; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --scrub-attributes --filter "call.*(cos|sin|tan|cbrt|erf|exp[^e]|gamma|log|sqrt|copysign|dim|min|mod|hypot|nextafter|pow|fma)" --version 2
; RUN: opt -mattr=+neon -vector-library=LIBMVEC -passes=inject-tli-mappings,loop-vectorize,simplifycfg -force-vector-interleave=1 -S < %s | FileCheck %s -check-prefix=LIBMVEC-NEON
; RUN: opt -mattr=+neon -vector-library=LIBMVEC -passes=inject-tli-mappings,loop-vectorize,simplifycfg -force-vector-interleave=1 -force-vector-width=2 -S < %s | FileCheck %s -check-prefix=LIBMVEC-NEON-WIDTH-2
; RUN: opt -mattr=+sve -vector-library=LIBMVEC -passes=inject-tli-mappings,loop-vectorize,simplifycfg -mcpu=neoverse-v1 -force-vector-interleave=1 -prefer-predicate-over-epilogue=predicate-dont-vectorize -S < %s | FileCheck %s -check-prefix=LIBMVEC-SVE
; 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 -mcpu=neoverse-v1 -force-vector-interleave=1 -prefer-predicate-over-epilogue=predicate-dont-vectorize -S < %s | FileCheck %s -check-prefix=SLEEF-SVE
; RUN: opt -mattr=+sve -vector-library=sleefgnuabi -passes=inject-tli-mappings,loop-vectorize,simplifycfg -mcpu=neoverse-v1 -force-vector-interleave=1 -S < %s | FileCheck %s -check-prefixes=SLEEF-SVE-NOPRED
; 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 -mcpu=neoverse-v1 -force-vector-interleave=1 -prefer-predicate-over-epilogue=predicate-dont-vectorize -S < %s | FileCheck %s -check-prefix=ARMPL-SVE
; RUN: opt -mattr=+sve -vector-library=ArmPL -passes=inject-tli-mappings,loop-vectorize,simplifycfg -mcpu=neoverse-v1 -force-vector-interleave=1 -S < %s | FileCheck %s -check-prefixes=ARMPL-SVE-NOPRED



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

; We are checking whether loops containing function 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 @acos(double)
declare float @acosf(float)

define void @acos_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @acos_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_acos(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @acos_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_acos(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @acos_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_acos(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @acos_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_acos(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @acos_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_acos(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @acos_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_acos(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @acos(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @acos_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vacosq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @acos_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svacos_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @acos_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0:[0-9]+]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svacos_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @acos(double [[IN:%.*]])
;
  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 @acos(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 @acos_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @acos_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_acosf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @acos_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_acosf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @acos_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_acosf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @acos_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_acosf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @acos_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_acosf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @acos_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_acosf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @acosf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @acos_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vacosq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @acos_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svacos_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @acos_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svacos_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @acosf(float [[IN:%.*]])
;
  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 @acosf(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 @acosh(double)
declare float @acoshf(float)

define void @acosh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @acosh_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_acosh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @acosh_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_acosh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @acosh_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_acosh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @acosh_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_acosh(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @acosh_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_acosh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @acosh_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_acosh(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @acosh(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @acosh_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vacoshq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @acosh_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svacosh_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @acosh_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svacosh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @acosh(double [[IN:%.*]])
;
  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 @acosh(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 @acosh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @acosh_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_acoshf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @acosh_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_acoshf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @acosh_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_acoshf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @acosh_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_acoshf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @acosh_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_acoshf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @acosh_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_acoshf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @acoshf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @acosh_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vacoshq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @acosh_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svacosh_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @acosh_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svacosh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @acoshf(float [[IN:%.*]])
;
  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 @acoshf(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 @asin(double)
declare float @asinf(float)

define void @asin_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @asin_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_asin(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @asin_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_asin(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @asin_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_asin(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @asin_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_asin(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @asin_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_asin(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @asin_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_asin(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @asin(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @asin_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vasinq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @asin_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svasin_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @asin_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svasin_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @asin(double [[IN:%.*]])
;
  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 @asin(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 @asin_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @asin_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_asinf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @asin_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_asinf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @asin_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_asinf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @asin_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_asinf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @asin_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_asinf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @asin_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_asinf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @asinf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @asin_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vasinq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @asin_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svasin_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @asin_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svasin_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @asinf(float [[IN:%.*]])
;
  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 @asinf(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 @asinh(double)
declare float @asinhf(float)

define void @asinh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @asinh_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_asinh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @asinh_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_asinh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @asinh_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_asinh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @asinh_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_asinh(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @asinh_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_asinh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @asinh_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_asinh(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @asinh(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @asinh_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vasinhq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @asinh_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svasinh_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @asinh_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svasinh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @asinh(double [[IN:%.*]])
;
  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 @asinh(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 @asinh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @asinh_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_asinhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @asinh_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_asinhf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @asinh_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_asinhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @asinh_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_asinhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @asinh_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_asinhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @asinh_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_asinhf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @asinhf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @asinh_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vasinhq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @asinh_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svasinh_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @asinh_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svasinh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @asinhf(float [[IN:%.*]])
;
  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 @asinhf(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 @atan(double)
declare float @atanf(float)

define void @atan_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @atan_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_atan(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @atan_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_atan(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @atan_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_atan(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @atan_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_atan(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @atan_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_atan(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @atan_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_atan(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @atan(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @atan_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vatanq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @atan_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svatan_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @atan_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svatan_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @atan(double [[IN:%.*]])
;
  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 @atan(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 @atan_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @atan_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_atanf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @atan_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_atanf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @atan_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_atanf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @atan_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_atanf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @atan_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_atanf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @atan_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_atanf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @atanf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @atan_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vatanq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @atan_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svatan_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @atan_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svatan_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @atanf(float [[IN:%.*]])
;
  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 @atanf(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 @atan2(double, double)
declare float @atan2f(float, float)

