_B32_]], implicit-def $scc 119 ; CHECK: %327.sub1:sreg_64 = S_ADDC_U32 undef %42:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 120 ; CHECK: undef %335.sub0:sreg_64 = S_ADD_U32 [[COPY8]], [[S_LSHL_B32_]], implicit-def $scc 121 ; CHECK: %335.sub1:sreg_64 = S_ADDC_U32 undef %39:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 122 ; CHECK: undef %343.sub0:sreg_64 = S_ADD_U32 [[COPY8]], [[S_LSHL_B32_1]], implicit-def $scc 123 ; CHECK: %343.sub1:sreg_64 = S_ADDC_U32 undef %39:sreg_32, [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc 124 ; CHECK: undef %351.sub0:sreg_64 = S_ADD_U32 [[COPY8]], [[S_LSHL_B32_2]], implicit-def $scc 125 ; CHECK: %351.sub1:sreg_64 = S_ADDC_U32 undef %39:sreg_32, [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 126 ; CHECK: [[S_LSHL_B32_3:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY9]], 4, implicit-def dead $scc 127 ; CHECK: [[S_ADD_I32_6:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_LSHL_B32_3]], 16, implicit-def dead $scc 128 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN5:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM5]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 129 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR6:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR undef %396:sgpr_128, [[S_ADD_I32_6]], 0, 0 :: (dereferenceable invariant load 4) 130 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN6:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM6]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 131 ; CHECK: [[S_LOAD_DWORDX4_IMM10:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %50, 224, 0, 0 :: (load 16 from %ir.155, addrspace 4) 132 ; CHECK: [[S_LOAD_DWORDX4_IMM11:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %217, 0, 0, 0 :: (load 16 from %ir.144, addrspace 4) 133 ; CHECK: [[S_LOAD_DWORDX4_IMM12:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %224, 0, 0, 0 :: (load 16 from %ir.150, addrspace 4) 134 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN7:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM7]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 135 ; CHECK: [[S_LOAD_DWORDX4_IMM13:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %241, 0, 0, 0 :: (load 16 from %ir.162, addrspace 4) 136 ; CHECK: [[S_LOAD_DWORDX4_IMM14:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %253, 0, 0, 0 :: (load 16 from %ir.170, addrspace 4) 137 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN8:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM8]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 138 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN9:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM9]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 139 ; CHECK: [[S_ADD_I32_7:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR4]], -217, implicit-def dead $scc 140 ; CHECK: [[S_ADD_I32_8:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -233, implicit-def dead $scc 141 ; CHECK: [[S_ADD_I32_9:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR5]], -249, implicit-def dead $scc 142 ; CHECK: [[S_ADD_I32_10:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM3]], -297, implicit-def dead $scc 143 ; CHECK: [[S_ADD_I32_11:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -313, implicit-def dead $scc 144 ; CHECK: [[S_ADD_I32_12:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -329, implicit-def dead $scc 145 ; CHECK: [[S_ADD_I32_13:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -345, implicit-def dead $scc 146 ; CHECK: [[S_ADD_I32_14:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR6]], -441, implicit-def dead $scc 147 ; CHECK: [[S_ADD_U32_3:%[0-9]+]]:sreg_32 = S_ADD_U32 [[COPY]], 160, implicit-def $scc 148 ; CHECK: [[S_ADDC_U32_3:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %36:sreg_32, 0, implicit-def dead $scc, implicit $scc 149 ; CHECK: undef %411.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_3]], [[S_LSHL_B32_2]], implicit-def $scc 150 ; CHECK: %411.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_3]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 151 ; CHECK: [[S_LSHL_B32_4:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY10]], 4, implicit-def dead $scc 152 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN10:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM11]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 153 ; CHECK: [[S_ASHR_I32_4:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_4]], 31, implicit-def dead $scc 154 ; CHECK: undef %425.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_3]], [[S_LSHL_B32_4]], implicit-def $scc 155 ; CHECK: %425.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_3]], [[S_ASHR_I32_4]], implicit-def dead $scc, implicit $scc 156 ; CHECK: [[S_ADD_U32_4:%[0-9]+]]:sreg_32 = S_ADD_U32 %56.sub0, 168, implicit-def $scc 157 ; CHECK: [[S_ADDC_U32_4:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %57:sreg_32, 0, implicit-def dead $scc, implicit $scc 158 ; CHECK: [[S_LSHL_B32_5:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY3]], 3, implicit-def dead $scc 159 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN11:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM12]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 160 ; CHECK: [[S_ASHR_I32_5:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_5]], 31, implicit-def dead $scc 161 ; CHECK: undef %441.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_5]], implicit-def $scc 162 ; CHECK: %441.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_5]], implicit-def dead $scc, implicit $scc 163 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN12:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM10]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 164 ; CHECK: %71.sub0:sgpr_128 = S_LOAD_DWORD_IMM %441, 0, 0, 0 :: (load 4 from %ir..i085.i, align 8, addrspace 4) 165 ; CHECK: [[S_LOAD_DWORDX4_IMM15:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %261, 0, 0, 0 :: (load 16 from %ir.176, addrspace 4) 166 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN13:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM13]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 167 ; CHECK: [[S_LOAD_DWORDX4_IMM16:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %273, 0, 0, 0 :: (load 16 from %ir.185, addrspace 4) 168 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN14:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM14]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 169 ; CHECK: [[S_LOAD_DWORDX4_IMM17:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %286, 0, 0, 0 :: (load 16 from %ir.194, addrspace 4) 170 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM4:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %71, 0, 0, 0 :: (dereferenceable invariant load 4) 171 