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
4 ; TODO: This was introduced in D88020 to catch a case that some unreachable
5 ; assert was hit during liverange split. But after D104509, there is some IR
6 ; change after register coalescer which make the case not work as before. We
7 ; need to find some other way to reproduce the bad case fixed by D88020.
9 %llpc.array.element = type <{ i32, [12 x i8] }>
10 %llpc.array.element.2 = type <{ i32, [12 x i8] }>
11 %llpc.array.element.5 = type <{ i32, [12 x i8] }>
13 define amdgpu_gs void @_amdgpu_gs_main(i32 inreg %primShaderTableAddrLow, <31 x i32> inreg %userData) {
14 ; CHECK-LABEL: name: _amdgpu_gs_main
15 ; CHECK: bb.0..expVert:
16 ; CHECK: liveins: $sgpr3, $sgpr4, $sgpr5, $sgpr8, $sgpr9, $sgpr10, $sgpr18, $sgpr19, $sgpr20, $sgpr21, $sgpr22, $sgpr23, $sgpr25, $sgpr27, $sgpr31
17 ; CHECK: undef %56.sub0:sgpr_64 = COPY $sgpr31
18 ; CHECK: [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr27
19 ; CHECK: [[COPY1:%[0-9]+]]:sgpr_32 = COPY $sgpr25
20 ; CHECK: [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr5
21 ; CHECK: [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr4
22 ; CHECK: [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr3
23 ; CHECK: [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr18
24 ; CHECK: undef %50.sub0:sgpr_64 = COPY $sgpr19
25 ; CHECK: [[COPY6:%[0-9]+]]:sgpr_32 = COPY $sgpr20
26 ; CHECK: [[COPY7:%[0-9]+]]:sgpr_32 = COPY $sgpr21
27 ; CHECK: [[COPY8:%[0-9]+]]:sgpr_32 = COPY $sgpr22
28 ; CHECK: [[COPY9:%[0-9]+]]:sgpr_32 = COPY $sgpr23
29 ; CHECK: [[COPY10:%[0-9]+]]:sgpr_32 = COPY $sgpr9
30 ; CHECK: [[COPY11:%[0-9]+]]:sgpr_32 = COPY $sgpr10
31 ; CHECK: [[COPY12:%[0-9]+]]:sgpr_32 = COPY $sgpr8
32 ; CHECK: undef %71.sub0_sub1:sgpr_128 = S_LOAD_DWORDX2_IMM %56, 232, 0 :: (load (s64) from %ir.40, addrspace 4)
33 ; CHECK: [[S_MOV_B32_:%[0-9]+]]:sreg_32 = S_MOV_B32 65535
34 ; CHECK: [[S_LSHL_B32_:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY4]], 4, implicit-def dead $scc
35 ; CHECK: [[S_LSHL_B32_1:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY3]], 4, implicit-def dead $scc
36 ; CHECK: [[S_LSHL_B32_2:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY2]], 4, implicit-def dead $scc
37 ; CHECK: [[S_ASHR_I32_:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_]], 31, implicit-def dead $scc
38 ; CHECK: [[S_ASHR_I32_1:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_1]], 31, implicit-def dead $scc
39 ; CHECK: %71.sub1:sgpr_128 = S_AND_B32 %71.sub1, [[S_MOV_B32_]], implicit-def dead $scc
40 ; CHECK: [[S_ASHR_I32_2:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_2]], 31, implicit-def dead $scc
41 ; CHECK: undef %130.sub0:sreg_64 = S_ADD_U32 [[COPY5]], [[S_LSHL_B32_2]], implicit-def $scc
42 ; CHECK: %130.sub1:sreg_64 = S_ADDC_U32 undef %54:sreg_32, [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
43 ; CHECK: [[S_LOAD_DWORDX4_IMM:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %130, 16, 0 :: (load (s128) from %ir.84, addrspace 4)
44 ; CHECK: [[S_LOAD_DWORDX4_IMM1:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM undef %74:sreg_64, 0, 0 :: (load (s128) from `<4 x i32> addrspace(4)* undef`, addrspace 4)
45 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM undef %132:sgpr_128, 0, 0 :: (dereferenceable invariant load (s32))
46 ; CHECK: KILL undef %74:sreg_64
47 ; CHECK: KILL undef %132:sgpr_128
48 ; CHECK: KILL %130.sub0, %130.sub1
49 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM1:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM [[S_LOAD_DWORDX4_IMM]], 0, 0 :: (dereferenceable invariant load (s32))
50 ; CHECK: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
51 ; CHECK: undef %302.sub1:sgpr_128 = S_MOV_B32 0
52 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
53 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
54 ; CHECK: KILL undef %89:sgpr_128
55 ; CHECK: [[S_SUB_I32_:%[0-9]+]]:sreg_32 = S_SUB_I32 [[S_BUFFER_LOAD_DWORD_IMM]], 29, implicit-def dead $scc
56 ; CHECK: [[S_SUB_I32_1:%[0-9]+]]:sreg_32 = S_SUB_I32 [[S_BUFFER_LOAD_DWORD_IMM]], 30, implicit-def dead $scc
57 ; CHECK: [[S_SUB_I32_2:%[0-9]+]]:sreg_32 = S_SUB_I32 [[S_BUFFER_LOAD_DWORD_IMM1]], 31, implicit-def dead $scc
58 ; CHECK: [[S_ADD_U32_:%[0-9]+]]:sreg_32 = S_ADD_U32 [[COPY5]], 64, implicit-def $scc
59 ; CHECK: [[S_ADDC_U32_:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %54:sreg_32, 0, implicit-def dead $scc, implicit $scc
60 ; CHECK: undef %149.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], [[S_LSHL_B32_]], implicit-def $scc
61 ; CHECK: %149.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
62 ; CHECK: undef %156.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], [[S_LSHL_B32_1]], implicit-def $scc
63 ; CHECK: [[S_LOAD_DWORDX4_IMM2:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %149, 0, 0 :: (load (s128) from %ir.91, addrspace 4)
64 ; CHECK: %156.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc
65 ; CHECK: undef %163.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], [[S_LSHL_B32_2]], implicit-def $scc
66 ; CHECK: %163.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
67 ; CHECK: [[S_ASHR_I32_3:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 undef %171:sreg_32, 31, implicit-def dead $scc
68 ; CHECK: undef %176.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_]], undef %171:sreg_32, implicit-def $scc
69 ; CHECK: %176.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_]], [[S_ASHR_I32_3]], implicit-def dead $scc, implicit $scc
70 ; CHECK: undef %183.sub0:sreg_64 = S_ADD_U32 %50.sub0, [[S_LSHL_B32_]], implicit-def $scc
71 ; CHECK: %183.sub1:sreg_64 = S_ADDC_U32 undef %51:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
72 ; CHECK: undef %190.sub0:sreg_64 = S_ADD_U32 %50.sub0, [[S_LSHL_B32_1]], implicit-def $scc
73 ; CHECK: %190.sub1:sreg_64 = S_ADDC_U32 undef %51:sreg_32, [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc
74 ; CHECK: undef %200.sub0:sreg_64 = S_ADD_U32 %50.sub0, undef %171:sreg_32, implicit-def $scc
75 ; CHECK: %200.sub1:sreg_64 = S_ADDC_U32 undef %51:sreg_32, [[S_ASHR_I32_3]], implicit-def dead $scc, implicit $scc
76 ; CHECK: [[S_ADD_U32_1:%[0-9]+]]:sreg_32 = S_ADD_U32 %50.sub0, 224, implicit-def $scc
77 ; CHECK: [[S_ADDC_U32_1:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %51:sreg_32, 0, implicit-def dead $scc, implicit $scc
78 ; CHECK: undef %210.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_1]], [[S_LSHL_B32_]], implicit-def $scc
79 ; CHECK: %210.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_1]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
80 ; CHECK: undef %217.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_1]], [[S_LSHL_B32_1]], implicit-def $scc
81 ; CHECK: %217.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_1]], [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc
82 ; CHECK: undef %224.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_1]], [[S_LSHL_B32_2]], implicit-def $scc
83 ; CHECK: %224.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_1]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
84 ; CHECK: [[S_ADD_U32_2:%[0-9]+]]:sreg_32 = S_ADD_U32 %50.sub0, 576, implicit-def $scc
85 ; CHECK: [[S_ADDC_U32_2:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %51:sreg_32, 0, implicit-def dead $scc, implicit $scc
86 ; CHECK: undef %241.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_2]], [[S_LSHL_B32_]], implicit-def $scc
87 ; CHECK: %241.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_2]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