define void @atan2_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @atan2_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2vv_atan2(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @atan2_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2vv_atan2(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; LIBMVEC-SVE-LABEL: define void @atan2_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_atan2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @atan2_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_atan2(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @atan2_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_atan2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @atan2_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_atan2(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @atan2(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @atan2_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vatan2q_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @atan2_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svatan2_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @atan2_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svatan2_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @atan2(double [[IN:%.*]], double [[IN]])
;
  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 @atan2(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 @atan2_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @atan2_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4vv_atan2f(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @atan2_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2vv_atan2f(<2 x float> [[WIDE_LOAD:%.*]], <2 x float> [[WIDE_LOAD]])
;
; LIBMVEC-SVE-LABEL: define void @atan2_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_atan2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @atan2_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_atan2f(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @atan2_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_atan2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @atan2_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_atan2f(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @atan2f(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @atan2_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vatan2q_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @atan2_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svatan2_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @atan2_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svatan2_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @atan2f(float [[IN:%.*]], float [[IN]])
;
  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 @atan2f(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 @atanh(double)
declare float @atanhf(float)

define void @atanh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @atanh_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_atanh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @atanh_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_atanh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @atanh_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_atanh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @atanh_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_atanh(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @atanh_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_atanh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @atanh_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_atanh(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @atanh(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @atanh_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vatanhq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @atanh_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svatanh_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @atanh_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svatanh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @atanh(double [[IN:%.*]])
;
  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 @atanh(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 @atanh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @atanh_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_atanhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @atanh_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_atanhf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @atanh_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_atanhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @atanh_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_atanhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @atanh_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_atanhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @atanh_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_atanhf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @atanhf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @atanh_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vatanhq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @atanh_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svatanh_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @atanh_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svatanh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @atanhf(float [[IN:%.*]])
;
  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 @atanhf(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 @cbrt(double)
declare float @cbrtf(float)

define void @cbrt_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cbrt_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_cbrt(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cbrt_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_cbrt(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cbrt_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cbrt(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @cbrt_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_cbrt(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cbrt_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cbrt(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cbrt_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cbrt(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cbrt(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cbrt_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vcbrtq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cbrt_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svcbrt_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cbrt_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svcbrt_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cbrt(double [[IN:%.*]])
;
  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 @cbrt(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 @cbrt_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cbrt_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_cbrtf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cbrt_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_cbrtf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cbrt_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cbrtf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @cbrt_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_cbrtf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cbrt_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cbrtf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cbrt_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cbrtf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @cbrtf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cbrt_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vcbrtq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cbrt_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svcbrt_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cbrt_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svcbrt_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @cbrtf(float [[IN:%.*]])
;
  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 @cbrtf(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 @copysign(double, double)
declare float @copysignf(float, float)

define void @copysign_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @copysign_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @copysign(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @copysign_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @copysign(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @copysign_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @copysign(double [[IN:%.*]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @copysign_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_copysign(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_copysign(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @copysign_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_copysign(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @copysign(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @copysign_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vcopysignq_f64(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svcopysign_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @copysign_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svcopysign_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @copysign(double [[IN:%.*]], double [[IN]])
;
  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 @copysign(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @copysign_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @copysignf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @copysign_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @copysignf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @copysign_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @copysignf(float [[IN:%.*]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @copysign_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_copysignf(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_copysignf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @copysign_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_copysignf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @copysignf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @copysign_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vcopysignq_f32(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svcopysign_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @copysign_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svcopysign_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @copysignf(float [[IN:%.*]], float [[IN]])
;
  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 @copysignf(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 @cos(double)
declare float @cosf(float)

define void @cos_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cos_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_cos(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cos_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_cos(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cos_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cos(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @cos_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cos(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cos_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cos(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cos(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cos_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cos_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svcos_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cos(double [[IN:%.*]])
;
  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 @cos(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cos_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_cosf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cos_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_cosf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cos_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cosf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @cos_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cosf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cos_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cosf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @cosf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cos_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cos_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svcos_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @cosf(float [[IN:%.*]])
;
  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 @cosf(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 @cosh(double)
declare float @coshf(float)

define void @cosh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cosh_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_cosh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cosh_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_cosh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cosh_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cosh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @cosh_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_cosh(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cosh_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cosh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cosh_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cosh(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cosh(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cosh_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vcoshq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cosh_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svcosh_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cosh_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svcosh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cosh(double [[IN:%.*]])
;
  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 @cosh(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 @cosh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cosh_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_coshf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cosh_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_coshf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cosh_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_coshf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @cosh_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_coshf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cosh_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_coshf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cosh_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_coshf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @coshf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cosh_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vcoshq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cosh_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svcosh_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cosh_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svcosh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @coshf(float [[IN:%.*]])
;
  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 @coshf(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 @cospi(double)
declare float @cospif(float)

define void @cospi_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cospi_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @cospi(double [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cospi_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @cospi(double [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cospi_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @cospi(double [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @cospi_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_cospi(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cospi_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cospi(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cospi_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_cospi(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cospi(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cospi_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vcospiq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cospi_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svcospi_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cospi_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svcospi_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @cospi(double [[IN:%.*]])
;
  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 @cospi(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 @cospi_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @cospi_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @cospif(float [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @cospi_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @cospif(float [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @cospi_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @cospif(float [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @cospi_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_cospif(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @cospi_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cospif(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @cospi_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_cospif(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @cospif(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @cospi_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vcospiq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @cospi_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svcospi_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @cospi_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svcospi_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @cospif(float [[IN:%.*]])
;
  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 @cospif(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 @erf(double)
declare float @erff(float)