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN15:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM15]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 172 ; CHECK: [[S_LOAD_DWORDX4_IMM18:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %293, 0, 0, 0 :: (load 16 from %ir.200, addrspace 4) 173 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN16:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM16]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 174 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET1:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET [[S_LOAD_DWORDX4_IMM1]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 175 ; CHECK: [[S_LSHL_B32_6:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY2]], 3, implicit-def dead $scc 176 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET2:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET [[S_LOAD_DWORDX4_IMM17]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 177 ; CHECK: [[S_ASHR_I32_6:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_6]], 31, implicit-def dead $scc 178 ; CHECK: [[S_ADD_I32_15:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM4]], -467, implicit-def dead $scc 179 ; CHECK: undef %453.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_6]], implicit-def $scc 180 ; CHECK: %453.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_6]], implicit-def dead $scc, implicit $scc 181 ; CHECK: %71.sub0_sub1:sgpr_128 = S_LOAD_DWORDX2_IMM %453, 0, 0, 0 :: (load 8 from %ir.304, addrspace 4) 182 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET3:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET [[S_LOAD_DWORDX4_IMM18]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 183 ; CHECK: [[S_LOAD_DWORDX4_IMM19:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %327, 0, 0, 0 :: (load 16 from %ir.223, addrspace 4) 184 ; CHECK: [[S_LOAD_DWORDX4_IMM20:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %335, 0, 0, 0 :: (load 16 from %ir.230, addrspace 4) 185 ; CHECK: [[S_LOAD_DWORDX4_IMM21:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %343, 0, 0, 0 :: (load 16 from %ir.236, addrspace 4) 186 ; CHECK: [[S_LOAD_DWORDX4_IMM22:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %351, 0, 0, 0 :: (load 16 from %ir.242, addrspace 4) 187 ; CHECK: %71.sub1:sgpr_128 = S_AND_B32 %71.sub1, [[S_MOV_B32_]], implicit-def dead $scc 188 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM5:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %71, 0, 0, 0 :: (dereferenceable invariant load 4) 189 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN17:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM19]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 190 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN18:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM20]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 191 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN19:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM21]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 192 ; CHECK: [[S_LSHL_B32_7:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY1]], 3, implicit-def dead $scc 193 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN20:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM22]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 194 ; CHECK: [[S_ASHR_I32_7:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_7]], 31, implicit-def dead $scc 195 ; CHECK: [[S_ADD_I32_16:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM5]], -468, implicit-def dead $scc 196 ; CHECK: undef %468.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_7]], implicit-def $scc 197 ; CHECK: %468.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_7]], implicit-def dead $scc, implicit $scc 198 ; CHECK: %71.sub0_sub1:sgpr_128 = S_LOAD_DWORDX2_IMM %468, 0, 0, 0 :: (load 8 from %ir.316, addrspace 4) 199 ; CHECK: %71.sub1:sgpr_128 = S_AND_B32 %71.sub1, [[S_MOV_B32_]], implicit-def dead $scc 200 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM6:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %71, 0, 0, 0 :: (dereferenceable invariant load 4) 201 ; CHECK: [[S_LOAD_DWORDX4_IMM23:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %411, 0, 0, 0 :: (load 16 from %ir.278, addrspace 4) 202 ; CHECK: [[S_LOAD_DWORD_IMM:%[0-9]+]]:sreg_32_xm0_xexec = S_LOAD_DWORD_IMM undef %488:sreg_64, 0, 0, 0 :: (load 4 from `i32 addrspace(4)* undef`, addrspace 4) 203 ; CHECK: [[S_LSHL_B32_8:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY11]], 3, implicit-def dead $scc 204 ; CHECK: [[S_LOAD_DWORDX4_IMM24:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %425, 0, 0, 0 :: (load 16 from %ir.287, addrspace 4) 205 ; CHECK: [[S_ASHR_I32_8:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_8]], 31, implicit-def dead $scc 206 ; CHECK: [[S_ADD_I32_17:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM6]], -469, implicit-def dead $scc 207 ; CHECK: undef %485.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_8]], implicit-def $scc 208 ; CHECK: %485.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_8]], implicit-def dead $scc, implicit $scc 209 ; CHECK: %71.sub0:sgpr_128 = S_LOAD_DWORD_IMM %485, 0, 0, 0 :: (load 4 from %ir..i0100.i, align 8, addrspace 4) 210 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN21:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM23]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 211 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN22:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM24]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 212 ; CHECK: %71.sub1:sgpr_128 = S_AND_B32 [[S_LOAD_DWORD_IMM]], [[S_MOV_B32_]], implicit-def dead $scc 213 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM7:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %71, 0, 0, 0 :: (dereferenceable invariant load 4) 214 ; CHECK: [[S_ADD_I32_18:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM]], -474, implicit-def dead $scc 215 ; CHECK: [[S_ADD_I32_19:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -475, implicit-def dead $scc 216 ; CHECK: [[S_ADD_I32_20:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -491, implicit-def dead $scc 217 ; CHECK: [[S_ADD_I32_21:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -507, implicit-def dead $scc 218 ; CHECK: [[S_ADD_I32_22:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -539, implicit-def dead $scc 219 ; CHECK: [[S_ADD_I32_23:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM7]], -473, implicit-def dead $scc 220 ; CHECK: [[SI_SPILL_S32_RESTORE:%[0-9]+]]:sgpr_32 = SI_SPILL_S32_RESTORE %stack.2, implicit $exec, implicit $sgpr32 :: (load 4 from %stack.2, addrspace 5) 221 ; CHECK: [[S_ADD_U32_5:%[0-9]+]]:sreg_32 = S_ADD_U32 [[SI_SPILL_S32_RESTORE]], 96, implicit-def $scc 222 ; CHECK: [[S_ADDC_U32_5:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %33:sreg_32, 0, implicit-def dead $scc, implicit $scc 223 ; CHECK: undef %514.