88 ; CHECK: undef %253.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_2]], [[S_LSHL_B32_2]], implicit-def $scc
89 ; CHECK: %253.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_2]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
90 ; CHECK: undef %261.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_2]], undef %171:sreg_32, implicit-def $scc
91 ; CHECK: %261.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_2]], [[S_ASHR_I32_3]], implicit-def dead $scc, implicit $scc
92 ; CHECK: undef %273.sub0:sreg_64 = S_ADD_U32 [[COPY6]], [[S_LSHL_B32_]], implicit-def $scc
93 ; CHECK: %273.sub1:sreg_64 = S_ADDC_U32 undef %48:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
94 ; CHECK: undef %286.sub0:sreg_64 = S_ADD_U32 [[COPY7]], [[S_LSHL_B32_1]], implicit-def $scc
95 ; CHECK: %286.sub1:sreg_64 = S_ADDC_U32 undef %45:sreg_32, [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc
96 ; CHECK: undef %293.sub0:sreg_64 = S_ADD_U32 [[COPY7]], [[S_LSHL_B32_2]], implicit-def $scc
97 ; CHECK: %293.sub1:sreg_64 = S_ADDC_U32 undef %45:sreg_32, [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
98 ; CHECK: [[S_ADD_I32_:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_LSHL_B32_]], 16, implicit-def dead $scc
99 ; CHECK: [[S_ADD_I32_1:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_LSHL_B32_2]], 16, implicit-def dead $scc
100 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR %302, [[S_ADD_I32_]], 0 :: (dereferenceable invariant load (s32))
101 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR1:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR %302, undef %314:sreg_32, 0 :: (dereferenceable invariant load (s32))
102 ; CHECK: [[S_BUFFER_LOAD_DWORD_SGPR2:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_SGPR %302, [[S_ADD_I32_1]], 0 :: (dereferenceable invariant load (s32))
103 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM2:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %302, 16, 0 :: (dereferenceable invariant load (s32))
104 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET undef %118:sgpr_128, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
105 ; 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 :: (dereferenceable invariant load (s32))
106 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM3:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM undef %380:sgpr_128, 16, 0 :: (dereferenceable invariant load (s32))
107 ; CHECK: [[S_LOAD_DWORDX4_IMM3:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %156, 0, 0 :: (load (s128) from %ir.97, addrspace 4)
108 ; CHECK: [[S_LOAD_DWORDX4_IMM4:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %163, 0, 0 :: (load (s128) from %ir.103, addrspace 4)
109 ; CHECK: [[S_LOAD_DWORDX4_IMM5:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %176, 0, 0 :: (load (s128) from %ir.111, addrspace 4)
110 ; CHECK: [[S_LOAD_DWORDX4_IMM6:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %183, 0, 0 :: (load (s128) from %ir.117, addrspace 4)
111 ; CHECK: [[S_LOAD_DWORDX4_IMM7:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %190, 0, 0 :: (load (s128) from %ir.123, addrspace 4)
112 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
113 ; 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 :: (dereferenceable invariant load (s32))
114 ; 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 :: (dereferenceable invariant load (s32))
115 ; CHECK: [[S_ADD_I32_2:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR]], -98, implicit-def dead $scc
116 ; CHECK: [[S_ADD_I32_3:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR1]], -114, implicit-def dead $scc
117 ; CHECK: [[S_ADD_I32_4:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR2]], -130, implicit-def dead $scc
118 ; CHECK: [[S_ADD_I32_5:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM2]], -178, implicit-def dead $scc
119 ; CHECK: undef %327.sub0:sreg_64 = S_ADD_U32 [[COPY8]], [[S_LSHL_B32_]], implicit-def $scc
120 ; CHECK: %327.sub1:sreg_64 = S_ADDC_U32 undef %42:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
121 ; CHECK: undef %335.sub0:sreg_64 = S_ADD_U32 [[COPY9]], [[S_LSHL_B32_]], implicit-def $scc
122 ; CHECK: %335.sub1:sreg_64 = S_ADDC_U32 undef %39:sreg_32, [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
123 ; CHECK: undef %343.sub0:sreg_64 = S_ADD_U32 [[COPY9]], [[S_LSHL_B32_1]], implicit-def $scc
124 ; CHECK: [[S_LOAD_DWORDX4_IMM8:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %200, 0, 0 :: (load (s128) from %ir.131, addrspace 4)
125 ; CHECK: %343.sub1:sreg_64 = S_ADDC_U32 undef %39:sreg_32, [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc
126 ; CHECK: undef %351.sub0:sreg_64 = S_ADD_U32 [[COPY9]], [[S_LSHL_B32_2]], implicit-def $scc
127 ; CHECK: %351.sub1:sreg_64 = S_ADDC_U32 undef %39:sreg_32, [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
128 ; CHECK: [[S_LSHL_B32_3:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY10]], 4, implicit-def dead $scc
129 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
130 ; CHECK: [[S_ADD_I32_6:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_LSHL_B32_3]], 16, implicit-def dead $scc
131 ; 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 :: (dereferenceable invariant load (s32))
132 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
133 ; CHECK: [[S_LOAD_DWORDX4_IMM9:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %50, 224, 0 :: (load (s128) from %ir.155, addrspace 4)
134 ; CHECK: [[S_LOAD_DWORDX4_IMM10:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %210, 0, 0 :: (load (s128) from %ir.138, addrspace 4)
135 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
136 ; CHECK: [[S_LOAD_DWORDX4_IMM11:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %217, 0, 0 :: (load (s128) from %ir.144, addrspace 4)
137 ; CHECK: [[S_LOAD_DWORDX4_IMM12:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %224, 0, 0 :: (load (s128) from %ir.150, addrspace 4)
138 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
139 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
140 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
141 ; CHECK: [[S_ADD_I32_7:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR4]], -217, implicit-def dead $scc
142 ; CHECK: [[S_ADD_I32_8:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -233, implicit-def dead $scc
143 ; CHECK: [[S_ADD_I32_9:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR5]], -249, implicit-def dead $scc
144 ; CHECK: [[S_ADD_I32_10:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM3]], -297, implicit-def dead $scc
145 ; CHECK: [[S_ADD_I32_11:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -313, implicit-def dead $scc
146 ; CHECK: [[S_ADD_I32_12:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -329, implicit-def dead $scc
147 ; CHECK: [[S_ADD_I32_13:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -345, implicit-def dead $scc
148 ; CHECK: [[S_ADD_I32_14:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR6]], -441, implicit-def dead $scc
149 ; CHECK: [[S_ADD_U32_3:%[0-9]+]]:sreg_32 = S_ADD_U32 [[COPY1]], 160, implicit-def $scc
150 ; CHECK: [[S_ADDC_U32_3:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %36:sreg_32, 0, implicit-def dead $scc, implicit $scc
151 ; CHECK: undef %411.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_3]], [[S_LSHL_B32_2]], implicit-def $scc
152 ; CHECK: %411.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_3]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