define void @erf_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @erf_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_erf(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @erf_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_erf(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @erf_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_erf(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @erf_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_erf(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @erf_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_erf(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @erf_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_erf(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @erf(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @erf_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_verfq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @erf_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_sverf_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @erf_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_sverf_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @erf(double [[IN:%.*]])
;
  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 @erf(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 @erf_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @erf_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_erff(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @erf_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_erff(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @erf_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_erff(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @erf_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_erff(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @erf_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_erff(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @erf_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_erff(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @erff(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @erf_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_verfq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @erf_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_sverf_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @erf_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_sverf_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @erff(float [[IN:%.*]])
;
  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 @erff(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 @erfc(double)
declare float @erfcf(float)

define void @erfc_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @erfc_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_erfc(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @erfc_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_erfc(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @erfc_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_erfc(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @erfc_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_erfc(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @erfc_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_erfc(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @erfc_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_erfc(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @erfc(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @erfc_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_verfcq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @erfc_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_sverfc_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @erfc_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_sverfc_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @erfc(double [[IN:%.*]])
;
  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 @erfc(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 @erfc_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @erfc_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_erfcf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @erfc_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_erfcf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @erfc_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_erfcf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @erfc_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_erfcf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @erfc_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_erfcf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @erfc_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_erfcf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @erfcf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @erfc_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_verfcq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @erfc_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_sverfc_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @erfc_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_sverfc_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @erfcf(float [[IN:%.*]])
;
  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 @erfcf(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 @exp(double)
declare float @expf(float)

define void @exp_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @exp_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_exp(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @exp_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_exp(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @exp_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @exp_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @exp_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @exp(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @exp_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svexp_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @exp(double [[IN:%.*]])
;
  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 @exp(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @exp_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_expf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @exp_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_expf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @exp_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @exp_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @exp_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @expf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @exp_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svexp_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @expf(float [[IN:%.*]])
;
  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 @expf(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 @exp10(double)
declare float @exp10f(float)

define void @exp10_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @exp10_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_exp10(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @exp10_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_exp10(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @exp10_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp10(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @exp10_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp10(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @exp10_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp10(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @exp10(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp10_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @exp10_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svexp10_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @exp10(double [[IN:%.*]])
;
  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 @exp10(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @exp10_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_exp10f(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @exp10_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_exp10f(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @exp10_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp10f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @exp10_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp10f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @exp10_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp10f(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @exp10f(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp10_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @exp10_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svexp10_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @exp10f(float [[IN:%.*]])
;
  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 @exp10f(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 @exp2(double)
declare float @exp2f(float)

define void @exp2_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @exp2_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_exp2(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @exp2_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_exp2(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @exp2_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @exp2_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @exp2_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_exp2(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @exp2(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp2_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @exp2_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svexp2_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @exp2(double [[IN:%.*]])
;
  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 @exp2(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @exp2_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_exp2f(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @exp2_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_exp2f(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @exp2_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @exp2_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @exp2_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_exp2f(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @exp2f(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @exp2_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @exp2_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svexp2_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @exp2f(float [[IN:%.*]])
;
  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 @exp2f(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 @expm1(double)
declare float @expm1f(float)

define void @expm1_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @expm1_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_expm1(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @expm1_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_expm1(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @expm1_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_expm1(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @expm1_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_expm1(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @expm1_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_expm1(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @expm1_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_expm1(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @expm1(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @expm1_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vexpm1q_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @expm1_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svexpm1_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @expm1_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svexpm1_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @expm1(double [[IN:%.*]])
;
  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 @expm1(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 @expm1_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @expm1_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_expm1f(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @expm1_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_expm1f(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @expm1_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expm1f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @expm1_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_expm1f(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @expm1_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expm1f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @expm1_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_expm1f(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @expm1f(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @expm1_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vexpm1q_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @expm1_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svexpm1_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @expm1_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svexpm1_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @expm1f(float [[IN:%.*]])
;
  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 @expm1f(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 @fdim(double, double)
declare float @fdimf(float, float)

define void @fdim_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fdim_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @fdim(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fdim_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @fdim(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fdim_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @fdim(double [[IN:%.*]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @fdim_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_fdim(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fdim_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fdim(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fdim_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fdim(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fdim(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @fdim_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vfdimq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fdim_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svfdim_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fdim_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svfdim_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fdim(double [[IN:%.*]], double [[IN]])
;
  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 @fdim(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 @fdim_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fdim_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @fdimf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fdim_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @fdimf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fdim_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @fdimf(float [[IN:%.*]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @fdim_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_fdimf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fdim_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fdimf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fdim_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fdimf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fdimf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @fdim_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vfdimq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fdim_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svfdim_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fdim_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svfdim_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fdimf(float [[IN:%.*]], float [[IN]])
;
  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 @fdimf(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 @fma(double, double, double)
declare float @fmaf(float, float, float)

define void @fma_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fma_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @fma(double [[IN:%.*]], double [[IN]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fma_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @fma(double [[IN:%.*]], double [[IN]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fma_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @fma(double [[IN:%.*]], double [[IN]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @fma_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vvv_fma(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvvv_fma(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fma_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvvv_fma(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fma(double [[IN:%.*]], double [[IN]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @fma_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vfmaq_f64(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svfma_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <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-SVE-NOPRED-LABEL: define void @fma_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svfma_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fma(double [[IN:%.*]], double [[IN]], double [[IN]])
;
  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 @fma(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fma_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @fmaf(float [[IN:%.*]], float [[IN]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fma_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @fmaf(float [[IN:%.*]], float [[IN]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fma_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @fmaf(float [[IN:%.*]], float [[IN]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @fma_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vvv_fmaf(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvvv_fmaf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fma_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvvv_fmaf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fmaf(float [[IN:%.*]], float [[IN]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @fma_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vfmaq_f32(<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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svfma_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <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-SVE-NOPRED-LABEL: define void @fma_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svfma_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fmaf(float [[IN:%.*]], float [[IN]], float [[IN]])
;
  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 @fmaf(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 @fmax(double, double)
declare float @fmaxf(float, float)