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_5]], [[S_LSHL_B32_]], implicit-def $scc 224 ; CHECK: %514.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_5]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc 225 ; CHECK: [[S_LOAD_DWORDX4_IMM25:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %514, 0, 0, 0 :: (load 16 from %ir.347, addrspace 4) 226 ; CHECK: undef %522.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_5]], [[S_LSHL_B32_1]], implicit-def $scc 227 ; CHECK: %522.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_5]], [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc 228 ; CHECK: [[S_LOAD_DWORDX4_IMM26:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %522, 0, 0, 0 :: (load 16 from %ir.353, addrspace 4) 229 ; CHECK: undef %530.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_5]], [[S_LSHL_B32_2]], implicit-def $scc 230 ; CHECK: %530.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_5]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc 231 ; CHECK: [[S_LOAD_DWORDX4_IMM27:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %530, 0, 0, 0 :: (load 16 from %ir.359, addrspace 4) 232 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN23:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM25]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 233 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN24:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM26]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 234 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN25:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM27]], 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4) 235 ; CHECK: [[V_ADD_U32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -2, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 236 ; CHECK: [[V_ADD_U32_e32_1:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -1, [[BUFFER_LOAD_FORMAT_X_IDXEN1]], implicit $exec 237 ; CHECK: [[V_ADD_U32_e32_2:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -3, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 238 ; CHECK: [[V_OR_B32_e32_:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_ADD_U32_e32_]], [[V_ADD_U32_e32_1]], implicit $exec 239 ; CHECK: [[V_ADD_U32_e32_3:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -4, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 240 ; CHECK: [[V_OR_B32_e32_1:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_]], [[V_ADD_U32_e32_2]], implicit $exec 241 ; CHECK: [[V_SUBREV_U32_e32_:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 27, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 242 ; CHECK: [[V_OR_B32_e32_2:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_1]], [[V_ADD_U32_e32_3]], implicit $exec 243 ; CHECK: [[V_SUBREV_U32_e32_1:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 28, [[BUFFER_LOAD_DWORD_OFFSET]], implicit $exec 244 ; CHECK: [[V_OR_B32_e32_3:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_2]], [[V_SUBREV_U32_e32_]], implicit $exec 245 ; CHECK: [[V_OR_B32_e32_4:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_3]], [[V_SUBREV_U32_e32_1]], implicit $exec 246 ; CHECK: [[V_OR_B32_e32_5:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_SUB_I32_]], [[V_OR_B32_e32_4]], implicit $exec 247 ; CHECK: [[V_OR_B32_e32_6:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_SUB_I32_1]], [[V_OR_B32_e32_5]], implicit $exec 248 ; CHECK: [[V_OR_B32_e32_7:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_SUB_I32_2]], [[V_OR_B32_e32_6]], implicit $exec 249 ; CHECK: [[V_SUBREV_U32_e32_2:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 32, [[BUFFER_LOAD_FORMAT_X_IDXEN2]], implicit $exec 250 ; CHECK: [[V_SUBREV_U32_e32_3:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 33, [[BUFFER_LOAD_FORMAT_X_IDXEN3]], implicit $exec 251 ; CHECK: [[V_OR_B32_e32_8:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_7]], [[V_SUBREV_U32_e32_2]], implicit $exec 252 ; CHECK: [[V_SUBREV_U32_e32_4:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 34, [[BUFFER_LOAD_FORMAT_X_IDXEN4]], implicit $exec 253 ; CHECK: [[V_OR_B32_e32_9:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_8]], [[V_SUBREV_U32_e32_3]], implicit $exec 254 ; CHECK: [[V_SUBREV_U32_e32_5:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 36, [[BUFFER_LOAD_FORMAT_X_IDXEN5]], implicit $exec 255 ; CHECK: [[V_OR_B32_e32_10:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_9]], [[V_SUBREV_U32_e32_4]], implicit $exec 256 ; CHECK: [[V_SUBREV_U32_e32_6:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 37, [[BUFFER_LOAD_FORMAT_X_IDXEN6]], implicit $exec 257 ; CHECK: [[V_OR_B32_e32_11:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_10]], [[V_SUBREV_U32_e32_5]], implicit $exec 258 ; CHECK: [[V_SUBREV_U32_e32_7:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 38, [[BUFFER_LOAD_FORMAT_X_IDXEN7]], implicit $exec 259 ; CHECK: [[V_OR_B32_e32_12:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_11]], [[V_SUBREV_U32_e32_6]], implicit $exec 260 ; CHECK: [[V_SUBREV_U32_e32_8:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 39, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 261 ; CHECK: [[V_OR_B32_e32_13:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_12]], [[V_SUBREV_U32_e32_7]], implicit $exec 262 ; CHECK: [[V_SUBREV_U32_e32_9:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 50, [[BUFFER_LOAD_FORMAT_X_IDXEN8]], implicit $exec 263 ; CHECK: [[V_OR_B32_e32_14:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_13]], [[V_SUBREV_U32_e32_8]], implicit $exec 264 ; CHECK: [[V_SUBREV_U32_e32_10:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 51, [[BUFFER_LOAD_FORMAT_X_IDXEN9]], implicit $exec 265 ; CHECK: [[V_OR_B32_e32_15:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_14]], [[V_SUBREV_U32_e32_9]], implicit $exec 266 ; CHECK: [[V_SUBREV_U32_e32_11:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 52, [[BUFFER_LOAD_FORMAT_X_IDXEN10]], implicit $exec 267 ; CHECK: [[V_OR_B32_e32_16:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_15]], [[V_SUBREV_U32_e32_10]], implicit $exec 268 ; CHECK: [[V_SUBREV_U32_e32_12:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 53, [[BUFFER_LOAD_FORMAT_X_IDXEN11]], implicit $exec 269 ; CHECK: [[V_OR_B32_e32_17:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_16]], [[V_SUBREV_U32_e32_11]], implicit $exec 270 ; CHECK: [[V_ADD_U32_e32_4:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -72, [[BUFFER_LOAD_FORMAT_X_IDXEN12]], implicit $exec 271 ; CHECK: [[V_OR_B32_e32_18:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_17]], [[V_SUBREV_U32_e32_12]], implicit $exec 272 ; CHECK: [[V_ADD_U32_e32_5:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -73, [[BUFFER_LOAD_FORMAT_X_IDXEN13]], implicit $exec 273 ; CHECK: [[V_OR_B32_e32_19:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_18]], [[V_ADD_U32_e32_4]], implicit $exec 274 ; CHECK: [[V_ADD_U32_e32_6:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -74, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 275 ; CHECK: [[V_OR_B32_e32_20:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_19]], [[V_ADD_U32_e32_5]], implicit $exec 276 ; CHECK: [[V_ADD_U32_e32_7:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -75, [[BUFFER_LOAD_FORMAT_X_IDXEN14]], implicit $exec 277 ; CHECK: [[V_OR_B32_e32_21:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_20]], [[V_ADD_U32_e32_6]], implicit $exec 278 ; CHECK: [[V_ADD_U32_e32_8:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -77, [[BUFFER_LOAD_FORMAT_X_IDXEN15]], implicit $exec 279 ; CHECK: [[V_OR_B32_e32_22:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_21]], [[V_ADD_U32_e32_7]], implicit $exec 280 ; CHECK: [[V_ADD_U32_e32_9:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -93, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 281 ; CHECK: [[V_OR_B32_e32_23:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_22]], [[V_ADD_U32_e32_8]], implicit $exec 282 ; CHECK: [[V_ADD_U32_e32_10:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -94, [[BUFFER_LOAD_FORMAT_X_IDXEN16]], implicit $exec 283 ; CHECK: [[V_OR_B32_e32_24:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_23]], [[V_ADD_U32_e32_9]], implicit $exec 284 ; CHECK: [[V_ADD_U32_e32_11:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -95, [[BUFFER_LOAD_DWORD_OFFSET1]], implicit $exec 285 ; CHECK: [[V_OR_B32_e32_25:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_24]], [[V_ADD_U32_e32_10]], implicit $exec 286 ; CHECK: [[V_ADD_U32_e32_12:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -96, [[BUFFER_LOAD_DWORD_OFFSET2]], implicit $exec 287 ; CHECK: [[V_OR_B32_e32_26:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_25]], [[V_ADD_U32_e32_11]], implicit $exec 288 ; CHECK: [[V_ADD_U32_e32_13:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -97, [[BUFFER_LOAD_DWORD_OFFSET3]], implicit $exec 289 ; CHECK: [[V_OR_B32_e32_27:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_26]], [[V_ADD_U32_e32_12]], implicit $exec 290 ; CHECK: [[V_OR_B32_e32_28:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_27]], [[V_ADD_U32_e32_13]], implicit $exec 291 ; CHECK: [[V_OR_B32_e32_29:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_2]], [[V_OR_B32_e32_28]], implicit $exec 292 ; CHECK: [[V_OR_B32_e32_30:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_3]], [[V_OR_B32_e32_29]], implicit $exec 293 ; CHECK: [[V_OR_B32_e32_31:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_4]], [[V_OR_B32_e32_30]], implicit $exec 294 ; CHECK: [[V_ADD_U32_e32_14:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -194, [[BUFFER_LOAD_FORMAT_X_IDXEN17]], implicit $exec 295 ; CHECK: [[V_OR_B32_e32_32:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_5]], [[V_OR_B32_e32_31]], implicit $exec 296 ; CHECK: [[V_ADD_U32_e32_15:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -195, [[BUFFER_LOAD_FORMAT_X_IDXEN18]], implicit $exec 297 ; CHECK: [[V_OR_B32_e32_33:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_32]], [[V_ADD_U32_e32_14]], implicit $exec 298 ; CHECK: [[V_ADD_U32_e32_16:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -196, [[BUFFER_LOAD_FORMAT_X_IDXEN19]], implicit $exec 299 ; CHECK: [[V_OR_B32_e32_34:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_33]], [[V_ADD_U32_e32_15]], implicit $exec 300 ; CHECK: [[V_ADD_U32_e32_17:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -197, [[BUFFER_LOAD_FORMAT_X_IDXEN20]], implicit $exec 301 ; CHECK: [[V_OR_B32_e32_35:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_34]], [[V_ADD_U32_e32_16]], implicit $exec 302 ; CHECK: [[V_ADD_U32_e32_18:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -216, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 303 ; CHECK: [[V_OR_B32_e32_36:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_35]], [[V_ADD_U32_e32_17]], implicit $exec 304 ; CHECK: [[V_OR_B32_e32_37:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_36]], [[V_ADD_U32_e32_18]], implicit $exec 305 ; CHECK: [[V_OR_B32_e32_38:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_7]], [[V_OR_B32_e32_37]], implicit $exec 306 ; CHECK: [[V_OR_B32_e32_39:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_8]], [[V_OR_B32_e32_38]], implicit $exec 307 ; CHECK: [[V_OR_B32_e32_40:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_9]], [[V_OR_B32_e32_39]], implicit $exec 308 ; CHECK: [[V_OR_B32_e32_41:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_10]], [[V_OR_B32_e32_40]], implicit $exec 309 ; CHECK: [[V_OR_B32_e32_42:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_11]], [[V_OR_B32_e32_41]], implicit $exec 310 ; CHECK: [[V_OR_B32_e32_43:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_12]], [[V_OR_B32_e32_42]], implicit $exec 311 ; CHECK: [[V_OR_B32_e32_44:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_13]], [[V_OR_B32_e32_43]], implicit $exec 312 ; CHECK: [[V_ADD_U32_e32_19:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -457, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 313 ; CHECK: [[V_OR_B32_e32_45:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_14]], [[V_OR_B32_e32_44]], implicit $exec 314 ; CHECK: [[V_ADD_U32_e32_20:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -458, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 315 ; CHECK: [[V_OR_B32_e32_46:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_45]], [[V_ADD_U32_e32_19]], implicit $exec 316 ; CHECK: [[V_ADD_U32_e32_21:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -459, [[BUFFER_LOAD_FORMAT_X_IDXEN21]], implicit $exec 317 ; CHECK: [[V_OR_B32_e32_47:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_46]], [[V_ADD_U32_e32_20]], implicit $exec 318 ; CHECK: [[V_ADD_U32_e32_22:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -466, [[BUFFER_LOAD_FORMAT_X_IDXEN22]], implicit $exec 319 ; CHECK: [[V_OR_B32_e32_48:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_47]], [[V_ADD_U32_e32_21]], implicit $exec 320 ; CHECK: [[V_OR_B32_e32_49:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_48]], [[V_ADD_U32_e32_22]], implicit $exec 321 ; CHECK: [[V_OR_B32_e32_50:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_15]], [[V_OR_B32_e32_49]], implicit $exec 322 ; CHECK: [[V_OR_B32_e32_51:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_16]], [[V_OR_B32_e32_50]], implicit $exec 323 ; CHECK: [[V_OR_B32_e32_52:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_17]], [[V_OR_B32_e32_51]], implicit $exec 324 ; CHECK: [[V_OR_B32_e32_53:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_23]], [[V_OR_B32_e32_52]], implicit $exec 325 ; CHECK: [[V_OR_B32_e32_54:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_18]], [[V_OR_B32_e32_53]], implicit $exec 326 ; CHECK: [[V_OR_B32_e32_55:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_19]], [[V_OR_B32_e32_54]], implicit $exec 