1 ; RUN: llc -mtriple=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN %s
2 ; RUN: llc -mtriple=amdgcn -mcpu=gfx90a -verify-machineinstrs -early-live-intervals < %s | FileCheck -check-prefixes=GCN %s
3 ; RUN: llc -global-isel -mtriple=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN %s
5 ; GCN-LABEL: {{^}}load_1d:
6 ; GCN: image_load v[0:3], v0, s[0:7] dmask:0xf unorm{{$}}
7 define amdgpu_ps <4 x float> @load_1d(<8 x i32> inreg %rsrc, i32 %s) {
9 %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
13 ; GCN-LABEL: {{^}}load_1d_lwe:
14 ; GCN: image_load v[0:4], v{{[0-9]+}}, s[0:7] dmask:0xf unorm lwe{{$}}
15 define amdgpu_ps <4 x float> @load_1d_lwe(<8 x i32> inreg %rsrc, ptr addrspace(1) inreg %out, i32 %s) {
17 %v = call {<4 x float>, i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 2, i32 0)
18 %v.vec = extractvalue {<4 x float>, i32} %v, 0
19 %v.err = extractvalue {<4 x float>, i32} %v, 1
20 store i32 %v.err, ptr addrspace(1) %out, align 4
21 ret <4 x float> %v.vec
24 ; GCN-LABEL: {{^}}load_2d:
25 ; GCN: image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm{{$}}
26 define amdgpu_ps <4 x float> @load_2d(<8 x i32> inreg %rsrc, i32 %s, i32 %t) {
28 %v = call <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
32 ; GCN-LABEL: {{^}}load_3d:
33 ; GCN: image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm{{$}}
34 define amdgpu_ps <4 x float> @load_3d(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %r) {
36 %v = call <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 0, i32 0)
40 ; GCN-LABEL: {{^}}load_cube:
41 ; GCN: image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm da{{$}}
42 define amdgpu_ps <4 x float> @load_cube(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice) {
44 %v = call <4 x float> @llvm.amdgcn.image.load.cube.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
48 ; GCN-LABEL: {{^}}load_cube_lwe:
49 ; GCN: image_load v[0:4], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0xf unorm lwe da{{$}}
50 define amdgpu_ps <4 x float> @load_cube_lwe(<8 x i32> inreg %rsrc, ptr addrspace(1) inreg %out, i32 %s, i32 %t, i32 %slice) {
52 %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.cube.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 2, i32 0)
53 %v.vec = extractvalue {<4 x float>, i32} %v, 0
54 %v.err = extractvalue {<4 x float>, i32} %v, 1
55 store i32 %v.err, ptr addrspace(1) %out, align 4
56 ret <4 x float> %v.vec
59 ; GCN-LABEL: {{^}}load_1darray:
60 ; GCN: image_load v[0:3], v[0:1], s[0:7] dmask:0xf unorm da{{$}}
61 define amdgpu_ps <4 x float> @load_1darray(<8 x i32> inreg %rsrc, i32 %s, i32 %slice) {
63 %v = call <4 x float> @llvm.amdgcn.image.load.1darray.v4f32.i32(i32 15, i32 %s, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
67 ; GCN-LABEL: {{^}}load_2darray:
68 ; GCN: image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm da{{$}}
69 define amdgpu_ps <4 x float> @load_2darray(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice) {
71 %v = call <4 x float> @llvm.amdgcn.image.load.2darray.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
75 ; GCN-LABEL: {{^}}load_2darray_lwe:
76 ; GCN: image_load v[0:4], v[{{[0-9]+:[0-9]+}}], s[0:7] dmask:0xf unorm lwe da{{$}}
77 define amdgpu_ps <4 x float> @load_2darray_lwe(<8 x i32> inreg %rsrc, ptr addrspace(1) inreg %out, i32 %s, i32 %t, i32 %slice) {
79 %v = call {<4 x float>,i32} @llvm.amdgcn.image.load.2darray.v4f32i32.i32(i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 2, i32 0)
80 %v.vec = extractvalue {<4 x float>, i32} %v, 0
81 %v.err = extractvalue {<4 x float>, i32} %v, 1
82 store i32 %v.err, ptr addrspace(1) %out, align 4
83 ret <4 x float> %v.vec
86 ; GCN-LABEL: {{^}}load_2dmsaa:
87 ; GCN: image_load v[0:3], v[0:2], s[0:7] dmask:0xf unorm{{$}}
88 define amdgpu_ps <4 x float> @load_2dmsaa(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %fragid) {