153 ; CHECK: [[S_LSHL_B32_4:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY11]], 4, implicit-def dead $scc
154 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN9:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM10]], 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
155 ; CHECK: [[S_ASHR_I32_4:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_4]], 31, implicit-def dead $scc
156 ; CHECK: undef %425.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_3]], [[S_LSHL_B32_4]], implicit-def $scc
157 ; CHECK: %425.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_3]], [[S_ASHR_I32_4]], implicit-def dead $scc, implicit $scc
158 ; CHECK: [[S_ADD_U32_4:%[0-9]+]]:sreg_32 = S_ADD_U32 %56.sub0, 168, implicit-def $scc
159 ; CHECK: [[S_ADDC_U32_4:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %57:sreg_32, 0, implicit-def dead $scc, implicit $scc
160 ; CHECK: [[S_LOAD_DWORDX4_IMM13:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %241, 0, 0 :: (load (s128) from %ir.162, addrspace 4)
161 ; CHECK: [[S_LSHL_B32_5:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY4]], 3, implicit-def dead $scc
162 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
163 ; CHECK: [[S_ASHR_I32_5:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_5]], 31, implicit-def dead $scc
164 ; CHECK: undef %441.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_5]], implicit-def $scc
165 ; CHECK: %441.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_5]], implicit-def dead $scc, implicit $scc
166 ; CHECK: [[S_LOAD_DWORD_IMM:%[0-9]+]]:sreg_32_xm0_xexec = S_LOAD_DWORD_IMM %441, 0, 0 :: (load (s32) from %ir..i085.i, align 8, addrspace 4)
167 ; CHECK: [[S_LOAD_DWORDX4_IMM14:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %253, 0, 0 :: (load (s128) from %ir.170, addrspace 4)
168 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
169 ; CHECK: [[S_LOAD_DWORDX4_IMM15:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %261, 0, 0 :: (load (s128) from %ir.176, addrspace 4)
170 ; CHECK: [[BUFFER_LOAD_FORMAT_X_IDXEN12:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_IDXEN [[V_MOV_B32_e32_]], [[S_LOAD_DWORDX4_IMM9]], 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
171 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
172 ; CHECK: %71.sub3:sgpr_128 = S_MOV_B32 553734060
173 ; CHECK: %71.sub2:sgpr_128 = S_MOV_B32 -1
174 ; CHECK: [[COPY13:%[0-9]+]]:sgpr_128 = COPY %71
175 ; CHECK: [[S_LOAD_DWORDX4_IMM16:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %273, 0, 0 :: (load (s128) from %ir.185, addrspace 4)
176 ; CHECK: [[COPY13]].sub1:sgpr_128 = COPY %302.sub1
177 ; CHECK: [[COPY13]].sub0:sgpr_128 = COPY [[S_LOAD_DWORD_IMM]]
178 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM4:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM [[COPY13]], 0, 0 :: (dereferenceable invariant load (s32))
179 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
180 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
181 ; CHECK: [[S_LOAD_DWORDX4_IMM17:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %286, 0, 0 :: (load (s128) from %ir.194, addrspace 4)
182 ; CHECK: [[S_LOAD_DWORDX4_IMM18:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %293, 0, 0 :: (load (s128) from %ir.200, addrspace 4)
183 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
184 ; CHECK: [[S_LSHL_B32_6:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY3]], 3, implicit-def dead $scc
185 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET1:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET [[S_LOAD_DWORDX4_IMM1]], 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
186 ; CHECK: [[S_ASHR_I32_6:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_6]], 31, implicit-def dead $scc
187 ; CHECK: [[S_ADD_I32_15:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM4]], -467, implicit-def dead $scc
188 ; CHECK: undef %453.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_6]], implicit-def $scc
189 ; CHECK: %453.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_6]], implicit-def dead $scc, implicit $scc
190 ; CHECK: [[S_LOAD_DWORDX2_IMM:%[0-9]+]]:sreg_64_xexec = S_LOAD_DWORDX2_IMM %453, 0, 0 :: (load (s64) from %ir.308, addrspace 4)
191 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET2:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET [[S_LOAD_DWORDX4_IMM17]], 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
192 ; CHECK: [[BUFFER_LOAD_DWORD_OFFSET3:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_DWORD_OFFSET [[S_LOAD_DWORDX4_IMM18]], 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
193 ; CHECK: [[S_LOAD_DWORDX4_IMM19:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %327, 0, 0 :: (load (s128) from %ir.223, addrspace 4)
194 ; CHECK: [[S_LOAD_DWORDX4_IMM20:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %335, 0, 0 :: (load (s128) from %ir.230, addrspace 4)
195 ; CHECK: [[COPY14:%[0-9]+]]:sgpr_128 = COPY %71
196 ; CHECK: [[S_LOAD_DWORDX4_IMM21:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %343, 0, 0 :: (load (s128) from %ir.236, addrspace 4)
197 ; CHECK: [[S_AND_B32_:%[0-9]+]]:sreg_32 = S_AND_B32 [[S_LOAD_DWORDX2_IMM]].sub1, [[S_MOV_B32_]], implicit-def dead $scc
198 ; CHECK: [[COPY14]].sub0:sgpr_128 = COPY [[S_LOAD_DWORDX2_IMM]].sub0
199 ; CHECK: [[COPY14]].sub1:sgpr_128 = COPY [[S_AND_B32_]]
200 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM5:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM [[COPY14]], 0, 0 :: (dereferenceable invariant load (s32))
201 ; CHECK: [[S_LOAD_DWORDX4_IMM22:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %351, 0, 0 :: (load (s128) from %ir.242, addrspace 4)
202 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
203 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
204 ; CHECK: [[S_LSHL_B32_7:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY2]], 3, implicit-def dead $scc
205 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
206 ; CHECK: [[S_ASHR_I32_7:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_7]], 31, implicit-def dead $scc
207 ; CHECK: [[S_ADD_I32_16:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM5]], -468, implicit-def dead $scc
208 ; CHECK: undef %468.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_7]], implicit-def $scc
209 ; CHECK: %468.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_7]], implicit-def dead $scc, implicit $scc
210 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
211 ; CHECK: [[S_LOAD_DWORDX2_IMM1:%[0-9]+]]:sreg_64_xexec = S_LOAD_DWORDX2_IMM %468, 0, 0 :: (load (s64) from %ir.320, addrspace 4)
212 ; CHECK: [[COPY15:%[0-9]+]]:sgpr_128 = COPY %71
213 ; CHECK: [[S_AND_B32_1:%[0-9]+]]:sreg_32 = S_AND_B32 [[S_LOAD_DWORDX2_IMM1]].sub1, [[S_MOV_B32_]], implicit-def dead $scc
214 ; CHECK: [[COPY15]].sub0:sgpr_128 = COPY [[S_LOAD_DWORDX2_IMM1]].sub0
215 ; CHECK: [[COPY15]].sub1:sgpr_128 = COPY [[S_AND_B32_1]]
216 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM6:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM [[COPY15]], 0, 0 :: (dereferenceable invariant load (s32))
217 ; CHECK: [[S_LOAD_DWORDX4_IMM23:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %411, 0, 0 :: (load (s128) from %ir.282, addrspace 4)
218 ; CHECK: [[S_LOAD_DWORD_IMM1:%[0-9]+]]:sreg_32_xm0_xexec = S_LOAD_DWORD_IMM undef %488:sreg_64, 0, 0 :: (load (s32) from `i32 addrspace(4)* undef`, addrspace 4)
219 ; CHECK: KILL %411.sub0, %411.sub1
220 ; CHECK: KILL undef %488:sreg_64
221 ; CHECK: KILL [[COPY15]].sub0_sub1, [[COPY15]].sub2_sub3
222 ; CHECK: [[S_LSHL_B32_8:%[0-9]+]]:sreg_32 = S_LSHL_B32 [[COPY12]], 3, implicit-def dead $scc