327 ; CHECK: [[V_OR_B32_e32_56:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_20]], [[V_OR_B32_e32_55]], implicit $exec 328 ; CHECK: [[V_OR_B32_e32_57:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_21]], [[V_OR_B32_e32_56]], implicit $exec 329 ; CHECK: [[V_OR_B32_e32_58:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_22]], [[V_OR_B32_e32_57]], implicit $exec 330 ; CHECK: [[V_ADD_U32_e32_23:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -555, [[BUFFER_LOAD_FORMAT_X_IDXEN23]], implicit $exec 331 ; CHECK: [[V_ADD_U32_e32_24:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -556, [[BUFFER_LOAD_FORMAT_X_IDXEN24]], implicit $exec 332 ; CHECK: [[V_OR_B32_e32_59:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_58]], [[V_ADD_U32_e32_23]], implicit $exec 333 ; CHECK: [[V_ADD_U32_e32_25:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -557, [[BUFFER_LOAD_FORMAT_X_IDXEN25]], implicit $exec 334 ; CHECK: [[V_OR_B32_e32_60:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_59]], [[V_ADD_U32_e32_24]], implicit $exec 335 ; CHECK: [[V_ADD_U32_e32_26:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -574, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 336 ; CHECK: [[V_OR_B32_e32_61:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_60]], [[V_ADD_U32_e32_25]], implicit $exec 337 ; CHECK: [[V_ADD_U32_e32_27:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -575, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 338 ; CHECK: [[V_OR_B32_e32_62:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_61]], [[V_ADD_U32_e32_26]], implicit $exec 339 ; CHECK: [[SI_SPILL_S32_RESTORE1:%[0-9]+]]:sreg_32_xm0_xexec = SI_SPILL_S32_RESTORE %stack.0, implicit $exec, implicit $sgpr32 :: (load 4 from %stack.0, addrspace 5) 340 ; CHECK: [[SI_SPILL_S128_RESTORE:%[0-9]+]]:sgpr_128 = SI_SPILL_S128_RESTORE %stack.1, implicit $exec, implicit $sgpr32 :: (load 16 from %stack.1, align 4, addrspace 5) 341 ; CHECK: undef %914.sub2_sub3:sgpr_128 = COPY [[SI_SPILL_S128_RESTORE]].sub2_sub3 { 342 ; CHECK: internal %914.sub0:sgpr_128 = COPY [[SI_SPILL_S128_RESTORE]].sub0 343 ; CHECK: } 344 ; CHECK: %914.sub1:sgpr_128 = COPY [[SI_SPILL_S32_RESTORE1]] 345 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM8:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %914, 0, 0, 0 :: (dereferenceable invariant load 4) 346 ; CHECK: [[V_ADD_U32_e32_28:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -576, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 347 ; CHECK: [[V_OR_B32_e32_63:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_62]], [[V_ADD_U32_e32_27]], implicit $exec 348 ; CHECK: [[V_ADD_U32_e32_29:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -577, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 349 ; CHECK: [[V_OR_B32_e32_64:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_63]], [[V_ADD_U32_e32_28]], implicit $exec 350 ; CHECK: [[V_ADD_U32_e32_30:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -593, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec 351 ; CHECK: [[V_OR_B32_e32_65:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_64]], [[V_ADD_U32_e32_29]], implicit $exec 352 ; CHECK: [[S_LOAD_DWORDX8_IMM:%[0-9]+]]:sgpr_256 = S_LOAD_DWORDX8_IMM undef %564:sreg_64, 0, 0, 0 :: (load 32 from `<8 x i32> addrspace(4)* undef`, addrspace 4) 353 ; CHECK: [[V_OR_B32_e32_66:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_OR_B32_e32_65]], [[V_ADD_U32_e32_30]], implicit $exec 354 ; CHECK: [[S_ADD_I32_24:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM8]], -594, implicit-def dead $scc 355 ; CHECK: [[V_OR_B32_e32_67:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_ADD_I32_24]], [[V_OR_B32_e32_66]], implicit $exec 356 ; CHECK: [[V_CMP_EQ_U32_e64_:%[0-9]+]]:sreg_32_xm0_xexec = V_CMP_EQ_U32_e64 0, [[V_OR_B32_e32_67]], implicit $exec 357 ; CHECK: undef %691.sub3:vreg_128 = V_CNDMASK_B32_e64 0, 0, 0, 1, [[V_CMP_EQ_U32_e64_]], implicit $exec 358 ; CHECK: IMAGE_STORE_V4_V2_gfx10 %691, undef %578:vreg_64, [[S_LOAD_DWORDX8_IMM]], 15, 1, -1, 0, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable store 16 into custom "TargetCustom8") 359 ; CHECK: S_ENDPGM 0 360.expVert: 361 %0 = extractelement <31 x i32> %userData, i64 2 362 %1 = extractelement <31 x i32> %userData, i64 3 363 %2 = extractelement <31 x i32> %userData, i64 4 364 %3 = extractelement <31 x i32> %userData, i64 7 365 %4 = extractelement <31 x i32> %userData, i64 8 366 %5 = extractelement <31 x i32> %userData, i64 9 367 %6 = extractelement <31 x i32> %userData, i64 17 368 %7 = extractelement <31 x i32> %userData, i64 18 369 %8 = extractelement <31 x i32> %userData, i64 19 370 %9 = extractelement <31 x i32> %userData, i64 20 371 %10 = extractelement <31 x i32> %userData, i64 21 372 %11 = extractelement <31 x i32> %userData, i64 22 373 %12 = extractelement <31 x i32> %userData, i64 24 374 %13 = extractelement <31 x i32> %userData, i64 26 375 %14 = extractelement <31 x i32> %userData, i64 30 376 %15 = insertelement <2 x i32> undef, i32 %13, i32 0 377 %16 = bitcast <2 x i32> %15 to i64 378 %17 = inttoptr i64 %16 to i8 addrspace(4)* 379 %18 = insertelement <2 x i32> undef, i32 %12, i32 0 380 %19 = bitcast <2 x i32> %18 to i64 381 %20 = inttoptr i64 %19 to i8 addrspace(4)* 382 %21 = insertelement <2 x i32> undef, i32 %11, i32 0 383 %22 = bitcast <2 x i32> %21 to i64 384 %23 = insertelement <2 x i32> undef, i32 %10, i32 0 385 %24 = bitcast <2 x i32> %23 to i64 386 %25 = insertelement <2 x i32> undef, i32 %9, i32 0 387 %26 = bitcast <2 x i32> %25 to i64 388 %27 = inttoptr i64 %26 to i8 addrspace(4)* 389 %28 = insertelement <2 x i32> undef, i32 %8, i32 0 390 %29 = bitcast <2 x i32> %28 to i64 391 %30 = insertelement <2 x i32> undef, i32 %7, i32 0 392 %31 = bitcast <2 x i32> %30 to i64 393 %32 = inttoptr i64 %31 to i8 addrspace(4)* 394 %33 = insertelement <2 x i32> undef, i32 %6, i32 0 395 %34 = bitcast <2 x i32> %33 to i64 396 %35 = inttoptr i64 %34 to i8 addrspace(4)* 397 %36 = insertelement <2 x i32> undef, i32 %14, i32 0 398 %37 = bitcast <2 x i32> %36 to i64 399 %38 = inttoptr i64 %37 to i8 addrspace(4)* 400 %39 = getelementptr i8, i8 addrspace(4)* %38, i64 232 401 %.i0.i = bitcast i8 addrspace(4)* %39 to i32 addrspace(4)* 402 %rootDesc58.ii0.i = load i32, i32 addrspace(4)* %.i0.i, align 8 403 %.i184.i = getelementptr i8, i8 addrspace(4)* %38, i64 236 404 %40 = bitcast i8 addrspace(4)* %.i184.i to i32 addrspace(4)* 405 %rootDesc58.ii1.i = load i32, i32 addrspace(4)* %40, align 4 406 %41 = and i32 %rootDesc58.ii1.i, 65535 407 %42 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %rootDesc58.ii0.i, i32 0 408 %43 = insertelement <4 x i32> %42, i32 %41, i32 1 409 %44 = and i32 undef, 65535 410 %45 = insertelement <4 x i32> undef, i32 %44, i32 1 411 %46 = load <4 x i32>, <4 x i32> addrspace(4)* undef, align 16 412 %47 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %46, i32 0, i32 0, i32 0, i32 0) 413 %48 = add i32 %47, -1 414 %49 = shl i32 %0, 4 415 %50 = call i32 @llvm.amdgcn.readfirstlane(i32 %49) 416 %51 = sext i32 %50 to i64 417 %52 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 418 %53 = add i32 %52, -2 419 %54 = or i32 %53, %48 420 %55 = shl i32 %1, 4 421 %56 = call i32 @llvm.amdgcn.readfirstlane(i32 %55) 422 %57 = sext i32 %56 to i64 423 %58 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 424 %59 = add i32 %58, -3 425 %60 = or i32 %54, %59 426 %61 = shl i32 %2, 4 427 %62 = call i32 @llvm.amdgcn.readfirstlane(i32 %61) 428 %63 = sext i32 %62 to i64 429 %64 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 430 %65 = add i32 %64, -4 431 %66 = or i32 %60, %65 432 %67 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 433 %68 = add i32 %67, -27 434 %69 = or i32 %66, %68 435 %70 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> undef, i32 0, i32 0, i32 0) 436 %71 = add i32 %70, -28 437 %72 = or i32 %69, %71 438 %73 = call i32 @llvm.amdgcn.readfirstlane(i32 %0) 439 %74 = getelementptr i8, i8 addrspace(4)* %35, i64 16 440 %75 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 0, i32 0) 441 %76 = add i32 %75, -29 442 %77 = or i32 %72, %76 443 %78 = call i32 @llvm.amdgcn.readfirstlane(i32 %1) 444 %79 = shl i32 %78, 4 445 %80 = sext i32 %79 to i64 446 %81 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 0, i32 0) 447 %82 = add i32 %81, -30 448 %83 = or i32 %77, %82 449 %84 = call i32 @llvm.amdgcn.readfirstlane(i32 %2) 450 %85 = shl i32 %84, 4 451 %86 = sext i32 %85 to i64 452 %87 = getelementptr i8, i8 addrspace(4)* %74, i64 %86 453 %88 = bitcast i8 addrspace(4)* %87 to <4 x i32> addrspace(4)* 454 %89 = load <4 x i32>, <4 x i32> addrspace(4)* %88, align 16 455 %90 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %89, i32 0, i32 0) 456 %91 = add i32 %90, -31 457 %92 = or i32 %83, %91 458 %93 = getelementptr i8, i8 addrspace(4)* %35, i64 64 459 %94 = getelementptr i8, i8 addrspace(4)* %93, i64 %51 460 %95 = bitcast i8 addrspace(4)* %94 to <4 x i32> addrspace(4)* 461 %96 = load <4 x i32>, <4 x i32> addrspace(4)* %95, align 16 462 %97 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %96, i32 0, i32 0, i32 0, i32 0) 463 %98 = add i32 %97, -32 464 %99 = or i32 %92, %98 465 %100 = getelementptr i8, i8 addrspace(4)* %93, i64 %57 466 %101 = bitcast i8 addrspace(4)* %100 to <4 x i32> addrspace(4)* 467 %102 = load <4 x i32>, <4 x i32> addrspace(4)* %101, align 16 468 %103 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %102, i32 0, i32 0, i32 0, i32 0) 469 %104 = add i32 %103, -33 470 %105 = or i32 %99, %104 471 %106 = getelementptr i8, i8 addrspace(4)* %93, i64 %63 472 %107 = bitcast i8 addrspace(4)* %106 to <4 x i32> addrspace(4)* 473 %108 = load <4 x i32>, <4 x i32> addrspace(4)* %107, align 16 474 %109 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %108, i32 0, i32 0, i32 0, i32 0) 475 %110 = add i32 %109, -34 476 %111 = or i32 %105, %110 477 %112 = call i32 @llvm.amdgcn.readfirstlane(i32 undef) 478 %113 = sext i32 %112 to i64 479 %114 = getelementptr i8, i8 addrspace(4)* %93, i64 %113 480 %115 = bitcast i8 addrspace(4)* %114 to <4 x i32> addrspace(4)* 481 %116 = load <4 x i32>, <4 x i32> addrspace(4)* %115, align 16 482 %117 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %116, i32 0, i32 0, i32 0, i32 0) 483 %118 = add i32 %117, -36 484 %119 = or i32 %111, %118 485 %120 = getelementptr i8, i8 addrspace(4)* %32, i64 %51 486 %121 = bitcast i8 addrspace(4)* %120 to <4 x i32> addrspace(4)* 487 %122 = load <4 x i32>, <4 x i32> addrspace(4)* %121, align 16 488 %123 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %122, i32 0, i32 0, i32 0, i32 0) 489 %124 = add i32 %123, -37 490 %125 = or i32 %119, %124 491 %126 = getelementptr i8, i8 addrspace(4)* %32, i64 %57 492 %127 = bitcast i8 addrspace(4)* %126 to <4 x i32> addrspace(4)* 493 %128 = load <4 x i32>, <4 x i32> addrspace(4)* %127, align 16 494 %129 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %128, i32 0, i32 0, i32 0, i32 0) 495 %130 = add i32 %129, -38 496 %131 = or i32 %125, %130 497 %132 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 498 %133 = add i32 %132, -39 499 %134 = or i32 %131, %133 500 %135 = call i32 @llvm.amdgcn.readfirstlane(i32 undef) 501 %136 = sext i32 %135 to i64 502 %137 = getelementptr i8, i8 addrspace(4)* %32, i64 %136 503 %138 = bitcast i8 addrspace(4)* %137 to <4 x i32> addrspace(4)* 504 %139 = load <4 x i32>, <4 x i32> addrspace(4)* %138, align 16 505 %140 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %139, i32 0, i32 0, i32 0, i32 0) 506 %141 = add i32 %140, -50 507 %142 = or i32 %134, %141 508 %143 = getelementptr i8, i8 addrspace(4)* %32, i64 224 509 %144 = getelementptr i8, i8 addrspace(4)* %143, i64 %51 510 %145 = bitcast i8 addrspace(4)* %144 to <4 x i32> addrspace(4)* 511 %146 = load <4 x i32>, <4 x i32> addrspace(4)* %145, align 16 512 %147 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %146, i32 0, i32 0, i32 0, i32 0) 513 %148 = add i32 %147, -51 514 %149 = or i32 %142, %148 515 %150 = getelementptr i8, i8 addrspace(4)* %143, i64 %57 516 %151 = bitcast i8 addrspace(4)* %150 to <4 x i32> addrspace(4)* 517 %152 = load <4 x i32>, <4 x i32> addrspace(4)* %151, align 16 518 %153 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %152, i32 0, i32 0, i32 0, i32 0) 519 %154 = add i32 %153, -52 520 %155 = or i32 %149, %154 521 %156 = getelementptr i8, i8 addrspace(4)* %143, i64 %63 522 %157 = bitcast i8 addrspace(4)* %156 to <4 x i32> addrspace(4)* 523 %158 = load <4 x i32>, <4 x i32> addrspace(4)* %157, align 16 524 %159 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %158, i32 0, i32 0, i32 0, i32 0) 525 %160 = add i32 %159, -53 526 %161 = or i32 %155, %160 527 %162 = sext i32 undef to i64 528 %163 = getelementptr i8, i8 addrspace(4)* %143, i64 %162 529 %164 = bitcast i8 addrspace(4)* %163 to <4 x i32> addrspace(4)* 530 %165 = load <4 x i32>, <4 x i32> addrspace(4)* %164, align 16 531 %166 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %165, i32 0, i32 0, i32 0, i32 0) 532 %167 = add i32 %166, -72 533 %168 = or i32 %161, %167 534 %169 = getelementptr i8, i8 addrspace(4)* %32, i64 576 535 %170 = getelementptr i8, i8 addrspace(4)* %169, i64 %51 536 %171 = bitcast i8 addrspace(4)* %170 to <4 x i32> addrspace(4)* 537 %172 = load <4 x i32>, <4 x i32> addrspace(4)* %171, align 16 538 %173 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %172, i32 0, i32 0, i32 0, i32 0) 539 %174 = add i32 %173, -73 540 %175 = or i32 %168, %174 541 %176 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 542 %177 = add i32 %176, -74 543 %178 = or i32 %175, %177 544 %179 = getelementptr i8, i8 addrspace(4)* %169, i64 %63 545 %180 = bitcast i8 addrspace(4)* %179 to <4 x i32> addrspace(4)* 546 %181 = load <4 x i32>, <4 x i32> addrspace(4)* %180, align 16 547 %182 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %181, i32 0, i32 0, i32 0, i32 0) 548 %183 = add i32 %182, -75 549 %184 = or i32 %178, %183 550 %185 = getelementptr i8, i8 addrspace(4)* %169, i64 %113 551 %186 = bitcast i8 addrspace(4)* %185 to <4 x i32> addrspace(4)* 552 %187 = load <4 x i32>, <4 x i32> addrspace(4)* %186, align 16 553 %188 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %187, i32 0, i32 0, i32 0, i32 0) 554 %189 = add i32 %188, -77 555 %190 = or i32 %184, %189 556 %191 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 557 %192 = add i32 %191, -93 558 %193 = or i32 %190, %192 559 %194 = inttoptr i64 %29 to i8 addrspace(4)* 560 %195 = getelementptr i8, i8 addrspace(4)* %194, i64 %51 561 %196 = bitcast i8 addrspace(4)* %195 to <4 x i32> addrspace(4)* 562 %197 = load <4 x i32>, <4 x i32> addrspace(4)* %196, align 16 563 %198 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %197, i32 0, i32 0, i32 0, i32 0) 564 %199 = add i32 %198, -94 565 %200 = or i32 %193, %199 566 %201 = load <4 x i32>, <4 x i32> addrspace(4)* undef, align 16 567 %202 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %201, i32 0, i32 0, i32 0) 568 %203 = add i32 %202, -95 569 %204 = or i32 %200, %203 570 %205 = getelementptr i8, i8 addrspace(4)* %27, i64 %80 571 %206 = bitcast i8 addrspace(4)* %205 to <4 x i32> addrspace(4)* 572 %207 = load <4 x i32>, <4 x i32> addrspace(4)* %206, align 16 573 %208 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %207, i32 0, i32 0, i32 0) 574 %209 = add i32 %208, -96 575 %210 = or i32 %204, %209 576 %211 = getelementptr i8, i8 addrspace(4)* %27, i64 %86 577 %212 = bitcast i8 addrspace(4)* %211 to <4 x i32> addrspace(4)* 578 %213 = load <4 x i32>, <4 x i32> addrspace(4)* %212, align 16 579 %214 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %213, i32 0, i32 0, i32 0) 580 %215 = add i32 %214, -97 581 %216 = or i32 %210, %215 582 %217 = getelementptr <{ [4 x i32], [6 x %llpc.array.element] }>, <{ [4 x i32], [6 x %llpc.array.element] }> addrspace(6)* null, i32 0, i32 1, i32 %0, i32 0 583 %218 = ptrtoint i32 addrspace(6)* %217 to i32 584 %219 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 %218, i32 0) 585 %220 = add i32 %219, -98 586 %221 = or i32 %216, %220 587 %222 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 undef, i32 0) 588 %223 = add i32 %222, -114 589 %224 = or i32 %221, %223 590 %225 = getelementptr <{ [4 x i32], [6 x %llpc.array.element] }>, <{ [4 x i32], [6 x %llpc.array.element] }> addrspace(6)* null, i32 0, i32 1, i32 %2, i32 0 591 %226 = ptrtoint i32 addrspace(6)* %225 to i32 592 %227 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 %226, i32 0) 593 %228 = add i32 %227, -130 594 %229 = or i32 %224, %228 595 %230 = getelementptr <{ [4 x i32], [6 x %llpc.array.element] }>, <{ [4 x i32], [6 x %llpc.array.element] }> addrspace(6)* null, i32 0, i32 1, i32 undef, i32 0 596 %231 = ptrtoint i32 addrspace(6)* %230 to i32 597 %232 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 %231, i32 0) 598 %233 = add i32 %232, -178 599 %234 = or i32 %229, %233 600 %235 = inttoptr i64 %24 to i8 addrspace(4)* 601 %236 = getelementptr i8, i8 addrspace(4)* %235, i64 %51 602 %237 = bitcast i8 addrspace(4)* %236 to <4 x i32> addrspace(4)* 603 %238 = load <4 x i32>, <4 x i32> addrspace(4)* %237, align 16 604 %239 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %238, i32 0, i32 0, i32 0, i32 0) 605 %240 = add i32 %239, -194 606 %241 = or i32 %234, %240 607 %242 = inttoptr i64 %22 to i8 addrspace(4)* 608 %243 = getelementptr i8, i8 addrspace(4)* %242, i64 %51 609 %244 = bitcast i8 addrspace(4)* %243 to <4 x i32> addrspace(4)* 610 %245 = load <4 x i32>, <4 x i32> addrspace(4)* %244, align 16 611 %246 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %245, i32 0, i32 0, i32 0, i32 0) 612 %247 = add i32 %246, -195 613 %248 = or i32 %241, %247 614 %249 = getelementptr i8, i8 addrspace(4)* %242, i64 %57 615 %250 = bitcast i8 addrspace(4)* %249 to <4 x i32> addrspace(4)* 616 %251 = load <4 x i32>, <4 x i32> addrspace(4)* %250, align 16 617 %252 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %251, i32 0, i32 0, i32 0, i32 0) 618 %253 = add i32 %252, -196 619 %254 = or i32 %248, %253 620 %255 = getelementptr i8, i8 addrspace(4)* %242, i64 %63 621 %256 = bitcast i8 addrspace(4)* %255 to <4 x i32> addrspace(4)* 622 %257 = load <4 x i32>, <4 x i32> addrspace(4)* %256, align 16 623 %258 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %257, i32 0, i32 0, i32 0, i32 0) 624 %259 = add i32 %258, -197 625 %260 = or i32 %254, %259 626 %261 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 627 %262 = add i32 %261, -216 628 %263 = or i32 %260, %262 629 %264 = getelementptr <{ [4 x i32], [6 x %llpc.array.element.2] }>, <{ [4 x i32], [6 x %llpc.array.element.2] }> addrspace(6)* null, i32 0, i32 1, i32 %0, i32 0 630 %265 = ptrtoint i32 addrspace(6)* %264 to i32 631 %266 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %265, i32 0) 632 %267 = add i32 %266, -217 633 %268 = or i32 %263, %267 634 %269 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 635 %270 = add i32 %269, -233 636 %271 = or i32 %268, %270 637 %272 = getelementptr <{ [4 x i32], [6 x %llpc.array.element.2] }>, <{ [4 x i32], [6 x %llpc.array.element.2] }> addrspace(6)* null, i32 0, i32 1, i32 %2, i32 0 638 %273 = ptrtoint i32 addrspace(6)* %272 to i32 639 %274 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %273, i32 0) 640 %275 = add i32 %274, -249 641 %276 = or i32 %271, %275 642 %277 = getelementptr <{ [4 x i32], [6 x %llpc.array.element.2] }>, <{ [4 x i32], [6 x %llpc.array.element.2] }> addrspace(6)* null, i32 0, i32 1, i32 undef, i32 0 643 %278 = ptrtoint i32 addrspace(6)* %277 to i32 644 %279 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %278, i32 0) 645 %280 = add i32 %279, -297 646 %281 = or i32 %276, %280 647 %282 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 648 %283 = add i32 %282, -313 649 %284 = or i32 %281, %283 650 %285 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 651 %286 = add i32 %285, -329 652 %287 = or i32 %284, %286 653 %288 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 654 %289 = add i32 %288, -345 655 %290 = or i32 %287, %289 656 %291 = getelementptr <{ [4 x i32], [9 x %llpc.array.element.5] }>, <{ [4 x i32], [9 x %llpc.array.element.5] }> addrspace(6)* null, i32 0, i32 1, i32 %4, i32 0 657 %292 = ptrtoint i32 addrspace(6)* %291 to i32 658 %293 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %292, i32 0) 659 %294 = add i32 %293, -441 660 %295 = or i32 %290, %294 661 %296 = getelementptr i8, i8 addrspace(4)* %20, i64 160 662 %297 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 663 %298 = add i32 %297, -457 664 %299 = or i32 %295, %298 665 %300 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 666 %301 = add i32 %300, -458 667 %302 = or i32 %299, %301 668 %303 = getelementptr i8, i8 addrspace(4)* %296, i64 %63 669 %304 = bitcast i8 addrspace(4)* %303 to <4 x i32> addrspace(4)* 670 %305 = load <4 x i32>, <4 x i32> addrspace(4)* %304, align 16 671 %306 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %305, i32 0, i32 0, i32 0, i32 0) 672 %307 = add i32 %306, -459 673 %308 = or i32 %302, %307 674 %309 = shl i32 %5, 4 675 %310 = call i32 @llvm.amdgcn.readfirstlane(i32 %309) 676 %311 = sext i32 %310 to i64 677 %312 = getelementptr i8, i8 addrspace(4)* %296, i64 %311 678 %313 = bitcast i8 addrspace(4)* %312 to <4 x i32> addrspace(4)* 679 %314 = load <4 x i32>, <4 x i32> addrspace(4)* %313, align 16 680 %315 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %314, i32 0, i32 0, i32 0, i32 0) 681 %316 = add i32 %315, -466 682 %317 = or i32 %308, %316 683 %318 = getelementptr i8, i8 addrspace(4)* %38, i64 168 684 %319 = shl i32 %73, 3 685 %320 = sext i32 %319 to i64 686 %321 = getelementptr i8, i8 addrspace(4)* %318, i64 %320 687 %.i085.i = bitcast i8 addrspace(4)* %321 to i32 addrspace(4)* 688 %.ii0.i = load i32, i32 addrspace(4)* %.i085.i, align 8 689 %322 = and i32 undef, 65535 690 %323 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii0.i, i32 0 691 %324 = insertelement <4 x i32> %323, i32 %322, i32 1 692 %325 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %324, i32 0, i32 0) 693 %326 = add i32 %325, -467 694 %327 = or i32 %317, %326 695 %328 = shl i32 %78, 3 696 %329 = sext i32 %328 to i64 697 %330 = getelementptr i8, i8 addrspace(4)* %318, i64 %329 698 %.i088.i = bitcast i8 addrspace(4)* %330 to i32 addrspace(4)* 699 %.ii090.i = load i32, i32 addrspace(4)* %.i088.i, align 8 700 %.i191.i = getelementptr i8, i8 addrspace(4)* %330, i64 4 701 %331 = bitcast i8 addrspace(4)* %.i191.i to i32 addrspace(4)* 702 %.ii192.i = load i32, i32 addrspace(4)* %331, align 4 703 %332 = and i32 %.ii192.i, 65535 704 %333 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii090.i, i32 0 705 %334 = insertelement <4 x i32> %333, i32 %332, i32 1 706 %335 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %334, i32 0, i32 0) 707 %336 = add i32 %335, -468 708 %337 = or i32 %327, %336 709 %338 = shl i32 %84, 3 710 %339 = sext i32 %338 to i64 711 %340 = getelementptr i8, i8 addrspace(4)* %318, i64 %339 712 %.i094.i = bitcast i8 addrspace(4)* %340 to i32 addrspace(4)* 713 %.ii096.i = load i32, i32 addrspace(4)* %.i094.i, align 8 714 %.i197.i = getelementptr i8, i8 addrspace(4)* %340, i64 4 715 %341 = bitcast i8 addrspace(4)* %.i197.i to i32 addrspace(4)* 716 %.ii198.i = load i32, i32 addrspace(4)* %341, align 4 717 %342 = and i32 %.ii198.i, 65535 718 %343 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii096.i, i32 0 719 %344 = insertelement <4 x i32> %343, i32 %342, i32 1 720 %345 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %344, i32 0, i32 0) 721 %346 = add i32 %345, -469 722 %347 = or i32 %337, %346 723 %348 = call i32 @llvm.amdgcn.readfirstlane(i32 %3) 724 %349 = shl i32 %348, 3 725 %350 = sext i32 %349 to i64 726 %351 = getelementptr i8, i8 addrspace(4)* %318, i64 %350 727 %.i0100.i = bitcast i8 addrspace(4)* %351 to i32 addrspace(4)* 728 %.ii0102.i = load i32, i32 addrspace(4)* %.i0100.i, align 8 729 %.ii1104.i = load i32, i32 addrspace(4)* undef, align 4 730 %352 = and i32 %.ii1104.i, 65535 731 %353 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii0102.i, i32 0 732 %354 = insertelement <4 x i32> %353, i32 %352, i32 1 733 %355 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %354, i32 0, i32 0) 734 %356 = add i32 %355, -473 735 %357 = or i32 %347, %356 736 %358 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 0, i32 0) 737 %359 = add i32 %358, -474 738 %360 = or i32 %357, %359 739 %361 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 740 %362 = add i32 %361, -475 741 %363 = or i32 %360, %362 742 %364 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 743 %365 = add i32 %364, -491 744 %366 = or i32 %363, %365 745 %367 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 746 %368 = add i32 %367, -507 747 %369 = or i32 %366, %368 748 %370 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0) 749 %371 = add i32 %370, -539 750 %372 = or i32 %369, %371 751 %373 = getelementptr i8, i8 addrspace(4)* %17, i64 96 752 %374 = getelementptr i8, i8 addrspace(4)* %373, i64 %51 753 %375 = bitcast i8 addrspace(4)* %374 to <4 x i32> addrspace(4)* 754 %376 = load <4 x i32>, <4 x i32> addrspace(4)* %375, align 16 755 %377 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %376, i32 0, i32 0, i32 0, i32 0) 756 %378 = add i32 %377, -555 757 %379 = or i32 %372, %378 758 %380 = getelementptr i8, i8 addrspace(4)* %373, i64 %57 759 %381 = bitcast i8 addrspace(4)* %380 to <4 x i32> addrspace(4)* 760 %382 = load <4 x i32>, <4 x i32> addrspace(4)* %381, align 16 761 %383 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %382, i32 0, i32 0, i32 0, i32 0) 762 %384 = add i32 %383, -556 763 %385 = or i32 %379, %384 764 %386 = getelementptr i8, i8 addrspace(4)* %373, i64 %63 765 %387 = bitcast i8 addrspace(4)* %386 to <4 x i32> addrspace(4)* 766 %388 = load <4 x i32>, <4 x i32> addrspace(4)* %387, align 16 767 %389 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %388, i32 0, i32 0, i32 0, i32 0) 768 %390 = add i32 %389, -557 769 %391 = or i32 %385, %390 770 %392 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 771 %393 = add i32 %392, -574 772 %394 = or i32 %391, %393 773 %395 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 774 %396 = add i32 %395, -575 775 %397 = or i32 %394, %396 776 %398 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 777 %399 = add i32 %398, -576 778 %400 = or i32 %397, %399 779 %401 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 780 %402 = add i32 %401, -577 781 %403 = or i32 %400, %402 782 %404 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0) 783 %405 = add i32 %404, -593 784 %406 = or i32 %403, %405 785 %407 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %43, i32 0, i32 0) 786 %408 = add i32 %407, -594 787 %409 = or i32 %406, %408 788 %.not.i = icmp eq i32 %409, 0 789 %410 = load <8 x i32>, <8 x i32> addrspace(4)* undef, align 32 790 %.i010.i = select i1 %.not.i, float 0x36A0000000000000, float 0.000000e+00 791 %411 = insertelement <4 x float> undef, float %.i010.i, i32 3 792 call void @llvm.amdgcn.image.store.2d.v4f32.i32(<4 x float> %411, i32 15, i32 undef, i32 undef, <8 x i32> %410, i32 0, i32 0) 793 ret void 794} 795 796declare i32 @llvm.amdgcn.readfirstlane(i32) 797declare void @llvm.amdgcn.image.store.2d.v4f32.i32(<4 x float>, i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) 798declare void @llvm.amdgcn.exp.f32(i32 immarg, i32 immarg, float, float, float, float, i1 immarg, i1 immarg) 799declare i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32>, i32, i32, i32, i32 immarg) 800declare i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32>, i32, i32, i32 immarg) 801declare i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32>, i32, i32 immarg) 802declare i32 @llvm.amdgcn.mbcnt.lo(i32, i32) 803