define void @fmax_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fmax_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @fmax(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fmax_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @fmax(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fmax_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @fmax(double [[IN:%.*]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @fmax_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_fmax(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fmax_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fmax(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fmax_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fmax(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fmax(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @fmax_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vfmaxq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fmax_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svfmax_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fmax_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svfmax_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fmax(double [[IN:%.*]], double [[IN]])
;
  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 @fmax(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 @fmax_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fmax_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @fmaxf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fmax_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @fmaxf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fmax_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @fmaxf(float [[IN:%.*]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @fmax_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_fmaxf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fmax_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fmaxf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fmax_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fmaxf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fmaxf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @fmax_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vfmaxq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fmax_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svfmax_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fmax_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svfmax_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fmaxf(float [[IN:%.*]], float [[IN]])
;
  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 @fmaxf(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 @fmin(double, double)
declare float @fminf(float, float)

define void @fmin_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fmin_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @fmin(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fmin_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @fmin(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fmin_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @fmin(double [[IN:%.*]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @fmin_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_fmin(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fmin_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fmin(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fmin_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fmin(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fmin(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @fmin_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vfminq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fmin_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svfmin_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fmin_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svfmin_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fmin(double [[IN:%.*]], double [[IN]])
;
  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 @fmin(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 @fmin_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fmin_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @fminf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fmin_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @fminf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fmin_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @fminf(float [[IN:%.*]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @fmin_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_fminf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fmin_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fminf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fmin_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fminf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fminf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @fmin_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vfminq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fmin_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svfmin_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fmin_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svfmin_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fminf(float [[IN:%.*]], float [[IN]])
;
  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 @fminf(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 @fmod(double, double)
declare float @fmodf(float, float)

define void @fmod_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fmod_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @fmod(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fmod_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @fmod(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fmod_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @fmod(double [[IN:%.*]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @fmod_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_fmod(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fmod_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fmod(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fmod_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_fmod(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fmod(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @fmod_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vfmodq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fmod_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svfmod_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fmod_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svfmod_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @fmod(double [[IN:%.*]], double [[IN]])
;
  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 @fmod(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 @fmod_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @fmod_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @fmodf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @fmod_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @fmodf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @fmod_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @fmodf(float [[IN:%.*]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @fmod_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_fmodf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @fmod_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fmodf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @fmod_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_fmodf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fmodf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @fmod_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vfmodq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @fmod_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svfmod_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @fmod_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svfmod_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @fmodf(float [[IN:%.*]], float [[IN]])
;
  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 @fmodf(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 @hypot(double, double)
declare float @hypotf(float, float)

define void @hypot_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @hypot_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2vv_hypot(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @hypot_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2vv_hypot(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; LIBMVEC-SVE-LABEL: define void @hypot_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_hypot(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @hypot_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_hypot(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @hypot_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_hypot(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @hypot_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_hypot(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @hypot(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @hypot_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vhypotq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @hypot_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svhypot_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @hypot_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svhypot_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @hypot(double [[IN:%.*]], double [[IN]])
;
  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 @hypot(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 @hypot_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @hypot_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4vv_hypotf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @hypot_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2vv_hypotf(<2 x float> [[WIDE_LOAD:%.*]], <2 x float> [[WIDE_LOAD]])
;
; LIBMVEC-SVE-LABEL: define void @hypot_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_hypotf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @hypot_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_hypotf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @hypot_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_hypotf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @hypot_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_hypotf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @hypotf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @hypot_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vhypotq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @hypot_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svhypot_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @hypot_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svhypot_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @hypotf(float [[IN:%.*]], float [[IN]])
;
  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 @hypotf(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 i32 @ilogb(double)
declare i32 @ilogbf(float)

define void @ilogb_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @ilogb_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call i32 @ilogb(double [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @ilogb_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call i32 @ilogb(double [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @ilogb_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call i32 @ilogb(double [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @ilogb_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x i32> @_ZGVnN2v_ilogb(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @ilogb_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x i32> @_ZGVsMxv_ilogb(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @ilogb_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x i32> @_ZGVsMxv_ilogb(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call i32 @ilogb(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @ilogb_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x i32> @armpl_vilogbq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @ilogb_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x i32> @armpl_svilogb_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @ilogb_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x i32> @armpl_svilogb_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call i32 @ilogb(double [[IN:%.*]])
;
  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 i32 @ilogb(double %in)
  %out.gep = getelementptr inbounds i32, ptr %out.ptr, i64 %iv
  store i32 %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 @ilogb_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @ilogb_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call i32 @ilogbf(float [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @ilogb_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call i32 @ilogbf(float [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @ilogb_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call i32 @ilogbf(float [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @ilogb_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x i32> @_ZGVnN4v_ilogbf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @ilogb_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x i32> @_ZGVsMxv_ilogbf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @ilogb_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x i32> @_ZGVsMxv_ilogbf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call i32 @ilogbf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @ilogb_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x i32> @armpl_vilogbq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @ilogb_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x i32> @armpl_svilogb_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @ilogb_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x i32> @armpl_svilogb_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call i32 @ilogbf(float [[IN:%.*]])
;
  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 i32 @ilogbf(float %in)
  %out.gep = getelementptr inbounds i32, ptr %out.ptr, i64 %iv
  store i32 %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 @ldexp(double, i32)
declare float @ldexpf(float, i32)