223 ; CHECK: [[S_LOAD_DWORDX4_IMM24:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %425, 0, 0 :: (load (s128) from %ir.291, addrspace 4)
224 ; CHECK: [[S_ASHR_I32_8:%[0-9]+]]:sreg_32_xm0 = S_ASHR_I32 [[S_LSHL_B32_8]], 31, implicit-def dead $scc
225 ; CHECK: [[S_ADD_I32_17:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM6]], -469, implicit-def dead $scc
226 ; CHECK: undef %485.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_4]], [[S_LSHL_B32_8]], implicit-def $scc
227 ; CHECK: %485.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_4]], [[S_ASHR_I32_8]], implicit-def dead $scc, implicit $scc
228 ; CHECK: [[S_LOAD_DWORD_IMM2:%[0-9]+]]:sreg_32_xm0_xexec = S_LOAD_DWORD_IMM %485, 0, 0 :: (load (s32) from %ir..i0100.i, align 8, addrspace 4)
229 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
230 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
231 ; CHECK: KILL [[S_LOAD_DWORDX4_IMM24]]
232 ; CHECK: KILL [[S_LOAD_DWORDX4_IMM23]]
233 ; CHECK: [[S_AND_B32_2:%[0-9]+]]:sreg_32 = S_AND_B32 [[S_LOAD_DWORD_IMM1]], [[S_MOV_B32_]], implicit-def dead $scc
234 ; CHECK: [[COPY16:%[0-9]+]]:sgpr_128 = COPY %71
235 ; CHECK: [[COPY16]].sub1:sgpr_128 = COPY [[S_AND_B32_2]]
236 ; CHECK: [[COPY16]].sub0:sgpr_128 = COPY [[S_LOAD_DWORD_IMM2]]
237 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM7:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM [[COPY16]], 0, 0 :: (dereferenceable invariant load (s32))
238 ; CHECK: [[S_ADD_I32_18:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM]], -474, implicit-def dead $scc
239 ; CHECK: [[S_ADD_I32_19:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -475, implicit-def dead $scc
240 ; CHECK: [[S_ADD_I32_20:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -491, implicit-def dead $scc
241 ; CHECK: [[S_ADD_I32_21:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -507, implicit-def dead $scc
242 ; CHECK: [[S_ADD_I32_22:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_SGPR3]], -539, implicit-def dead $scc
243 ; CHECK: [[S_ADD_I32_23:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM7]], -473, implicit-def dead $scc
244 ; CHECK: [[S_ADD_U32_5:%[0-9]+]]:sreg_32 = S_ADD_U32 [[COPY]], 96, implicit-def $scc
245 ; CHECK: [[S_ADDC_U32_5:%[0-9]+]]:sreg_32 = S_ADDC_U32 undef %33:sreg_32, 0, implicit-def dead $scc, implicit $scc
246 ; CHECK: undef %514.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_5]], [[S_LSHL_B32_]], implicit-def $scc
247 ; CHECK: %514.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_5]], [[S_ASHR_I32_]], implicit-def dead $scc, implicit $scc
248 ; CHECK: [[S_LOAD_DWORDX4_IMM25:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %514, 0, 0 :: (load (s128) from %ir.351, addrspace 4)
249 ; CHECK: undef %522.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_5]], [[S_LSHL_B32_1]], implicit-def $scc
250 ; CHECK: %522.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_5]], [[S_ASHR_I32_1]], implicit-def dead $scc, implicit $scc
251 ; CHECK: [[S_LOAD_DWORDX4_IMM26:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %522, 0, 0 :: (load (s128) from %ir.357, addrspace 4)
252 ; CHECK: undef %530.sub0:sreg_64 = S_ADD_U32 [[S_ADD_U32_5]], [[S_LSHL_B32_2]], implicit-def $scc
253 ; CHECK: %530.sub1:sreg_64 = S_ADDC_U32 [[S_ADDC_U32_5]], [[S_ASHR_I32_2]], implicit-def dead $scc, implicit $scc
254 ; CHECK: [[S_LOAD_DWORDX4_IMM27:%[0-9]+]]:sgpr_128 = S_LOAD_DWORDX4_IMM %530, 0, 0 :: (load (s128) from %ir.363, addrspace 4)
255 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
256 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
257 ; 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, implicit $exec :: (dereferenceable load (s32) from custom "BufferResource", align 1, addrspace 4)
258 ; CHECK: KILL [[S_LOAD_DWORDX4_IMM27]]
259 ; CHECK: KILL [[S_LOAD_DWORDX4_IMM25]]
260 ; CHECK: KILL [[V_MOV_B32_e32_]]
261 ; CHECK: KILL [[S_LOAD_DWORDX4_IMM26]]
262 ; CHECK: [[V_ADD_U32_e32_:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -2, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
263 ; CHECK: [[V_ADD_U32_e32_1:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -1, [[BUFFER_LOAD_FORMAT_X_IDXEN1]], implicit $exec
264 ; CHECK: [[V_ADD_U32_e32_2:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -3, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
265 ; CHECK: [[V_OR_B32_e32_:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[V_ADD_U32_e32_]], [[V_ADD_U32_e32_1]], implicit $exec
266 ; CHECK: [[V_ADD_U32_e32_3:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -4, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
267 ; 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
268 ; CHECK: [[V_SUBREV_U32_e32_:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 27, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
269 ; 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
270 ; CHECK: [[V_SUBREV_U32_e32_1:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 28, [[BUFFER_LOAD_DWORD_OFFSET]], implicit $exec
271 ; 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
272 ; 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
273 ; CHECK: [[V_OR_B32_e32_5:%[0-9]+]]:vgpr_32 = V_OR_B32_e32 [[S_SUB_I32_]], [[V_OR_B32_e32_4]], implicit $exec
274 ; 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
275 ; 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
276 ; CHECK: [[V_SUBREV_U32_e32_2:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 32, [[BUFFER_LOAD_FORMAT_X_IDXEN2]], implicit $exec
277 ; CHECK: [[V_SUBREV_U32_e32_3:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 33, [[BUFFER_LOAD_FORMAT_X_IDXEN3]], implicit $exec
278 ; 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
279 ; CHECK: [[V_SUBREV_U32_e32_4:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 34, [[BUFFER_LOAD_FORMAT_X_IDXEN4]], implicit $exec
280 ; 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
281 ; CHECK: [[V_SUBREV_U32_e32_5:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 36, [[BUFFER_LOAD_FORMAT_X_IDXEN5]], implicit $exec
282 ; 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
283 ; CHECK: [[V_SUBREV_U32_e32_6:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 37, [[BUFFER_LOAD_FORMAT_X_IDXEN6]], implicit $exec
284 ; 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
285 ; CHECK: [[V_SUBREV_U32_e32_7:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 38, [[BUFFER_LOAD_FORMAT_X_IDXEN7]], implicit $exec
286 ; 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
287 ; CHECK: [[V_SUBREV_U32_e32_8:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 39, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
288 ; 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
289 ; CHECK: [[V_SUBREV_U32_e32_9:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 50, [[BUFFER_LOAD_FORMAT_X_IDXEN8]], implicit $exec
290 ; 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
291 ; CHECK: [[V_SUBREV_U32_e32_10:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 51, [[BUFFER_LOAD_FORMAT_X_IDXEN9]], implicit $exec
292 ; 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
293 ; CHECK: [[V_SUBREV_U32_e32_11:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 52, [[BUFFER_LOAD_FORMAT_X_IDXEN10]], implicit $exec
294 ; 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
295 ; CHECK: [[V_SUBREV_U32_e32_12:%[0-9]+]]:vgpr_32 = V_SUBREV_U32_e32 53, [[BUFFER_LOAD_FORMAT_X_IDXEN11]], implicit $exec
296 ; 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
