llvm/test/Transforms/LoopVectorize/AArch64/veclib-intrinsic-calls.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --filter "call.*(cos|exp|log|sin|pow|ceil|copysign|fabs|floor|fma|m..num|nearbyint|rint|round|sqrt|trunc)" --version 2

; RUN: opt -mattr=+neon -vector-library=sleefgnuabi -passes=inject-tli-mappings,loop-vectorize,simplifycfg -force-vector-interleave=1 -S < %s | FileCheck %s --check-prefix=SLEEF-NEON
; RUN: opt -mattr=+sve -vector-library=sleefgnuabi -passes=inject-tli-mappings,loop-vectorize,simplifycfg -force-vector-interleave=1 -prefer-predicate-over-epilogue=predicate-dont-vectorize -S < %s | FileCheck %s --check-prefix=SLEEF-SVE
; RUN: opt -mattr=+neon -vector-library=ArmPL -passes=inject-tli-mappings,loop-vectorize,simplifycfg -force-vector-interleave=1 -S < %s | FileCheck %s --check-prefix=ARMPL-NEON
; RUN: opt -mattr=+sve -vector-library=ArmPL -passes=inject-tli-mappings,loop-vectorize,simplifycfg -force-vector-interleave=1 -prefer-predicate-over-epilogue=predicate-dont-vectorize -S < %s | FileCheck %s --check-prefix=ARMPL-SVE

target triple = "aarch64-unknown-linux-gnu"

; We are checking whether loops containing intrinsic calls can be vectorized,
; when the compiler provides TLI mappings to their vector variants. The tests
; are checking fixed width vectorization with NEON and scalable vectorization
; with SVE.

declare double @llvm.ceil.f64(double)
declare float @llvm.ceil.f32(float)

define void @ceil_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @ceil_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1:[0-9]+]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.ceil.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @ceil_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1:[0-9]+]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.ceil.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @ceil_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1:[0-9]+]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.ceil.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @ceil_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1:[0-9]+]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.ceil.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.ceil.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @ceil_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @ceil_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.ceil.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @ceil_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.ceil.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @ceil_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.ceil.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @ceil_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.ceil.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.ceil.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.copysign.f64(double, double)
declare float @llvm.copysign.f32(float, float)

define void @copysign_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @copysign_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.copysign.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @copysign_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.copysign.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @copysign_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.copysign.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @copysign_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.copysign.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.copysign.f64(double %in, double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @copysign_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @copysign_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.copysign.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @copysign_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.copysign.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @copysign_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.copysign.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @copysign_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.copysign.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.copysign.f32(float %in, float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.cos.f64(double)
declare float @llvm.cos.f32(float)

define void @cos_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @cos_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_cos(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cos_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cos(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @cos_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vcosq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cos_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svcos_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.cos.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @cos_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @cos_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_cosf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cos_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cosf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @cos_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vcosq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cos_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svcos_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.cos.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.exp.f64(double)
declare float @llvm.exp.f32(float)

define void @exp_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @exp_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_exp(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @exp_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vexpq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @exp_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svexp_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.exp.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @exp_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @exp_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_expf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @exp_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vexpq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @exp_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svexp_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.exp.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.exp10.f64(double)
declare float @llvm.exp10.f32(float)

define void @exp10_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @exp10_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_exp10(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @exp10_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp10(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp10_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vexp10q_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @exp10_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svexp10_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.exp10.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @exp10_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @exp10_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_exp10f(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @exp10_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp10f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp10_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vexp10q_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @exp10_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svexp10_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.exp10.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.exp2.f64(double)
declare float @llvm.exp2.f32(float)

define void @exp2_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @exp2_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_exp2(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @exp2_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp2_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vexp2q_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @exp2_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svexp2_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.exp2.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @exp2_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @exp2_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_exp2f(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @exp2_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp2_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vexp2q_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @exp2_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svexp2_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.exp2.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.fabs.f64(double)
declare float @llvm.fabs.f32(float)

define void @fabs_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @fabs_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.fabs.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @fabs_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.fabs.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @fabs_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.fabs.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @fabs_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.fabs.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.fabs.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @fabs_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @fabs_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.fabs.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @fabs_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.fabs.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @fabs_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.fabs.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @fabs_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.fabs.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.fabs.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.floor.f64(double)
declare float @llvm.floor.f32(float)

define void @floor_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @floor_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.floor.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @floor_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.floor.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @floor_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.floor.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @floor_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.floor.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.floor.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @floor_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @floor_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.floor.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @floor_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.floor.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @floor_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.floor.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @floor_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.floor.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.floor.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.fma.f64(double, double, double)
declare float @llvm.fma.f32(float, float, float)

define void @fma_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @fma_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.fma.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fma_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.fma.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @fma_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.fma.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fma_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.fma.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.fma.f64(double %in, double %in, double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @fma_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @fma_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.fma.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fma_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.fma.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @fma_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.fma.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fma_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.fma.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.fma.f32(float %in, float %in, float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.log.f64(double)
declare float @llvm.log.f32(float)

define void @log_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @log_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_log(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @log_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vlogq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svlog_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.log.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @log_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @log_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_logf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_logf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @log_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vlogq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svlog_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.log.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.log10.f64(double)
declare float @llvm.log10.f32(float)

define void @log10_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @log10_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_log10(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log10_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log10(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @log10_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vlog10q_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log10_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svlog10_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.log10.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @log10_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @log10_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_log10f(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log10_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log10f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @log10_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vlog10q_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log10_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svlog10_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.log10.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.log2.f64(double)
declare float @llvm.log2.f32(float)