define void @ldexp_f64(ptr noalias %in1.ptr, ptr noalias %in2.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @ldexp_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @ldexp(double [[IN1:%.*]], i32 [[IN2:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @ldexp_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @ldexp(double [[IN1:%.*]], i32 [[IN2:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @ldexp_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @ldexp(double [[IN1:%.*]], i32 [[IN2:%.*]])
;
; SLEEF-NEON-LABEL: define void @ldexp_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP4:%.*]] = call <2 x double> @_ZGVnN2vv_ldexp(<2 x double> [[WIDE_LOAD:%.*]], <2 x i32> [[WIDE_LOAD1:%.*]])
;
; SLEEF-SVE-LABEL: define void @ldexp_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP14:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_ldexp(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i32> [[WIDE_MASKED_LOAD1:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @ldexp_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP10:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_ldexp(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i32> [[WIDE_LOAD1:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @ldexp(double [[IN1:%.*]], i32 [[IN2:%.*]])
;
; ARMPL-NEON-LABEL: define void @ldexp_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP4:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vldexpq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x i32> [[WIDE_LOAD1:%.*]])
;
; ARMPL-SVE-LABEL: define void @ldexp_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP14:%.*]] = call <vscale x 2 x double> @armpl_svldexp_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i32> [[WIDE_MASKED_LOAD1:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @ldexp_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP10:%.*]] = call <vscale x 2 x double> @armpl_svldexp_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i32> [[WIDE_LOAD1:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @ldexp(double [[IN1:%.*]], i32 [[IN2:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in1.gep = getelementptr inbounds double, ptr %in1.ptr, i64 %iv
  %in1 = load double, ptr %in1.gep, align 8
  %in2.gep = getelementptr inbounds i32, ptr %in2.ptr, i64 %iv
  %in2 = load i32, ptr %in2.gep, align 8
  %call = tail call double @ldexp(double %in1, i32 %in2)
  %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 @ldexp_f32(ptr noalias %in1.ptr, ptr noalias %in2.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @ldexp_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @ldexpf(float [[IN1:%.*]], i32 [[IN2:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @ldexp_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @ldexpf(float [[IN1:%.*]], i32 [[IN2:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @ldexp_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @ldexpf(float [[IN1:%.*]], i32 [[IN2:%.*]])
;
; SLEEF-NEON-LABEL: define void @ldexp_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP4:%.*]] = call <4 x float> @_ZGVnN4vv_ldexpf(<4 x float> [[WIDE_LOAD:%.*]], <4 x i32> [[WIDE_LOAD1:%.*]])
;
; SLEEF-SVE-LABEL: define void @ldexp_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP14:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_ldexpf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i32> [[WIDE_MASKED_LOAD1:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @ldexp_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP10:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_ldexpf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i32> [[WIDE_LOAD1:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @ldexpf(float [[IN1:%.*]], i32 [[IN2:%.*]])
;
; ARMPL-NEON-LABEL: define void @ldexp_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP4:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vldexpq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x i32> [[WIDE_LOAD1:%.*]])
;
; ARMPL-SVE-LABEL: define void @ldexp_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP14:%.*]] = call <vscale x 4 x float> @armpl_svldexp_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i32> [[WIDE_MASKED_LOAD1:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @ldexp_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN1_PTR:%.*]], ptr noalias [[IN2_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP10:%.*]] = call <vscale x 4 x float> @armpl_svldexp_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i32> [[WIDE_LOAD1:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @ldexpf(float [[IN1:%.*]], i32 [[IN2:%.*]])
;
  entry:
  br label %for.body

  for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %in1.gep = getelementptr inbounds float, ptr %in1.ptr, i64 %iv
  %in1 = load float, ptr %in1.gep, align 8
  %in2.gep = getelementptr inbounds i32, ptr %in2.ptr, i64 %iv
  %in2 = load i32, ptr %in2.gep, align 8
  %call = tail call float @ldexpf(float %in1, i32 %in2)
  %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 @lgamma(double)
declare float @lgammaf(float)

define void @lgamma_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @lgamma_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @lgamma(double [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @lgamma_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @lgamma(double [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @lgamma_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @lgamma(double [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @lgamma_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_lgamma(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @lgamma_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_lgamma(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @lgamma_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_lgamma(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @lgamma(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @lgamma_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vlgammaq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @lgamma_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svlgamma_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @lgamma_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svlgamma_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @lgamma(double [[IN:%.*]])
;
  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 @lgamma(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 @lgamma_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @lgamma_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @lgammaf(float [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @lgamma_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @lgammaf(float [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @lgamma_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @lgammaf(float [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @lgamma_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_lgammaf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @lgamma_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_lgammaf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @lgamma_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_lgammaf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @lgammaf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @lgamma_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vlgammaq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @lgamma_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svlgamma_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @lgamma_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svlgamma_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @lgammaf(float [[IN:%.*]])
;
  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 @lgammaf(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 @log(double)
declare float @logf(float)

define void @log_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svlog_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log(double [[IN:%.*]])
;
  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 @log(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_logf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_logf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_logf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_logf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_logf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @logf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svlog_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @logf(float [[IN:%.*]])
;
  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 @logf(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 @log10(double)
declare float @log10f(float)

define void @log10_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log10_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log10(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log10_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log10(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log10_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log10(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log10_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log10(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log10_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log10(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log10(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log10_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log10_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svlog10_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log10(double [[IN:%.*]])
;
  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 @log10(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log10_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_log10f(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log10_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_log10f(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log10_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log10f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log10_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log10f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log10_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log10f(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @log10f(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log10_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log10_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svlog10_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @log10f(float [[IN:%.*]])
;
  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 @log10f(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 @log1p(double)
declare float @log1pf(float)