297 ; CHECK: [[V_ADD_U32_e32_4:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -72, [[BUFFER_LOAD_FORMAT_X_IDXEN12]], implicit $exec
298 ; 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
299 ; CHECK: [[V_ADD_U32_e32_5:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -73, [[BUFFER_LOAD_FORMAT_X_IDXEN13]], implicit $exec
300 ; 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
301 ; CHECK: [[V_ADD_U32_e32_6:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -74, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
302 ; 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
303 ; CHECK: [[V_ADD_U32_e32_7:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -75, [[BUFFER_LOAD_FORMAT_X_IDXEN14]], implicit $exec
304 ; 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
305 ; CHECK: [[V_ADD_U32_e32_8:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -77, [[BUFFER_LOAD_FORMAT_X_IDXEN15]], implicit $exec
306 ; 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
307 ; CHECK: [[V_ADD_U32_e32_9:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -93, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
308 ; 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
309 ; CHECK: [[V_ADD_U32_e32_10:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -94, [[BUFFER_LOAD_FORMAT_X_IDXEN16]], implicit $exec
310 ; 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
311 ; CHECK: [[V_ADD_U32_e32_11:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -95, [[BUFFER_LOAD_DWORD_OFFSET1]], implicit $exec
312 ; 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
313 ; CHECK: [[V_ADD_U32_e32_12:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -96, [[BUFFER_LOAD_DWORD_OFFSET2]], implicit $exec
314 ; 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
315 ; CHECK: [[V_ADD_U32_e32_13:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -97, [[BUFFER_LOAD_DWORD_OFFSET3]], implicit $exec
316 ; 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
317 ; 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
318 ; 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
319 ; 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
320 ; 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
321 ; CHECK: [[V_ADD_U32_e32_14:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -194, [[BUFFER_LOAD_FORMAT_X_IDXEN17]], implicit $exec
322 ; 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
323 ; CHECK: [[V_ADD_U32_e32_15:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -195, [[BUFFER_LOAD_FORMAT_X_IDXEN18]], implicit $exec
324 ; 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
325 ; CHECK: [[V_ADD_U32_e32_16:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -196, [[BUFFER_LOAD_FORMAT_X_IDXEN19]], implicit $exec
326 ; 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
327 ; CHECK: [[V_ADD_U32_e32_17:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -197, [[BUFFER_LOAD_FORMAT_X_IDXEN20]], implicit $exec
328 ; 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
329 ; CHECK: [[V_ADD_U32_e32_18:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -216, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
330 ; 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
331 ; 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
332 ; 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
333 ; 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
334 ; 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
335 ; 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
336 ; 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
337 ; 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
338 ; 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
339 ; CHECK: [[V_ADD_U32_e32_19:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -457, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
340 ; 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
341 ; CHECK: [[V_ADD_U32_e32_20:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -458, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
342 ; 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
343 ; CHECK: [[V_ADD_U32_e32_21:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -459, [[BUFFER_LOAD_FORMAT_X_IDXEN21]], implicit $exec
344 ; 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
345 ; CHECK: [[V_ADD_U32_e32_22:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -466, [[BUFFER_LOAD_FORMAT_X_IDXEN22]], implicit $exec
346 ; 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
347 ; 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
348 ; 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
349 ; 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
350 ; 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
351 ; 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
352 ; 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
353 ; 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
354 ; 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
355 ; 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
356 ; 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
357 ; CHECK: [[V_ADD_U32_e32_23:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -555, [[BUFFER_LOAD_FORMAT_X_IDXEN23]], implicit $exec
358 ; CHECK: [[V_ADD_U32_e32_24:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -556, [[BUFFER_LOAD_FORMAT_X_IDXEN24]], implicit $exec
359 ; 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
360 ; CHECK: [[V_ADD_U32_e32_25:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -557, [[BUFFER_LOAD_FORMAT_X_IDXEN25]], implicit $exec
361 ; 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
362 ; CHECK: [[V_ADD_U32_e32_26:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -574, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
363 ; 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
364 ; CHECK: [[V_ADD_U32_e32_27:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -575, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
365 ; 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
366 ; CHECK: [[S_BUFFER_LOAD_DWORD_IMM8:%[0-9]+]]:sreg_32_xm0_xexec = S_BUFFER_LOAD_DWORD_IMM %71, 0, 0 :: (dereferenceable invariant load (s32))
367 ; CHECK: [[V_ADD_U32_e32_28:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -576, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
368 ; 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
369 ; CHECK: [[V_ADD_U32_e32_29:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -577, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
370 ; 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
371 ; CHECK: [[V_ADD_U32_e32_30:%[0-9]+]]:vgpr_32 = V_ADD_U32_e32 -593, [[BUFFER_LOAD_FORMAT_X_IDXEN]], implicit $exec
372 ; 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
373 ; CHECK: [[S_LOAD_DWORDX8_IMM:%[0-9]+]]:sgpr_256 = S_LOAD_DWORDX8_IMM undef %564:sreg_64, 0, 0 :: (load (s256) from `<8 x i32> addrspace(4)* undef`, addrspace 4)
374 ; 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
375 ; CHECK: [[S_ADD_I32_24:%[0-9]+]]:sreg_32 = S_ADD_I32 [[S_BUFFER_LOAD_DWORD_IMM8]], -594, implicit-def dead $scc
376 ; 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
377 ; 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
378 ; CHECK: undef %691.sub3:vreg_128 = V_CNDMASK_B32_e64 0, 0, 0, 1, [[V_CMP_EQ_U32_e64_]], implicit $exec
379 ; CHECK: IMAGE_STORE_V4_V2_gfx10 %691, undef %578:vreg_64, [[S_LOAD_DWORDX8_IMM]], 15, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable store (s128) into custom "ImageResource")
382 %0 = extractelement <31 x i32> %userData, i64 2
