; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -mtriple=x86_64-unknown-linux -mcpu=skylake-avx512 -passes=slp-vectorizer -S | FileCheck %s
; RUN: opt < %s -mtriple=x86_64-unknown-linux -mcpu=skylake-avx512 -passes=inject-tli-mappings,slp-vectorizer -vector-library=SVML -S | FileCheck %s --check-prefix=VECLIB
; RUN: opt < %s -mtriple=x86_64-unknown-linux -mcpu=skylake-avx512 -passes=inject-tli-mappings,slp-vectorizer -vector-library=AMDLIBM -S | FileCheck %s --check-prefix=AMDLIBM

@src = common global [8 x double] zeroinitializer, align 64
@dst = common global [8 x double] zeroinitializer, align 64

declare double @llvm.sqrt.f64(double)
declare double @llvm.sin.f64(double)

define void @test() {
; CHECK-LABEL: @test(
; CHECK-NEXT:    [[A2:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 2), align 8
; CHECK-NEXT:    [[A3:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 3), align 8
; CHECK-NEXT:    [[A6:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 6), align 8
; CHECK-NEXT:    [[A7:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 7), align 8
; CHECK-NEXT:    [[SIN0:%.*]] = call fast double @llvm.sin.f64(double [[A2]])
; CHECK-NEXT:    [[SIN1:%.*]] = call fast double @llvm.sin.f64(double [[A3]])
; CHECK-NEXT:    [[SIN2:%.*]] = call fast double @llvm.sin.f64(double [[A6]])
; CHECK-NEXT:    [[SIN3:%.*]] = call fast double @llvm.sin.f64(double [[A7]])
; CHECK-NEXT:    [[TMP1:%.*]] = load <6 x double>, ptr @src, align 8
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <6 x double> [[TMP1]], <6 x double> poison, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
; CHECK-NEXT:    [[TMP6:%.*]] = shufflevector <6 x double> [[TMP1]], <6 x double> poison, <2 x i32> <i32 0, i32 4>
; CHECK-NEXT:    [[TMP7:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP6]])
; CHECK-NEXT:    [[TMP8:%.*]] = shufflevector <6 x double> [[TMP1]], <6 x double> poison, <2 x i32> <i32 1, i32 5>
; CHECK-NEXT:    [[TMP9:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP8]])
; CHECK-NEXT:    [[TMP13:%.*]] = insertelement <2 x double> poison, double [[SIN1]], i32 0
; CHECK-NEXT:    [[TMP5:%.*]] = insertelement <2 x double> [[TMP13]], double [[SIN3]], i32 1
; CHECK-NEXT:    [[TMP10:%.*]] = fadd fast <2 x double> [[TMP7]], [[TMP5]]
; CHECK-NEXT:    [[TMP14:%.*]] = insertelement <2 x double> poison, double [[SIN0]], i32 0
; CHECK-NEXT:    [[TMP3:%.*]] = insertelement <2 x double> [[TMP14]], double [[SIN2]], i32 1
; CHECK-NEXT:    [[TMP11:%.*]] = fadd fast <2 x double> [[TMP3]], [[TMP9]]
; CHECK-NEXT:    [[TMP12:%.*]] = fadd fast <2 x double> [[TMP10]], [[TMP11]]
; CHECK-NEXT:    store <2 x double> [[TMP12]], ptr @dst, align 8
; CHECK-NEXT:    ret void
;
; VECLIB-LABEL: @test(
; VECLIB-NEXT:    [[TMP1:%.*]] = load <8 x double>, ptr @src, align 8
; VECLIB-NEXT:    [[TMP2:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 2, i32 6>
; VECLIB-NEXT:    [[TMP3:%.*]] = call fast <2 x double> @__svml_sin2(<2 x double> [[TMP2]])
; VECLIB-NEXT:    [[TMP4:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 3, i32 7>
; VECLIB-NEXT:    [[TMP5:%.*]] = call fast <2 x double> @__svml_sin2(<2 x double> [[TMP4]])