define void @log1p_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log1p_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log1p(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log1p_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log1p(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log1p_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log1p(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log1p_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_log1p(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log1p_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log1p(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log1p_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log1p(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log1p(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log1p_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vlog1pq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log1p_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svlog1p_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log1p_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svlog1p_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log1p(double [[IN:%.*]])
;
  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 @log1p(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 @log1p_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log1p_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_log1pf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log1p_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_log1pf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log1p_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log1pf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log1p_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_log1pf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @log1p_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log1pf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log1p_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log1pf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @log1pf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log1p_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vlog1pq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @log1p_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svlog1p_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log1p_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svlog1p_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @log1pf(float [[IN:%.*]])
;
  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 @log1pf(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 @log2(double)
declare float @log2f(float)

define void @log2_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log2_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log2(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log2_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_log2(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log2_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log2_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log2(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log2_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_log2(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log2(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log2_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log2_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svlog2_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @log2(double [[IN:%.*]])
;
  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 @log2(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @log2_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_log2f(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @log2_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_log2f(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @log2_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @log2_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log2f(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @log2_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_log2f(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @log2f(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @log2_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @log2_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svlog2_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @log2f(float [[IN:%.*]])
;
  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 @log2f(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
}

; FIXME: Re-enable modf[f] vectorization once aliasing issues due to output
; pointers have been resolved.

declare double @modf(double, ptr)
declare float @modff(float, ptr)

define void @modf_f64(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
; LIBMVEC-NEON-LABEL: define void @modf_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @modf_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @modf_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; SLEEF-NEON-LABEL: define void @modf_f64
; SLEEF-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; SLEEF-SVE-LABEL: define void @modf_f64
; SLEEF-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @modf_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; ARMPL-NEON-LABEL: define void @modf_f64
; ARMPL-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; ARMPL-SVE-LABEL: define void @modf_f64
; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @modf_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]])
;
entry:
  br label %for.body

for.body:
  %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
  %gepa = getelementptr double, ptr %a, i64 %indvars.iv
  %num = load double, ptr %gepa, align 8
  %gepb = getelementptr double, ptr %b, i64 %indvars.iv
  %data = call double @modf(double %num, ptr %gepb)
  %gepc = getelementptr inbounds double, ptr %c, i64 %indvars.iv
  store double %data, ptr %gepc, align 8
  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
  %exitcond = icmp eq i64 %indvars.iv.next, 1000
  br i1 %exitcond, label %for.cond.cleanup, label %for.body

for.cond.cleanup:
  ret void
}

define void @modf_f32(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
; LIBMVEC-NEON-LABEL: define void @modf_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @modf_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @modf_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; SLEEF-NEON-LABEL: define void @modf_f32
; SLEEF-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; SLEEF-SVE-LABEL: define void @modf_f32
; SLEEF-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @modf_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; ARMPL-NEON-LABEL: define void @modf_f32
; ARMPL-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; ARMPL-SVE-LABEL: define void @modf_f32
; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @modf_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]])
;
entry:
  br label %for.body

for.body:
  %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
  %gepa = getelementptr float, ptr %a, i64 %indvars.iv
  %num = load float, ptr %gepa, align 8
  %gepb = getelementptr float, ptr %b, i64 %indvars.iv
  %data = call float @modff(float %num, ptr %gepb)
  %gepc = getelementptr inbounds float, ptr %c, i64 %indvars.iv
  store float %data, ptr %gepc, align 8
  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
  %exitcond = icmp eq i64 %indvars.iv.next, 1000
  br i1 %exitcond, label %for.cond.cleanup, label %for.body

for.cond.cleanup:
  ret void
}

declare double @nextafter(double, double)
declare float @nextafterf(float, float)

define void @nextafter_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @nextafter_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @nextafter_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @nextafter_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]])
;
; SLEEF-NEON-LABEL: define void @nextafter_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2vv_nextafter(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @nextafter_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_nextafter(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x double> [[WIDE_MASKED_LOAD]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @nextafter_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_nextafter(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @nextafter_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vnextafterq_f64(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @nextafter_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svnextafter_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @nextafter_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svnextafter_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]])
;
  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 @nextafter(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 @nextafter_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @nextafter_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @nextafter_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]])
;
; LIBMVEC-SVE-LABEL: define void @nextafter_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]])
;
; SLEEF-NEON-LABEL: define void @nextafter_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4vv_nextafterf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; SLEEF-SVE-LABEL: define void @nextafter_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_nextafterf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x float> [[WIDE_MASKED_LOAD]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @nextafter_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_nextafterf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @nextafter_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vnextafterq_f32(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; ARMPL-SVE-LABEL: define void @nextafter_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svnextafter_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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @nextafter_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svnextafter_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]])
;
  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 @nextafterf(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 @pow(double, double)
declare float @powf(float, float)

define void @pow_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @pow_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2vv_pow(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @pow_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2vv_pow(<2 x double> [[WIDE_LOAD:%.*]], <2 x double> [[WIDE_LOAD]])
;
; LIBMVEC-SVE-LABEL: define void @pow_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; SLEEF-NEON-LABEL: define void @pow_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @pow_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxvv_pow(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @pow(double [[IN:%.*]], double [[IN]])
;
; ARMPL-NEON-LABEL: define void @pow_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @pow_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svpow_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @pow(double [[IN:%.*]], double [[IN]])
;
  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 @pow(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @pow_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4vv_powf(<4 x float> [[WIDE_LOAD:%.*]], <4 x float> [[WIDE_LOAD]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @pow_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2vv_powf(<2 x float> [[WIDE_LOAD:%.*]], <2 x float> [[WIDE_LOAD]])
;
; LIBMVEC-SVE-LABEL: define void @pow_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; SLEEF-NEON-LABEL: define void @pow_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @pow_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxvv_powf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @powf(float [[IN:%.*]], float [[IN]])
;
; ARMPL-NEON-LABEL: define void @pow_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @pow_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svpow_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @powf(float [[IN:%.*]], float [[IN]])
;
  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 @powf(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 @sin(double)
declare float @sinf(float)

define void @sin_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sin_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_sin(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sin_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_sin(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sin_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sin(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @sin_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sin(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sin_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sin(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sin(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sin_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sin_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svsin_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sin(double [[IN:%.*]])
;
  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 @sin(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sin_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_sinf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sin_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_sinf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sin_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @sin_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = 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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sin_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sin_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <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 noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = 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:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sin_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svsin_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinf(float [[IN:%.*]])
;
  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 @sinf(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
}