define void @log2_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @log2_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_log2(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log2_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @log2_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vlog2q_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log2_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svlog2_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.log2.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @log2_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @log2_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_log2f(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log2_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @log2_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vlog2q_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log2_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svlog2_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.log2.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.maxnum.f64(double, double)
declare float @llvm.maxnum.f32(float, float)

define void @maxnum_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @maxnum_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.maxnum.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @maxnum_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.maxnum.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @maxnum_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.maxnum.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @maxnum_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.maxnum.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.maxnum.f64(double %in, double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @maxnum_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @maxnum_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.maxnum.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @maxnum_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.maxnum.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @maxnum_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.maxnum.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @maxnum_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.maxnum.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.maxnum.f32(float %in, float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.minnum.f64(double, double)
declare float @llvm.minnum.f32(float, float)

define void @minnum_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @minnum_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.minnum.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @minnum_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.minnum.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @minnum_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.minnum.v2f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @minnum_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.minnum.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.minnum.f64(double %in, double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @minnum_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @minnum_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.minnum.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @minnum_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.minnum.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
; ARMPL-NEON-LABEL: define void @minnum_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.minnum.v4f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @minnum_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.minnum.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.minnum.f32(float %in, float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.nearbyint.f64(double)
declare float @llvm.nearbyint.f32(float)

define void @nearbyint_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @nearbyint_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.nearbyint.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @nearbyint_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.nearbyint.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @nearbyint_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.nearbyint.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @nearbyint_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.nearbyint.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.nearbyint.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @nearbyint_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @nearbyint_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.nearbyint.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @nearbyint_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.nearbyint.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @nearbyint_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.nearbyint.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @nearbyint_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.nearbyint.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.nearbyint.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.pow.f64(double, double)
declare float @llvm.pow.f32(float, float)

define void @pow_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @pow_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2vv_pow(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @pow_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_pow(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @pow_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vpowq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @pow_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svpow_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.pow.f64(double %in, double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @pow_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @pow_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4vv_powf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @pow_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_powf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @pow_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vpowq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @pow_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svpow_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.pow.f32(float %in, float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.rint.f64(double)
declare float @llvm.rint.f32(float)

define void @rint_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @rint_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.rint.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @rint_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.rint.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @rint_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.rint.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @rint_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.rint.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.rint.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @rint_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @rint_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.rint.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @rint_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.rint.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @rint_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.rint.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @rint_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.rint.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.rint.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.round.f64(double)
declare float @llvm.round.f32(float)

define void @round_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @round_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.round.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @round_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.round.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @round_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.round.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @round_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.round.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.round.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @round_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @round_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.round.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @round_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.round.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @round_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.round.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @round_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.round.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.round.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.sin.f64(double)
declare float @llvm.sin.f32(float)

define void @sin_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @sin_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @_ZGVnN2v_sin(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sin_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sin(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @sin_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @armpl_vsinq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sin_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @armpl_svsin_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.sin.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @sin_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @sin_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @_ZGVnN4v_sinf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sin_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-NEON-LABEL: define void @sin_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @armpl_vsinq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sin_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @armpl_svsin_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.sin.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.sqrt.f64(double)
declare float @llvm.sqrt.f32(float)

define void @sqrt_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @sqrt_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.sqrt.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sqrt_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.sqrt.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @sqrt_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.sqrt.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sqrt_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.sqrt.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.sqrt.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @sqrt_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @sqrt_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.sqrt.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sqrt_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.sqrt.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @sqrt_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.sqrt.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sqrt_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.sqrt.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.sqrt.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

declare double @llvm.trunc.f64(double)
declare float @llvm.trunc.f32(float)

define void @trunc_f64(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @trunc_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.trunc.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @trunc_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.trunc.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @trunc_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <2 x double> @llvm.trunc.v2f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @trunc_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 2 x double> @llvm.trunc.nxv2f64(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds double, ptr %in.ptr, i64 %iv
  %in = load double, ptr %in.gep, align 8
  %call = tail call double @llvm.trunc.f64(double %in)
  %out.gep = getelementptr inbounds double, ptr %out.ptr, i64 %iv
  store double %call, ptr %out.gep, align 8
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}

define void @trunc_f32(ptr noalias %in.ptr, ptr %out.ptr) {
; SLEEF-NEON-LABEL: define void @trunc_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.trunc.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @trunc_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; SLEEF-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.trunc.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
; ARMPL-NEON-LABEL: define void @trunc_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-NEON:    [[TMP3:%.*]] = call <4 x float> @llvm.trunc.v4f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @trunc_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr [[OUT_PTR:%.*]]) #[[ATTR1]] {
; ARMPL-SVE:    [[TMP15:%.*]] = call <vscale x 4 x float> @llvm.trunc.nxv4f32(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in.gep = getelementptr inbounds float, ptr %in.ptr, i64 %iv
  %in = load float, ptr %in.gep, align 8
  %call = tail call float @llvm.trunc.f32(float %in)
  %out.gep = getelementptr inbounds float, ptr %out.ptr, i64 %iv
  store float %call, ptr %out.gep, align 4
  %iv.next = add nuw nsw i64 %iv, 1
  %exitcond = icmp eq i64 %iv.next, 1000
  br i1 %exitcond, label %for.end, label %for.body

  for.end:
  ret void
}