383 %1 = extractelement <31 x i32> %userData, i64 3
384 %2 = extractelement <31 x i32> %userData, i64 4
385 %3 = extractelement <31 x i32> %userData, i64 7
386 %4 = extractelement <31 x i32> %userData, i64 8
387 %5 = extractelement <31 x i32> %userData, i64 9
388 %6 = extractelement <31 x i32> %userData, i64 17
389 %7 = extractelement <31 x i32> %userData, i64 18
390 %8 = extractelement <31 x i32> %userData, i64 19
391 %9 = extractelement <31 x i32> %userData, i64 20
392 %10 = extractelement <31 x i32> %userData, i64 21
393 %11 = extractelement <31 x i32> %userData, i64 22
394 %12 = extractelement <31 x i32> %userData, i64 24
395 %13 = extractelement <31 x i32> %userData, i64 26
396 %14 = extractelement <31 x i32> %userData, i64 30
397 %15 = insertelement <2 x i32> undef, i32 %13, i32 0
398 %16 = bitcast <2 x i32> %15 to i64
399 %17 = inttoptr i64 %16 to i8 addrspace(4)*
400 %18 = insertelement <2 x i32> undef, i32 %12, i32 0
401 %19 = bitcast <2 x i32> %18 to i64
402 %20 = inttoptr i64 %19 to i8 addrspace(4)*
403 %21 = insertelement <2 x i32> undef, i32 %11, i32 0
404 %22 = bitcast <2 x i32> %21 to i64
405 %23 = insertelement <2 x i32> undef, i32 %10, i32 0
406 %24 = bitcast <2 x i32> %23 to i64
407 %25 = insertelement <2 x i32> undef, i32 %9, i32 0
408 %26 = bitcast <2 x i32> %25 to i64
409 %27 = inttoptr i64 %26 to i8 addrspace(4)*
410 %28 = insertelement <2 x i32> undef, i32 %8, i32 0
411 %29 = bitcast <2 x i32> %28 to i64
412 %30 = insertelement <2 x i32> undef, i32 %7, i32 0
413 %31 = bitcast <2 x i32> %30 to i64
414 %32 = inttoptr i64 %31 to i8 addrspace(4)*
415 %33 = insertelement <2 x i32> undef, i32 %6, i32 0
416 %34 = bitcast <2 x i32> %33 to i64
417 %35 = inttoptr i64 %34 to i8 addrspace(4)*
418 %36 = insertelement <2 x i32> undef, i32 %14, i32 0
419 %37 = bitcast <2 x i32> %36 to i64
420 %38 = inttoptr i64 %37 to i8 addrspace(4)*
421 %39 = getelementptr i8, i8 addrspace(4)* %38, i64 232
422 %.i0.i = bitcast i8 addrspace(4)* %39 to i32 addrspace(4)*
423 %rootDesc58.ii0.i = load i32, i32 addrspace(4)* %.i0.i, align 8
424 %.i184.i = getelementptr i8, i8 addrspace(4)* %38, i64 236
425 %40 = bitcast i8 addrspace(4)* %.i184.i to i32 addrspace(4)*
426 %rootDesc58.ii1.i = load i32, i32 addrspace(4)* %40, align 4
427 %41 = and i32 %rootDesc58.ii1.i, 65535
428 %42 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %rootDesc58.ii0.i, i32 0
429 %43 = insertelement <4 x i32> %42, i32 %41, i32 1
430 %44 = and i32 undef, 65535
431 %45 = insertelement <4 x i32> undef, i32 %44, i32 1
432 %46 = load <4 x i32>, <4 x i32> addrspace(4)* undef, align 16
433 %47 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %46, i32 0, i32 0, i32 0, i32 0)
434 %48 = add i32 %47, -1
436 %50 = call i32 @llvm.amdgcn.readfirstlane(i32 %49)
437 %51 = sext i32 %50 to i64
438 %52 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
439 %53 = add i32 %52, -2
440 %54 = or i32 %53, %48
442 %56 = call i32 @llvm.amdgcn.readfirstlane(i32 %55)
443 %57 = sext i32 %56 to i64
444 %58 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
445 %59 = add i32 %58, -3
446 %60 = or i32 %54, %59
448 %62 = call i32 @llvm.amdgcn.readfirstlane(i32 %61)
449 %63 = sext i32 %62 to i64
450 %64 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
451 %65 = add i32 %64, -4
452 %66 = or i32 %60, %65
453 %67 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
454 %68 = add i32 %67, -27
455 %69 = or i32 %66, %68
456 %70 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> undef, i32 0, i32 0, i32 0)
457 %71 = add i32 %70, -28
458 %72 = or i32 %69, %71
459 %73 = call i32 @llvm.amdgcn.readfirstlane(i32 %0)
460 %74 = getelementptr i8, i8 addrspace(4)* %35, i64 16
461 %75 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 0, i32 0)
462 %76 = add i32 %75, -29
463 %77 = or i32 %72, %76
464 %78 = call i32 @llvm.amdgcn.readfirstlane(i32 %1)
466 %80 = sext i32 %79 to i64
467 %81 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 0, i32 0)
468 %82 = add i32 %81, -30
469 %83 = or i32 %77, %82
470 %84 = call i32 @llvm.amdgcn.readfirstlane(i32 %2)
472 %86 = sext i32 %85 to i64
473 %87 = getelementptr i8, i8 addrspace(4)* %74, i64 %86
474 %88 = bitcast i8 addrspace(4)* %87 to <4 x i32> addrspace(4)*
475 %89 = load <4 x i32>, <4 x i32> addrspace(4)* %88, align 16
476 %90 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %89, i32 0, i32 0)
477 %91 = add i32 %90, -31
478 %92 = or i32 %83, %91
479 %93 = getelementptr i8, i8 addrspace(4)* %35, i64 64
480 %94 = getelementptr i8, i8 addrspace(4)* %93, i64 %51
481 %95 = bitcast i8 addrspace(4)* %94 to <4 x i32> addrspace(4)*
482 %96 = load <4 x i32>, <4 x i32> addrspace(4)* %95, align 16
483 %97 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %96, i32 0, i32 0, i32 0, i32 0)
484 %98 = add i32 %97, -32
485 %99 = or i32 %92, %98
486 %100 = getelementptr i8, i8 addrspace(4)* %93, i64 %57
487 %101 = bitcast i8 addrspace(4)* %100 to <4 x i32> addrspace(4)*
488 %102 = load <4 x i32>, <4 x i32> addrspace(4)* %101, align 16
489 %103 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %102, i32 0, i32 0, i32 0, i32 0)
490 %104 = add i32 %103, -33
491 %105 = or i32 %99, %104
492 %106 = getelementptr i8, i8 addrspace(4)* %93, i64 %63
493 %107 = bitcast i8 addrspace(4)* %106 to <4 x i32> addrspace(4)*
494 %108 = load <4 x i32>, <4 x i32> addrspace(4)* %107, align 16
495 %109 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %108, i32 0, i32 0, i32 0, i32 0)
496 %110 = add i32 %109, -34
497 %111 = or i32 %105, %110
498 %112 = call i32 @llvm.amdgcn.readfirstlane(i32 undef)
499 %113 = sext i32 %112 to i64
500 %114 = getelementptr i8, i8 addrspace(4)* %93, i64 %113
501 %115 = bitcast i8 addrspace(4)* %114 to <4 x i32> addrspace(4)*
502 %116 = load <4 x i32>, <4 x i32> addrspace(4)* %115, align 16
503 %117 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %116, i32 0, i32 0, i32 0, i32 0)
504 %118 = add i32 %117, -36
505 %119 = or i32 %111, %118
506 %120 = getelementptr i8, i8 addrspace(4)* %32, i64 %51
507 %121 = bitcast i8 addrspace(4)* %120 to <4 x i32> addrspace(4)*
508 %122 = load <4 x i32>, <4 x i32> addrspace(4)* %121, align 16
509 %123 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %122, i32 0, i32 0, i32 0, i32 0)
510 %124 = add i32 %123, -37
511 %125 = or i32 %119, %124
512 %126 = getelementptr i8, i8 addrspace(4)* %32, i64 %57
513 %127 = bitcast i8 addrspace(4)* %126 to <4 x i32> addrspace(4)*
514 %128 = load <4 x i32>, <4 x i32> addrspace(4)* %127, align 16
515 %129 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %128, i32 0, i32 0, i32 0, i32 0)
516 %130 = add i32 %129, -38
517 %131 = or i32 %125, %130
518 %132 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
519 %133 = add i32 %132, -39
520 %134 = or i32 %131, %133
521 %135 = call i32 @llvm.amdgcn.readfirstlane(i32 undef)
522 %136 = sext i32 %135 to i64
523 %137 = getelementptr i8, i8 addrspace(4)* %32, i64 %136
524 %138 = bitcast i8 addrspace(4)* %137 to <4 x i32> addrspace(4)*
525 %139 = load <4 x i32>, <4 x i32> addrspace(4)* %138, align 16
526 %140 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %139, i32 0, i32 0, i32 0, i32 0)
527 %141 = add i32 %140, -50
528 %142 = or i32 %134, %141
529 %143 = getelementptr i8, i8 addrspace(4)* %32, i64 224
530 %144 = getelementptr i8, i8 addrspace(4)* %143, i64 %51
531 %145 = bitcast i8 addrspace(4)* %144 to <4 x i32> addrspace(4)*
532 %146 = load <4 x i32>, <4 x i32> addrspace(4)* %145, align 16
533 %147 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %146, i32 0, i32 0, i32 0, i32 0)