90 %v = call <4 x float> @llvm.amdgcn.image.load.2dmsaa.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
94 ; GCN-LABEL: {{^}}load_2darraymsaa:
95 ; GCN: image_load v[0:3], v[0:3], s[0:7] dmask:0xf unorm da{{$}}
96 define amdgpu_ps <4 x float> @load_2darraymsaa(<8 x i32> inreg %rsrc, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
98 %v = call <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i32(i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
102 ; GCN-LABEL: {{^}}load_1d_addr_align:
103 ; GCN: v_mov_b32_e32 [[VADDR:v[0-9]?[02468]]], v1
104 ; GCN: image_load v[0:3], [[VADDR]], s[0:7] dmask:0xf unorm{{$}}
105 define amdgpu_ps <4 x float> @load_1d_addr_align(<8 x i32> inreg %rsrc, <2 x i32> %s) {
107 %s1 = extractelement <2 x i32> %s, i32 1
108 %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s1, <8 x i32> %rsrc, i32 0, i32 0)
112 ; GCN-LABEL: {{^}}store_1d:
113 ; GCN: image_store v[0:3], v4, s[0:7] dmask:0xf unorm{{$}}
114 define amdgpu_ps void @store_1d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
116 call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
120 ; GCN-LABEL: {{^}}store_2d:
121 ; GCN: image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm{{$}}
122 define amdgpu_ps void @store_2d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t) {
124 call void @llvm.amdgcn.image.store.2d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
128 ; GCN-LABEL: {{^}}store_3d:
129 ; GCN: image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm{{$}}
130 define amdgpu_ps void @store_3d(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %r) {
132 call void @llvm.amdgcn.image.store.3d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %r, <8 x i32> %rsrc, i32 0, i32 0)
136 ; GCN-LABEL: {{^}}store_cube:
137 ; GCN: image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm da{{$}}
138 define amdgpu_ps void @store_cube(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice) {
140 call void @llvm.amdgcn.image.store.cube.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
144 ; GCN-LABEL: {{^}}store_1darray:
145 ; GCN: image_store v[0:3], v[4:5], s[0:7] dmask:0xf unorm da{{$}}
146 define amdgpu_ps void @store_1darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %slice) {
148 call void @llvm.amdgcn.image.store.1darray.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
152 ; GCN-LABEL: {{^}}store_2darray:
153 ; GCN: image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm da{{$}}
154 define amdgpu_ps void @store_2darray(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice) {
156 call void @llvm.amdgcn.image.store.2darray.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, <8 x i32> %rsrc, i32 0, i32 0)
160 ; GCN-LABEL: {{^}}store_2dmsaa:
161 ; GCN: image_store v[0:3], v[4:6], s[0:7] dmask:0xf unorm{{$}}
162 define amdgpu_ps void @store_2dmsaa(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %fragid) {
164 call void @llvm.amdgcn.image.store.2dmsaa.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
168 ; GCN-LABEL: {{^}}store_2darraymsaa:
169 ; GCN: image_store v[0:3], v[4:7], s[0:7] dmask:0xf unorm da{{$}}
170 define amdgpu_ps void @store_2darraymsaa(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s, i32 %t, i32 %slice, i32 %fragid) {
172 call void @llvm.amdgcn.image.store.2darraymsaa.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, i32 %t, i32 %slice, i32 %fragid, <8 x i32> %rsrc, i32 0, i32 0)
176 ; GCN-LABEL: {{^}}load_1d_V1:
177 ; GCN: image_load v0, v0, s[0:7] dmask:0x8 unorm{{$}}
178 define amdgpu_ps float @load_1d_V1(<8 x i32> inreg %rsrc, i32 %s) {
180 %v = call float @llvm.amdgcn.image.load.1d.f32.i32(i32 8, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
184 ; GCN-LABEL: {{^}}load_1d_V2:
185 ; GCN: image_load v[0:1], v0, s[0:7] dmask:0x9 unorm{{$}}
186 define amdgpu_ps <2 x float> @load_1d_V2(<8 x i32> inreg %rsrc, i32 %s) {