; VECLIB-NEXT:    [[TMP6:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 0, i32 4>
; VECLIB-NEXT:    [[TMP7:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP6]])
; VECLIB-NEXT:    [[TMP8:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 1, i32 5>
; VECLIB-NEXT:    [[TMP9:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP8]])
; VECLIB-NEXT:    [[TMP10:%.*]] = fadd fast <2 x double> [[TMP7]], [[TMP5]]
; VECLIB-NEXT:    [[TMP11:%.*]] = fadd fast <2 x double> [[TMP3]], [[TMP9]]
; VECLIB-NEXT:    [[TMP12:%.*]] = fadd fast <2 x double> [[TMP10]], [[TMP11]]
; VECLIB-NEXT:    store <2 x double> [[TMP12]], ptr @dst, align 8
; VECLIB-NEXT:    ret void
;
; AMDLIBM-LABEL: @test(
; AMDLIBM-NEXT:    [[TMP1:%.*]] = load <8 x double>, ptr @src, align 8
; AMDLIBM-NEXT:    [[TMP2:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 2, i32 6>
; AMDLIBM-NEXT:    [[TMP3:%.*]] = call fast <2 x double> @amd_vrd2_sin(<2 x double> [[TMP2]])
; AMDLIBM-NEXT:    [[TMP4:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 3, i32 7>
; AMDLIBM-NEXT:    [[TMP5:%.*]] = call fast <2 x double> @amd_vrd2_sin(<2 x double> [[TMP4]])
; AMDLIBM-NEXT:    [[TMP6:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 0, i32 4>
; AMDLIBM-NEXT:    [[TMP7:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP6]])
; AMDLIBM-NEXT:    [[TMP8:%.*]] = shufflevector <8 x double> [[TMP1]], <8 x double> poison, <2 x i32> <i32 1, i32 5>
; AMDLIBM-NEXT:    [[TMP9:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP8]])
; AMDLIBM-NEXT:    [[TMP10:%.*]] = fadd fast <2 x double> [[TMP7]], [[TMP5]]
; AMDLIBM-NEXT:    [[TMP11:%.*]] = fadd fast <2 x double> [[TMP3]], [[TMP9]]
; AMDLIBM-NEXT:    [[TMP12:%.*]] = fadd fast <2 x double> [[TMP10]], [[TMP11]]
; AMDLIBM-NEXT:    store <2 x double> [[TMP12]], ptr @dst, align 8
; AMDLIBM-NEXT:    ret void
;
  %a0 = load double, ptr @src, align 8
  %a1 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 1), align 8
  %a2 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 2), align 8
  %a3 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 3), align 8
  %a4 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 4), align 8
  %a5 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 5), align 8
  %a6 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 6), align 8
  %a7 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 7), align 8
  %sin0 = call fast double @llvm.sin.f64(double %a2)
  %sin1 = call fast double @llvm.sin.f64(double %a3)
  %sqrt0 = call fast double @llvm.sqrt.f64(double %a0)
  %sqrt1 = call fast double @llvm.sqrt.f64(double %a1)
  %sin2 = call fast double @llvm.sin.f64(double %a6)
  %sin3 = call fast double @llvm.sin.f64(double %a7)
  %sqrt2 = call fast double @llvm.sqrt.f64(double %a4)
  %sqrt3 = call fast double @llvm.sqrt.f64(double %a5)
  %res1 = fadd fast double %sqrt0, %sin1
  %res2 = fadd fast double %sin0, %sqrt1
  %res00 = fadd fast double %res1, %res2
  %res3 = fadd fast double %sqrt2, %sin3
  %res4 = fadd fast double %sin2, %sqrt3
  %res01 = fadd fast double %res3, %res4
  store double %res00, ptr @dst, align 8
  store double %res01, ptr getelementptr inbounds ([8 x double], ptr @dst, i32 0, i64 1), align 8
  ret void
}