534 %148 = add i32 %147, -51
535 %149 = or i32 %142, %148
536 %150 = getelementptr i8, i8 addrspace(4)* %143, i64 %57
537 %151 = bitcast i8 addrspace(4)* %150 to <4 x i32> addrspace(4)*
538 %152 = load <4 x i32>, <4 x i32> addrspace(4)* %151, align 16
539 %153 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %152, i32 0, i32 0, i32 0, i32 0)
540 %154 = add i32 %153, -52
541 %155 = or i32 %149, %154
542 %156 = getelementptr i8, i8 addrspace(4)* %143, i64 %63
543 %157 = bitcast i8 addrspace(4)* %156 to <4 x i32> addrspace(4)*
544 %158 = load <4 x i32>, <4 x i32> addrspace(4)* %157, align 16
545 %159 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %158, i32 0, i32 0, i32 0, i32 0)
546 %160 = add i32 %159, -53
547 %161 = or i32 %155, %160
548 %162 = sext i32 undef to i64
549 %163 = getelementptr i8, i8 addrspace(4)* %143, i64 %162
550 %164 = bitcast i8 addrspace(4)* %163 to <4 x i32> addrspace(4)*
551 %165 = load <4 x i32>, <4 x i32> addrspace(4)* %164, align 16
552 %166 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %165, i32 0, i32 0, i32 0, i32 0)
553 %167 = add i32 %166, -72
554 %168 = or i32 %161, %167
555 %169 = getelementptr i8, i8 addrspace(4)* %32, i64 576
556 %170 = getelementptr i8, i8 addrspace(4)* %169, i64 %51
557 %171 = bitcast i8 addrspace(4)* %170 to <4 x i32> addrspace(4)*
558 %172 = load <4 x i32>, <4 x i32> addrspace(4)* %171, align 16
559 %173 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %172, i32 0, i32 0, i32 0, i32 0)
560 %174 = add i32 %173, -73
561 %175 = or i32 %168, %174
562 %176 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
563 %177 = add i32 %176, -74
564 %178 = or i32 %175, %177
565 %179 = getelementptr i8, i8 addrspace(4)* %169, i64 %63
566 %180 = bitcast i8 addrspace(4)* %179 to <4 x i32> addrspace(4)*
567 %181 = load <4 x i32>, <4 x i32> addrspace(4)* %180, align 16
568 %182 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %181, i32 0, i32 0, i32 0, i32 0)
569 %183 = add i32 %182, -75
570 %184 = or i32 %178, %183
571 %185 = getelementptr i8, i8 addrspace(4)* %169, i64 %113
572 %186 = bitcast i8 addrspace(4)* %185 to <4 x i32> addrspace(4)*
573 %187 = load <4 x i32>, <4 x i32> addrspace(4)* %186, align 16
574 %188 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %187, i32 0, i32 0, i32 0, i32 0)
575 %189 = add i32 %188, -77
576 %190 = or i32 %184, %189
577 %191 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
578 %192 = add i32 %191, -93
579 %193 = or i32 %190, %192
580 %194 = inttoptr i64 %29 to i8 addrspace(4)*
581 %195 = getelementptr i8, i8 addrspace(4)* %194, i64 %51
582 %196 = bitcast i8 addrspace(4)* %195 to <4 x i32> addrspace(4)*
583 %197 = load <4 x i32>, <4 x i32> addrspace(4)* %196, align 16
584 %198 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %197, i32 0, i32 0, i32 0, i32 0)
585 %199 = add i32 %198, -94
586 %200 = or i32 %193, %199
587 %201 = load <4 x i32>, <4 x i32> addrspace(4)* undef, align 16
588 %202 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %201, i32 0, i32 0, i32 0)
589 %203 = add i32 %202, -95
590 %204 = or i32 %200, %203
591 %205 = getelementptr i8, i8 addrspace(4)* %27, i64 %80
592 %206 = bitcast i8 addrspace(4)* %205 to <4 x i32> addrspace(4)*
593 %207 = load <4 x i32>, <4 x i32> addrspace(4)* %206, align 16
594 %208 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %207, i32 0, i32 0, i32 0)
595 %209 = add i32 %208, -96
596 %210 = or i32 %204, %209
597 %211 = getelementptr i8, i8 addrspace(4)* %27, i64 %86
598 %212 = bitcast i8 addrspace(4)* %211 to <4 x i32> addrspace(4)*
599 %213 = load <4 x i32>, <4 x i32> addrspace(4)* %212, align 16
600 %214 = call i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32> %213, i32 0, i32 0, i32 0)
601 %215 = add i32 %214, -97
602 %216 = or i32 %210, %215
603 %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
604 %218 = ptrtoint i32 addrspace(6)* %217 to i32
605 %219 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 %218, i32 0)
606 %220 = add i32 %219, -98
607 %221 = or i32 %216, %220
608 %222 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 undef, i32 0)
609 %223 = add i32 %222, -114
610 %224 = or i32 %221, %223
611 %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
612 %226 = ptrtoint i32 addrspace(6)* %225 to i32
613 %227 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 %226, i32 0)
614 %228 = add i32 %227, -130
615 %229 = or i32 %224, %228
616 %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
617 %231 = ptrtoint i32 addrspace(6)* %230 to i32
618 %232 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %45, i32 %231, i32 0)
619 %233 = add i32 %232, -178
620 %234 = or i32 %229, %233
621 %235 = inttoptr i64 %24 to i8 addrspace(4)*
622 %236 = getelementptr i8, i8 addrspace(4)* %235, i64 %51
623 %237 = bitcast i8 addrspace(4)* %236 to <4 x i32> addrspace(4)*
624 %238 = load <4 x i32>, <4 x i32> addrspace(4)* %237, align 16
625 %239 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %238, i32 0, i32 0, i32 0, i32 0)
626 %240 = add i32 %239, -194
627 %241 = or i32 %234, %240
628 %242 = inttoptr i64 %22 to i8 addrspace(4)*
629 %243 = getelementptr i8, i8 addrspace(4)* %242, i64 %51
630 %244 = bitcast i8 addrspace(4)* %243 to <4 x i32> addrspace(4)*
631 %245 = load <4 x i32>, <4 x i32> addrspace(4)* %244, align 16
632 %246 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %245, i32 0, i32 0, i32 0, i32 0)
633 %247 = add i32 %246, -195
634 %248 = or i32 %241, %247
635 %249 = getelementptr i8, i8 addrspace(4)* %242, i64 %57
636 %250 = bitcast i8 addrspace(4)* %249 to <4 x i32> addrspace(4)*
637 %251 = load <4 x i32>, <4 x i32> addrspace(4)* %250, align 16
638 %252 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %251, i32 0, i32 0, i32 0, i32 0)
639 %253 = add i32 %252, -196
640 %254 = or i32 %248, %253
641 %255 = getelementptr i8, i8 addrspace(4)* %242, i64 %63
642 %256 = bitcast i8 addrspace(4)* %255 to <4 x i32> addrspace(4)*
643 %257 = load <4 x i32>, <4 x i32> addrspace(4)* %256, align 16
644 %258 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %257, i32 0, i32 0, i32 0, i32 0)
645 %259 = add i32 %258, -197
646 %260 = or i32 %254, %259
647 %261 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
648 %262 = add i32 %261, -216
649 %263 = or i32 %260, %262
650 %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
651 %265 = ptrtoint i32 addrspace(6)* %264 to i32
652 %266 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %265, i32 0)
653 %267 = add i32 %266, -217
654 %268 = or i32 %263, %267
655 %269 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
656 %270 = add i32 %269, -233
657 %271 = or i32 %268, %270
658 %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
659 %273 = ptrtoint i32 addrspace(6)* %272 to i32
660 %274 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %273, i32 0)
661 %275 = add i32 %274, -249
662 %276 = or i32 %271, %275
663 %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
664 %278 = ptrtoint i32 addrspace(6)* %277 to i32
665 %279 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %278, i32 0)
666 %280 = add i32 %279, -297
667 %281 = or i32 %276, %280
668 %282 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
669 %283 = add i32 %282, -313
670 %284 = or i32 %281, %283
671 %285 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
672 %286 = add i32 %285, -329
673 %287 = or i32 %284, %286
674 %288 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
675 %289 = add i32 %288, -345
676 %290 = or i32 %287, %289
677 %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
