1 ; RUN: llc < %s -mtriple=thumbv7-elf -mattr=+neon -arm-atomic-cfg-tidy=0 | FileCheck %s
4 %bar = type { float, float, float }
5 %baz = type { i32, [16 x %bar], [16 x float], [16 x i32], i8 }
6 %foo = type { <4 x float> }
7 %quux = type { ptr, ptr, i32 }
8 %quuz = type { %quux, i32, %bar, [128 x i8], [16 x %foo], %foo, %foo, %foo }
10 declare <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr, i32) nounwind readonly
12 define void @aaa(ptr %this, ptr %block) {
14 ; CHECK: bfc r4, #0, #4
15 ; CHECK: vst1.64 {{.*}}[{{.*}}:128]
16 ; CHECK: vld1.64 {{.*}}[{{.*}}:128]
18 %aligned_vec = alloca <4 x float>, align 16
19 %"alloca point" = bitcast i32 0 to i32
20 %0 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr %aligned_vec, i32 1) nounwind
21 store float 6.300000e+01, ptr undef, align 4
22 %1 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind ; <<4 x float>> [#uses=1]
23 store float 0.000000e+00, ptr undef, align 4
24 %2 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind ; <<4 x float>> [#uses=1]
25 %ld3 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
26 store float 0.000000e+00, ptr undef, align 4
27 %ld4 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
28 store float 0.000000e+00, ptr undef, align 4
29 %ld5 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
30 store float 0.000000e+00, ptr undef, align 4
31 %ld6 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
32 store float 0.000000e+00, ptr undef, align 4
33 %ld7 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
34 store float 0.000000e+00, ptr undef, align 4
35 %ld8 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
36 store float 0.000000e+00, ptr undef, align 4
37 %ld9 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
38 store float 0.000000e+00, ptr undef, align 4
39 %ld10 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
40 store float 0.000000e+00, ptr undef, align 4
41 %ld11 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
42 store float 0.000000e+00, ptr undef, align 4
43 %ld12 = call <4 x float> @llvm.arm.neon.vld1.v4f32.p0(ptr undef, i32 1) nounwind
44 store float 0.000000e+00, ptr undef, align 4
45 %val173 = load <4 x float>, ptr undef ; <<4 x float>> [#uses=1]
48 bb4: ; preds = %bb193, %entry
49 %besterror.0.2264 = phi <4 x float> [ undef, %entry ], [ %besterror.0.0, %bb193 ] ; <<4 x float>> [#uses=2]
50 %part0.0.0261 = phi <4 x float> [ zeroinitializer, %entry ], [ %23, %bb193 ] ; <<4 x float>> [#uses=2]
51 %3 = fmul <4 x float> zeroinitializer, %0 ; <<4 x float>> [#uses=2]
52 %4 = fadd <4 x float> %3, %part0.0.0261 ; <<4 x float>> [#uses=1]
53 %5 = shufflevector <4 x float> %3, <4 x float> undef, <2 x i32> <i32 2, i32 3> ; <<2 x float>> [#uses=1]
54 %6 = shufflevector <2 x float> %5, <2 x float> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1> ; <<4 x float>> [#uses=1]
55 %7 = fmul <4 x float> %1, %1
56 %8 = fadd <4 x float> %7, <float 5.000000e-01, float 5.000000e-01, float 5.000000e-01, float 5.000000e-01> ; <<4 x float>> [#uses=1]
57 %9 = fptosi <4 x float> %8 to <4 x i32> ; <<4 x i32>> [#uses=1]
58 %10 = sitofp <4 x i32> %9 to <4 x float> ; <<4 x float>> [#uses=1]
59 %11 = fmul <4 x float> %10, %2 ; <<4 x float>> [#uses=1]
60 %12 = fmul <4 x float> %6, %6
61 %13 = fmul <4 x float> %11, %4 ; <<4 x float>> [#uses=1]
62 %14 = fsub <4 x float> %12, %13 ; <<4 x float>> [#uses=1]
63 %15 = fsub <4 x float> %14, %14
64 %16 = fmul <4 x float> %15, <float 2.000000e+00, float 2.000000e+00, float 2.000000e+00, float 2.000000e+00> ; <<4 x float>> [#uses=1]
65 %17 = fadd <4 x float> %16, %16
66 %18 = fmul <4 x float> %17, %val173 ; <<4 x float>> [#uses=1]
67 %19 = shufflevector <4 x float> %18, <4 x float> undef, <2 x i32> <i32 2, i32 3> ; <<2 x float>> [#uses=1]
68 %20 = shufflevector <2 x float> %19, <2 x float> undef, <4 x i32> zeroinitializer ; <<4 x float>> [#uses=1]
69 %tmp1 = fadd <4 x float> %20, %ld3
70 %tmp2 = fadd <4 x float> %tmp1, %ld4
71 %tmp3 = fadd <4 x float> %tmp2, %ld5
72 %tmp4 = fadd <4 x float> %tmp3, %ld6
73 %tmp5 = fadd <4 x float> %tmp4, %ld7
74 %tmp6 = fadd <4 x float> %tmp5, %ld8
75 %tmp7 = fadd <4 x float> %tmp6, %ld9
76 %tmp8 = fadd <4 x float> %tmp7, %ld10
77 %tmp9 = fadd <4 x float> %tmp8, %ld11
78 %21 = fadd <4 x float> %tmp9, %ld12
79 %22 = fcmp ogt <4 x float> %besterror.0.2264, %21 ; <<4 x i1>> [#uses=0]
80 %tmp = extractelement <4 x i1> %22, i32 0
81 br i1 %tmp, label %bb193, label %bb186
86 bb193: ; preds = %bb186, %bb4
87 %besterror.0.0 = phi <4 x float> [ %21, %bb186 ], [ %besterror.0.2264, %bb4 ] ; <<4 x float>> [#uses=1]
88 %23 = fadd <4 x float> %part0.0.0261, zeroinitializer ; <<4 x float>> [#uses=1]