; FIXME: Re-enable sincos[f] vectorization once aliasing issues with output
; pointers have been resolved.

declare void @sincos(double, ptr, ptr)
declare void @sincosf(float, ptr, ptr)

define void @sincos_f64(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
; LIBMVEC-NEON-LABEL: define void @sincos_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sincos_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sincos_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-NEON-LABEL: define void @sincos_f64
; SLEEF-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-LABEL: define void @sincos_f64
; SLEEF-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sincos_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-NEON-LABEL: define void @sincos_f64
; ARMPL-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-LABEL: define void @sincos_f64
; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sincos_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
entry:
  br label %for.body

for.body:
  %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
  %gepa = getelementptr double, ptr %a, i64 %indvars.iv
  %num = load double, ptr %gepa, align 8
  %gepb = getelementptr double, ptr %b, i64 %indvars.iv
  %gepc = getelementptr double, ptr %c, i64 %indvars.iv
  call void @sincos(double %num, ptr %gepb, ptr %gepc)
  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
  %exitcond = icmp eq i64 %indvars.iv.next, 1000
  br i1 %exitcond, label %for.cond.cleanup, label %for.body

for.cond.cleanup:
  ret void
}

define void @sincos_f32(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
; LIBMVEC-NEON-LABEL: define void @sincos_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sincos_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sincos_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-NEON-LABEL: define void @sincos_f32
; SLEEF-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-LABEL: define void @sincos_f32
; SLEEF-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sincos_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-NEON-LABEL: define void @sincos_f32
; ARMPL-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-LABEL: define void @sincos_f32
; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sincos_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
entry:
  br label %for.body

for.body:
  %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
  %gepa = getelementptr float, ptr %a, i64 %indvars.iv
  %num = load float, ptr %gepa, align 8
  %gepb = getelementptr float, ptr %b, i64 %indvars.iv
  %gepc = getelementptr float, ptr %c, i64 %indvars.iv
  call void @sincosf(float %num, ptr %gepb, ptr %gepc)
  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
  %exitcond = icmp eq i64 %indvars.iv.next, 1000
  br i1 %exitcond, label %for.cond.cleanup, label %for.body

for.cond.cleanup:
  ret void
}

; FIXME: Re-enable sincospi[f] vectorization once aliasing issues with output
; pointers have been resolved.

declare void @sincospi(double, ptr, ptr)
declare void @sincospif(float, ptr, ptr)

define void @sincospi_f64(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
; LIBMVEC-NEON-LABEL: define void @sincospi_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sincospi_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sincospi_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-NEON-LABEL: define void @sincospi_f64
; SLEEF-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-LABEL: define void @sincospi_f64
; SLEEF-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sincospi_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-NEON-LABEL: define void @sincospi_f64
; ARMPL-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-LABEL: define void @sincospi_f64
; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sincospi_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
entry:
  br label %for.body

for.body:
  %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
  %gepa = getelementptr double, ptr %a, i64 %indvars.iv
  %num = load double, ptr %gepa, align 8
  %gepb = getelementptr double, ptr %b, i64 %indvars.iv
  %gepc = getelementptr double, ptr %c, i64 %indvars.iv
  call void @sincospi(double %num, ptr %gepb, ptr %gepc)
  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
  %exitcond = icmp eq i64 %indvars.iv.next, 1000
  br i1 %exitcond, label %for.cond.cleanup, label %for.body

for.cond.cleanup:
  ret void
}

define void @sincospi_f32(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
; LIBMVEC-NEON-LABEL: define void @sincospi_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sincospi_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sincospi_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-NEON-LABEL: define void @sincospi_f32
; SLEEF-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-LABEL: define void @sincospi_f32
; SLEEF-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sincospi_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-NEON-LABEL: define void @sincospi_f32
; ARMPL-NEON-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-LABEL: define void @sincospi_f32
; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sincospi_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]])
;
entry:
  br label %for.body

for.body:
  %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
  %gepa = getelementptr float, ptr %a, i64 %indvars.iv
  %num = load float, ptr %gepa, align 8
  %gepb = getelementptr float, ptr %b, i64 %indvars.iv
  %gepc = getelementptr float, ptr %c, i64 %indvars.iv
  call void @sincospif(float %num, ptr %gepb, ptr %gepc)
  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
  %exitcond = icmp eq i64 %indvars.iv.next, 1000
  br i1 %exitcond, label %for.cond.cleanup, label %for.body

for.cond.cleanup:
  ret void
}

declare double @sinh(double)
declare float @sinhf(float)

define void @sinh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sinh_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_sinh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sinh_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_sinh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sinh_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sinh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @sinh_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_sinh(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sinh_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sinh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sinh_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sinh(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinh(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sinh_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vsinhq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sinh_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svsinh_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sinh_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svsinh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinh(double [[IN:%.*]])
;
  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 @sinh(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 @sinh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sinh_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_sinhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sinh_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_sinhf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sinh_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @sinh_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_sinhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sinh_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sinh_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinhf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinhf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sinh_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vsinhq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sinh_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svsinh_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sinh_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svsinh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinhf(float [[IN:%.*]])
;
  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 @sinhf(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 @sinpi(double)
declare float @sinpif(float)