678 %292 = ptrtoint i32 addrspace(6)* %291 to i32
679 %293 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 %292, i32 0)
680 %294 = add i32 %293, -441
681 %295 = or i32 %290, %294
682 %296 = getelementptr i8, i8 addrspace(4)* %20, i64 160
683 %297 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
684 %298 = add i32 %297, -457
685 %299 = or i32 %295, %298
686 %300 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
687 %301 = add i32 %300, -458
688 %302 = or i32 %299, %301
689 %303 = getelementptr i8, i8 addrspace(4)* %296, i64 %63
690 %304 = bitcast i8 addrspace(4)* %303 to <4 x i32> addrspace(4)*
691 %305 = load <4 x i32>, <4 x i32> addrspace(4)* %304, align 16
692 %306 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %305, i32 0, i32 0, i32 0, i32 0)
693 %307 = add i32 %306, -459
694 %308 = or i32 %302, %307
696 %310 = call i32 @llvm.amdgcn.readfirstlane(i32 %309)
697 %311 = sext i32 %310 to i64
698 %312 = getelementptr i8, i8 addrspace(4)* %296, i64 %311
699 %313 = bitcast i8 addrspace(4)* %312 to <4 x i32> addrspace(4)*
700 %314 = load <4 x i32>, <4 x i32> addrspace(4)* %313, align 16
701 %315 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %314, i32 0, i32 0, i32 0, i32 0)
702 %316 = add i32 %315, -466
703 %317 = or i32 %308, %316
704 %318 = getelementptr i8, i8 addrspace(4)* %38, i64 168
705 %319 = shl i32 %73, 3
706 %320 = sext i32 %319 to i64
707 %321 = getelementptr i8, i8 addrspace(4)* %318, i64 %320
708 %.i085.i = bitcast i8 addrspace(4)* %321 to i32 addrspace(4)*
709 %.ii0.i = load i32, i32 addrspace(4)* %.i085.i, align 8
710 %322 = and i32 undef, 65535
711 %323 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii0.i, i32 0
712 %324 = insertelement <4 x i32> %323, i32 %322, i32 1
713 %325 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %324, i32 0, i32 0)
714 %326 = add i32 %325, -467
715 %327 = or i32 %317, %326
716 %328 = shl i32 %78, 3
717 %329 = sext i32 %328 to i64
718 %330 = getelementptr i8, i8 addrspace(4)* %318, i64 %329
719 %.i088.i = bitcast i8 addrspace(4)* %330 to i32 addrspace(4)*
720 %.ii090.i = load i32, i32 addrspace(4)* %.i088.i, align 8
721 %.i191.i = getelementptr i8, i8 addrspace(4)* %330, i64 4
722 %331 = bitcast i8 addrspace(4)* %.i191.i to i32 addrspace(4)*
723 %.ii192.i = load i32, i32 addrspace(4)* %331, align 4
724 %332 = and i32 %.ii192.i, 65535
725 %333 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii090.i, i32 0
726 %334 = insertelement <4 x i32> %333, i32 %332, i32 1
727 %335 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %334, i32 0, i32 0)
728 %336 = add i32 %335, -468
729 %337 = or i32 %327, %336
730 %338 = shl i32 %84, 3
731 %339 = sext i32 %338 to i64
732 %340 = getelementptr i8, i8 addrspace(4)* %318, i64 %339
733 %.i094.i = bitcast i8 addrspace(4)* %340 to i32 addrspace(4)*
734 %.ii096.i = load i32, i32 addrspace(4)* %.i094.i, align 8
735 %.i197.i = getelementptr i8, i8 addrspace(4)* %340, i64 4
736 %341 = bitcast i8 addrspace(4)* %.i197.i to i32 addrspace(4)*
737 %.ii198.i = load i32, i32 addrspace(4)* %341, align 4
738 %342 = and i32 %.ii198.i, 65535
739 %343 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii096.i, i32 0
740 %344 = insertelement <4 x i32> %343, i32 %342, i32 1
741 %345 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %344, i32 0, i32 0)
742 %346 = add i32 %345, -469
743 %347 = or i32 %337, %346
744 %348 = call i32 @llvm.amdgcn.readfirstlane(i32 %3)
745 %349 = shl i32 %348, 3
746 %350 = sext i32 %349 to i64
747 %351 = getelementptr i8, i8 addrspace(4)* %318, i64 %350
748 %.i0100.i = bitcast i8 addrspace(4)* %351 to i32 addrspace(4)*
749 %.ii0102.i = load i32, i32 addrspace(4)* %.i0100.i, align 8
750 %.ii1104.i = load i32, i32 addrspace(4)* undef, align 4
751 %352 = and i32 %.ii1104.i, 65535
752 %353 = insertelement <4 x i32> <i32 undef, i32 undef, i32 -1, i32 553734060>, i32 %.ii0102.i, i32 0
753 %354 = insertelement <4 x i32> %353, i32 %352, i32 1
754 %355 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %354, i32 0, i32 0)
755 %356 = add i32 %355, -473
756 %357 = or i32 %347, %356
757 %358 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 0, i32 0)
758 %359 = add i32 %358, -474
759 %360 = or i32 %357, %359
760 %361 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
761 %362 = add i32 %361, -475
762 %363 = or i32 %360, %362
763 %364 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
764 %365 = add i32 %364, -491
765 %366 = or i32 %363, %365
766 %367 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
767 %368 = add i32 %367, -507
768 %369 = or i32 %366, %368
769 %370 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> undef, i32 undef, i32 0)
770 %371 = add i32 %370, -539
771 %372 = or i32 %369, %371
772 %373 = getelementptr i8, i8 addrspace(4)* %17, i64 96
773 %374 = getelementptr i8, i8 addrspace(4)* %373, i64 %51
774 %375 = bitcast i8 addrspace(4)* %374 to <4 x i32> addrspace(4)*
775 %376 = load <4 x i32>, <4 x i32> addrspace(4)* %375, align 16
776 %377 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %376, i32 0, i32 0, i32 0, i32 0)
777 %378 = add i32 %377, -555
778 %379 = or i32 %372, %378
779 %380 = getelementptr i8, i8 addrspace(4)* %373, i64 %57
780 %381 = bitcast i8 addrspace(4)* %380 to <4 x i32> addrspace(4)*
781 %382 = load <4 x i32>, <4 x i32> addrspace(4)* %381, align 16
782 %383 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %382, i32 0, i32 0, i32 0, i32 0)
783 %384 = add i32 %383, -556
784 %385 = or i32 %379, %384
785 %386 = getelementptr i8, i8 addrspace(4)* %373, i64 %63
786 %387 = bitcast i8 addrspace(4)* %386 to <4 x i32> addrspace(4)*
787 %388 = load <4 x i32>, <4 x i32> addrspace(4)* %387, align 16
788 %389 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> %388, i32 0, i32 0, i32 0, i32 0)
789 %390 = add i32 %389, -557
790 %391 = or i32 %385, %390
791 %392 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
792 %393 = add i32 %392, -574
793 %394 = or i32 %391, %393
794 %395 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
795 %396 = add i32 %395, -575
796 %397 = or i32 %394, %396
797 %398 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
798 %399 = add i32 %398, -576
799 %400 = or i32 %397, %399
800 %401 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
801 %402 = add i32 %401, -577
802 %403 = or i32 %400, %402
803 %404 = call i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32> undef, i32 0, i32 0, i32 0, i32 0)
804 %405 = add i32 %404, -593
805 %406 = or i32 %403, %405
806 %407 = call i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32> %43, i32 0, i32 0)
807 %408 = add i32 %407, -594
808 %409 = or i32 %406, %408
809 %.not.i = icmp eq i32 %409, 0
810 %410 = load <8 x i32>, <8 x i32> addrspace(4)* undef, align 32
811 %.i010.i = select i1 %.not.i, float 0x36A0000000000000, float 0.000000e+00
812 %411 = insertelement <4 x float> undef, float %.i010.i, i32 3
813 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)
817 declare i32 @llvm.amdgcn.readfirstlane(i32)
818 declare void @llvm.amdgcn.image.store.2d.v4f32.i32(<4 x float>, i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg)
819 declare void @llvm.amdgcn.exp.f32(i32 immarg, i32 immarg, float, float, float, float, i1 immarg, i1 immarg)
820 declare i32 @llvm.amdgcn.struct.buffer.load.format.i32(<4 x i32>, i32, i32, i32, i32 immarg)
821 declare i32 @llvm.amdgcn.raw.buffer.load.i32(<4 x i32>, i32, i32, i32 immarg)
822 declare i32 @llvm.amdgcn.s.buffer.load.i32(<4 x i32>, i32, i32 immarg)
823 declare i32 @llvm.amdgcn.mbcnt.lo(i32, i32)