188 %v = call <2 x float> @llvm.amdgcn.image.load.1d.v2f32.i32(i32 9, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
192 ; GCN-LABEL: {{^}}store_1d_V1:
193 ; GCN: v_mov_b32_e32 [[VADDR:v[0-9]?[02468]]], v1
194 ; GCN: image_store v0, [[VADDR]], s[0:7] dmask:0x2 unorm{{$}}
195 define amdgpu_ps void @store_1d_V1(<8 x i32> inreg %rsrc, float %vdata, i32 %s) {
197 call void @llvm.amdgcn.image.store.1d.f32.i32(float %vdata, i32 2, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
201 ; GCN-LABEL: {{^}}store_1d_V2:
202 ; GCN: image_store v[0:1], v2, s[0:7] dmask:0xc unorm{{$}}
203 define amdgpu_ps void @store_1d_V2(<8 x i32> inreg %rsrc, <2 x float> %vdata, i32 %s) {
205 call void @llvm.amdgcn.image.store.1d.v2f32.i32(<2 x float> %vdata, i32 12, i32 %s, <8 x i32> %rsrc, i32 0, i32 0)
209 ; GCN-LABEL: {{^}}load_1d_glc:
210 ; GCN: image_load v[0:3], v0, s[0:7] dmask:0xf unorm glc{{$}}
211 define amdgpu_ps <4 x float> @load_1d_glc(<8 x i32> inreg %rsrc, i32 %s) {
213 %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 1)
217 ; GCN-LABEL: {{^}}load_1d_slc:
218 ; GCN: image_load v[0:3], v0, s[0:7] dmask:0xf unorm slc{{$}}
219 define amdgpu_ps <4 x float> @load_1d_slc(<8 x i32> inreg %rsrc, i32 %s) {
221 %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 2)
225 ; GCN-LABEL: {{^}}load_1d_glc_slc:
226 ; GCN: image_load v[0:3], v0, s[0:7] dmask:0xf unorm glc slc{{$}}
227 define amdgpu_ps <4 x float> @load_1d_glc_slc(<8 x i32> inreg %rsrc, i32 %s) {
229 %v = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 3)
233 ; GCN-LABEL: {{^}}store_1d_glc:
234 ; GCN: image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc{{$}}
235 define amdgpu_ps void @store_1d_glc(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
237 call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 1)
241 ; GCN-LABEL: {{^}}store_1d_slc:
242 ; GCN: image_store v[0:3], v4, s[0:7] dmask:0xf unorm slc{{$}}
243 define amdgpu_ps void @store_1d_slc(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
245 call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 2)
249 ; GCN-LABEL: {{^}}store_1d_glc_slc:
250 ; GCN: image_store v[0:3], v4, s[0:7] dmask:0xf unorm glc slc{{$}}
251 define amdgpu_ps void @store_1d_glc_slc(<8 x i32> inreg %rsrc, <4 x float> %vdata, i32 %s) {
253 call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %vdata, i32 15, i32 %s, <8 x i32> %rsrc, i32 0, i32 3)
257 ; GCN-LABEL: {{^}}image_store_wait:
258 ; GCN: image_store v[0:3], v4, s[0:7] dmask:0xf
259 ; SI: s_waitcnt expcnt(0)
260 ; GCN: image_load v[0:3], v4, s[8:15] dmask:0xf
261 ; GCN: s_waitcnt vmcnt(0)
262 ; GCN: image_store v[0:3], v4, s[16:23] dmask:0xf
263 define amdgpu_ps void @image_store_wait(<8 x i32> inreg %arg, <8 x i32> inreg %arg1, <8 x i32> inreg %arg2, <4 x float> %arg3, i32 %arg4) #0 {
265 call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %arg3, i32 15, i32 %arg4, <8 x i32> %arg, i32 0, i32 0)
266 %data = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %arg4, <8 x i32> %arg1, i32 0, i32 0)
267 call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %data, i32 15, i32 %arg4, <8 x i32> %arg2, i32 0, i32 0)
271 ; GCN-LABEL: image_load_mmo
272 ; GCN: image_load v1, v[{{[0-9:]+}}], s[0:7] dmask:0x1 unorm
273 define amdgpu_ps float @image_load_mmo(<8 x i32> inreg %rsrc, ptr addrspace(3) %lds, <2 x i32> %c) #0 {
274 store float 0.000000e+00, ptr addrspace(3) %lds
275 %c0 = extractelement <2 x i32> %c, i32 0
276 %c1 = extractelement <2 x i32> %c, i32 1
277 %tex = call float @llvm.amdgcn.image.load.2d.f32.i32(i32 1, i32 %c0, i32 %c1, <8 x i32> %rsrc, i32 0, i32 0)
278 %tmp2 = getelementptr float, ptr addrspace(3) %lds, i32 4
279 store float 0.000000e+00, ptr addrspace(3) %tmp2
283 ; GCN: v_mov_b32_e32 [[VADDR:v[0-9]?[02468]]], v1
284 ; GCN: image_get_resinfo v[0:3], [[VADDR]], s[0:7] dmask:0xf unorm