define void @sinpi_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sinpi_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sinpi_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sinpi_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @sinpi_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_sinpi(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sinpi_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sinpi(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sinpi_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sinpi(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sinpi_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vsinpiq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sinpi_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svsinpi_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sinpi_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svsinpi_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]])
;
  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 @sinpi(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 @sinpi_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sinpi_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sinpi_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sinpi_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @sinpi_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_sinpif(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sinpi_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinpif(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sinpi_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sinpif(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sinpi_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vsinpiq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sinpi_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svsinpi_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sinpi_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svsinpi_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]])
;
  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 @sinpif(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 @sqrt(double)
declare float @sqrtf(float)

define void @sqrt_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sqrt_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sqrt_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sqrt_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @sqrt_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_sqrt(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sqrt_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sqrt(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sqrt_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_sqrt(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sqrt_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vsqrtq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sqrt_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svsqrt_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sqrt_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svsqrt_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]])
;
  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 @sqrt(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 noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @sqrt_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @sqrt_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @sqrt_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @sqrt_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_sqrtf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @sqrt_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sqrtf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @sqrt_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_sqrtf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @sqrt_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vsqrtq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @sqrt_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svsqrt_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @sqrt_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svsqrt_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]])
;
  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 @sqrtf(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 @tan(double)
declare float @tanf(float)

define void @tan_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @tan_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_tan(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @tan_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_tan(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @tan_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tan(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @tan_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_tan(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @tan_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tan(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @tan_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tan(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tan(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @tan_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vtanq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @tan_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svtan_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @tan_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svtan_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tan(double [[IN:%.*]])
;
  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 @tan(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 @tan_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @tan_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_tanf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @tan_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_tanf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @tan_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tanf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @tan_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_tanf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @tan_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tanf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @tan_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tanf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @tanf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @tan_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vtanq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @tan_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svtan_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @tan_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svtan_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @tanf(float [[IN:%.*]])
;
  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 @tanf(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 @tanh(double)
declare float @tanhf(float)

define void @tanh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @tanh_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_tanh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @tanh_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @_ZGVnN2v_tanh(<2 x double> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @tanh_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tanh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @tanh_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_tanh(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @tanh_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tanh(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @tanh_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tanh(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tanh(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @tanh_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vtanhq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @tanh_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svtanh_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @tanh_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svtanh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tanh(double [[IN:%.*]])
;
  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 @tanh(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 @tanh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @tanh_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @_ZGVnN4v_tanhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @tanh_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x float> @_ZGVnN2v_tanhf(<2 x float> [[WIDE_LOAD:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @tanh_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tanhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-NEON-LABEL: define void @tanh_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_tanhf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @tanh_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tanhf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @tanh_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tanhf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @tanhf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @tanh_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vtanhq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @tanh_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svtanh_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @tanh_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svtanh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @tanhf(float [[IN:%.*]])
;
  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 @tanhf(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 @tgamma(double)
declare float @tgammaf(float)

define void @tgamma_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @tgamma_f64
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call double @tgamma(double [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @tgamma_f64
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call double @tgamma(double [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @tgamma_f64
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call double @tgamma(double [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @tgamma_f64
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <2 x double> @_ZGVnN2v_tgamma(<2 x double> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @tgamma_f64
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tgamma(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @tgamma_f64
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @_ZGVsMxv_tgamma(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tgamma(double [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @tgamma_f64
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <2 x double> @armpl_vtgammaq_f64(<2 x double> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @tgamma_f64
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 2 x double> @armpl_svtgamma_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @tgamma_f64
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 2 x double> @armpl_svtgamma_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tgamma(double [[IN:%.*]])
;
  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 @tgamma(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 @tgamma_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
; LIBMVEC-NEON-LABEL: define void @tgamma_f32
; LIBMVEC-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON:    [[CALL:%.*]] = tail call float @tgammaf(float [[IN:%.*]])
;
; LIBMVEC-NEON-WIDTH-2-LABEL: define void @tgamma_f32
; LIBMVEC-NEON-WIDTH-2-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-NEON-WIDTH-2:    [[CALL:%.*]] = tail call float @tgammaf(float [[IN:%.*]])
;
; LIBMVEC-SVE-LABEL: define void @tgamma_f32
; LIBMVEC-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; LIBMVEC-SVE:    [[CALL:%.*]] = tail call float @tgammaf(float [[IN:%.*]])
;
; SLEEF-NEON-LABEL: define void @tgamma_f32
; SLEEF-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-NEON:    [[TMP2:%.*]] = call <4 x float> @_ZGVnN4v_tgammaf(<4 x float> [[WIDE_LOAD:%.*]])
;
; SLEEF-SVE-LABEL: define void @tgamma_f32
; SLEEF-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tgammaf(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; SLEEF-SVE-NOPRED-LABEL: define void @tgamma_f32
; SLEEF-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; SLEEF-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @_ZGVsMxv_tgammaf(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; SLEEF-SVE-NOPRED:    [[CALL:%.*]] = tail call float @tgammaf(float [[IN:%.*]])
;
; ARMPL-NEON-LABEL: define void @tgamma_f32
; ARMPL-NEON-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-NEON:    [[TMP2:%.*]] = call aarch64_vector_pcs <4 x float> @armpl_vtgammaq_f32(<4 x float> [[WIDE_LOAD:%.*]])
;
; ARMPL-SVE-LABEL: define void @tgamma_f32
; ARMPL-SVE-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE:    [[TMP12:%.*]] = call <vscale x 4 x float> @armpl_svtgamma_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
;
; ARMPL-SVE-NOPRED-LABEL: define void @tgamma_f32
; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
; ARMPL-SVE-NOPRED:    [[TMP8:%.*]] = call <vscale x 4 x float> @armpl_svtgamma_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> splat (i1 true))
; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @tgammaf(float [[IN:%.*]])
;
  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 @tgammaf(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
}