285 define amdgpu_ps <4 x float> @getresinfo_1d(<8 x i32> inreg %rsrc, <2 x i32> %s) {
287 %s1 = extractelement <2 x i32> %s, i32 1
288 %v = call <4 x float> @llvm.amdgcn.image.getresinfo.1d.v4f32.i32(i32 15, i32 %s1, <8 x i32> %rsrc, i32 0, i32 0)
292 declare <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32, i32, <8 x i32>, i32, i32) #1
293 declare {float,i32} @llvm.amdgcn.image.load.1d.f32i32.i32(i32, i32, <8 x i32>, i32, i32) #1
294 declare {<2 x float>,i32} @llvm.amdgcn.image.load.1d.v2f32i32.i32(i32, i32, <8 x i32>, i32, i32) #1
295 declare {<4 x float>,i32} @llvm.amdgcn.image.load.1d.v4f32i32.i32(i32, i32, <8 x i32>, i32, i32) #1
296 declare <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32, i32, i32, <8 x i32>, i32, i32) #1
297 declare {<4 x float>,i32} @llvm.amdgcn.image.load.2d.v4f32i32.i32(i32, i32, i32, <8 x i32>, i32, i32) #1
298 declare <4 x float> @llvm.amdgcn.image.load.3d.v4f32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
299 declare {<4 x float>,i32} @llvm.amdgcn.image.load.3d.v4f32i32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
300 declare <4 x float> @llvm.amdgcn.image.load.cube.v4f32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
301 declare {<4 x float>,i32} @llvm.amdgcn.image.load.cube.v4f32i32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
302 declare <4 x float> @llvm.amdgcn.image.load.1darray.v4f32.i32(i32, i32, i32, <8 x i32>, i32, i32) #1
303 declare {<4 x float>,i32} @llvm.amdgcn.image.load.1darray.v4f32i32.i32(i32, i32, i32, <8 x i32>, i32, i32) #1
304 declare <4 x float> @llvm.amdgcn.image.load.2darray.v4f32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
305 declare {<4 x float>,i32} @llvm.amdgcn.image.load.2darray.v4f32i32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
306 declare <4 x float> @llvm.amdgcn.image.load.2dmsaa.v4f32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
307 declare {<4 x float>,i32} @llvm.amdgcn.image.load.2dmsaa.v4f32i32.i32(i32, i32, i32, i32, <8 x i32>, i32, i32) #1
308 declare <4 x float> @llvm.amdgcn.image.load.2darraymsaa.v4f32.i32(i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #1
309 declare {<4 x float>,i32} @llvm.amdgcn.image.load.2darraymsaa.v4f32i32.i32(i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #1
311 declare void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float>, i32, i32, <8 x i32>, i32, i32) #0
312 declare void @llvm.amdgcn.image.store.2d.v4f32.i32(<4 x float>, i32, i32, i32, <8 x i32>, i32, i32) #0
313 declare void @llvm.amdgcn.image.store.3d.v4f32.i32(<4 x float>, i32, i32, i32, i32, <8 x i32>, i32, i32) #0
314 declare void @llvm.amdgcn.image.store.cube.v4f32.i32(<4 x float>, i32, i32, i32, i32, <8 x i32>, i32, i32) #0
315 declare void @llvm.amdgcn.image.store.1darray.v4f32.i32(<4 x float>, i32, i32, i32, <8 x i32>, i32, i32) #0
316 declare void @llvm.amdgcn.image.store.2darray.v4f32.i32(<4 x float>, i32, i32, i32, i32, <8 x i32>, i32, i32) #0
317 declare void @llvm.amdgcn.image.store.2dmsaa.v4f32.i32(<4 x float>, i32, i32, i32, i32, <8 x i32>, i32, i32) #0
318 declare void @llvm.amdgcn.image.store.2darraymsaa.v4f32.i32(<4 x float>, i32, i32, i32, i32, i32, <8 x i32>, i32, i32) #0
320 declare float @llvm.amdgcn.image.load.1d.f32.i32(i32, i32, <8 x i32>, i32, i32) #1
321 declare float @llvm.amdgcn.image.load.2d.f32.i32(i32, i32, i32, <8 x i32>, i32, i32) #1
322 declare <2 x float> @llvm.amdgcn.image.load.1d.v2f32.i32(i32, i32, <8 x i32>, i32, i32) #1
323 declare void @llvm.amdgcn.image.store.1d.f32.i32(float, i32, i32, <8 x i32>, i32, i32) #0
324 declare void @llvm.amdgcn.image.store.1d.v2f32.i32(<2 x float>, i32, i32, <8 x i32>, i32, i32) #0
326 declare <4 x float> @llvm.amdgcn.image.getresinfo.1d.v4f32.i32(i32, i32, <8 x i32>, i32, i32) #2
328 attributes #0 = { nounwind }
329 attributes #1 = { nounwind readonly }
330 attributes #2 = { nounwind readnone }