; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sme2 -verify-machineinstrs < %s | FileCheck %s

; == 8 to 64-bit elements ==

define { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @sel_x4_i8(target("aarch64.svcount") %pn, <vscale x 16 x i8> %unused, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, <vscale x 16 x i8> %zn4, <vscale x 16 x i8> %zm1, <vscale x 16 x i8> %zm2, <vscale x 16 x i8> %zm3, <vscale x 16 x i8> %zm4) nounwind {
; CHECK-LABEL: sel_x4_i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.b
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1b { z27.b }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.b - z3.b }, pn8, { z28.b - z31.b }, { z24.b - z27.b }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x4.nxv16i8(target("aarch64.svcount") %pn, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, <vscale x 16 x i8> %zn4, <vscale x 16 x i8> %zm1, <vscale x 16 x i8> %zm2, <vscale x 16 x i8> %zm3, <vscale x 16 x i8> %zm4)
  ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } %res
}

define { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @sel_x4_i16(target("aarch64.svcount") %pn, <vscale x 8 x i16> %unused, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, <vscale x 8 x i16> %zn4, <vscale x 8 x i16> %zm1, <vscale x 8 x i16> %zm2, <vscale x 8 x i16> %zm3, <vscale x 8 x i16> %zm4) nounwind {
; CHECK-LABEL: sel_x4_i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.h
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1h { z27.h }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.h - z3.h }, pn8, { z28.h - z31.h }, { z24.h - z27.h }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") %pn, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, <vscale x 8 x i16> %zn4, <vscale x 8 x i16> %zm1, <vscale x 8 x i16> %zm2, <vscale x 8 x i16> %zm3, <vscale x 8 x i16> %zm4)
  ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } %res
}

define { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @sel_x4_f16(target("aarch64.svcount") %pn, <vscale x 8 x half> %unused, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, <vscale x 8 x half> %zn4, <vscale x 8 x half> %zm1, <vscale x 8 x half> %zm2, <vscale x 8 x half> %zm3, <vscale x 8 x half> %zm4) nounwind {
; CHECK-LABEL: sel_x4_f16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.h
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1h { z27.h }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.h - z3.h }, pn8, { z28.h - z31.h }, { z24.h - z27.h }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.sel.x4.nxv8f16(target("aarch64.svcount") %pn, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, <vscale x 8 x half> %zn4, <vscale x 8 x half> %zm1, <vscale x 8 x half> %zm2, <vscale x 8 x half> %zm3, <vscale x 8 x half> %zm4)
  ret { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } %res
}

define { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @sel_x4_bf16(target("aarch64.svcount") %pn, <vscale x 8 x bfloat> %unused, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, <vscale x 8 x bfloat> %zn4, <vscale x 8 x bfloat> %zm1, <vscale x 8 x bfloat> %zm2, <vscale x 8 x bfloat> %zm3, <vscale x 8 x bfloat> %zm4) nounwind {
; CHECK-LABEL: sel_x4_bf16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.h
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1h { z27.h }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.h - z3.h }, pn8, { z28.h - z31.h }, { z24.h - z27.h }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.sel.x4.nxv8bf16(target("aarch64.svcount") %pn, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, <vscale x 8 x bfloat> %zn4, <vscale x 8 x bfloat> %zm1, <vscale x 8 x bfloat> %zm2, <vscale x 8 x bfloat> %zm3, <vscale x 8 x bfloat> %zm4)
  ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } %res
}

define { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @sel_x4_i32(target("aarch64.svcount") %pn, <vscale x 4 x i32> %unused, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, <vscale x 4 x i32> %zn4, <vscale x 4 x i32> %zm1, <vscale x 4 x i32> %zm2, <vscale x 4 x i32> %zm3, <vscale x 4 x i32> %zm4) nounwind {
; CHECK-LABEL: sel_x4_i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.s
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1w { z27.s }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.s - z3.s }, pn8, { z28.s - z31.s }, { z24.s - z27.s }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") %pn, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, <vscale x 4 x i32> %zn4, <vscale x 4 x i32> %zm1, <vscale x 4 x i32> %zm2, <vscale x 4 x i32> %zm3, <vscale x 4 x i32> %zm4)
  ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } %res
}

define { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @sel_x4_f32(target("aarch64.svcount") %pn, <vscale x 4 x float> %unused, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, <vscale x 4 x float> %zn4, <vscale x 4 x float> %zm1, <vscale x 4 x float> %zm2, <vscale x 4 x float> %zm3, <vscale x 4 x float> %zm4) nounwind {
; CHECK-LABEL: sel_x4_f32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.s
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1w { z27.s }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.s - z3.s }, pn8, { z28.s - z31.s }, { z24.s - z27.s }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @llvm.aarch64.sve.sel.x4.nxv4f32(target("aarch64.svcount") %pn, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, <vscale x 4 x float> %zn4, <vscale x 4 x float> %zm1, <vscale x 4 x float> %zm2, <vscale x 4 x float> %zm3, <vscale x 4 x float> %zm4)
  ret { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } %res
}

define { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @sel_x4_i64(target("aarch64.svcount") %pn, <vscale x 2 x i64> %unused, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, <vscale x 2 x i64> %zn4, <vscale x 2 x i64> %zm1, <vscale x 2 x i64> %zm2, <vscale x 2 x i64> %zm3, <vscale x 2 x i64> %zm4) nounwind {
; CHECK-LABEL: sel_x4_i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.d
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1d { z27.d }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.d - z3.d }, pn8, { z28.d - z31.d }, { z24.d - z27.d }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") %pn, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, <vscale x 2 x i64> %zn4, <vscale x 2 x i64> %zm1, <vscale x 2 x i64> %zm2, <vscale x 2 x i64> %zm3, <vscale x 2 x i64> %zm4)
  ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } %res
}

define { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @sel_x4_f64(target("aarch64.svcount") %pn, <vscale x 2 x double> %unused, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, <vscale x 2 x double> %zn4, <vscale x 2 x double> %zm1, <vscale x 2 x double> %zm2, <vscale x 2 x double> %zm3, <vscale x 2 x double> %zm4) nounwind {
; CHECK-LABEL: sel_x4_f64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    str x29, [sp, #-16]! // 8-byte Folded Spill
; CHECK-NEXT:    addvl sp, sp, #-1
; CHECK-NEXT:    ptrue p1.d
; CHECK-NEXT:    str p8, [sp, #7, mul vl] // 2-byte Folded Spill
; CHECK-NEXT:    mov z26.d, z7.d
; CHECK-NEXT:    mov z31.d, z4.d
; CHECK-NEXT:    mov p8.b, p0.b
; CHECK-NEXT:    mov z25.d, z6.d
; CHECK-NEXT:    mov z30.d, z3.d
; CHECK-NEXT:    mov z24.d, z5.d
; CHECK-NEXT:    mov z29.d, z2.d
; CHECK-NEXT:    mov z28.d, z1.d
; CHECK-NEXT:    ld1d { z27.d }, p1/z, [x0]
; CHECK-NEXT:    sel { z0.d - z3.d }, pn8, { z28.d - z31.d }, { z24.d - z27.d }
; CHECK-NEXT:    ldr p8, [sp, #7, mul vl] // 2-byte Folded Reload
; CHECK-NEXT:    addvl sp, sp, #1
; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
; CHECK-NEXT:    ret
  %res = call { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @llvm.aarch64.sve.sel.x4.nxv2f64(target("aarch64.svcount") %pn, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, <vscale x 2 x double> %zn4, <vscale x 2 x double> %zm1, <vscale x 2 x double> %zm2, <vscale x 2 x double> %zm3, <vscale x 2 x double> %zm4)
  ret { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } %res
}


; == 8 to 64-bit elements ==
declare { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x4.nxv16i8(target("aarch64.svcount") %pn, <vscale x 16 x i8> %zn1, <vscale x 16 x i8> %zn2, <vscale x 16 x i8> %zn3, <vscale x 16 x i8> %zn4, <vscale x 16 x i8> %zm1, <vscale x 16 x i8> %zm2, <vscale x 16 x i8> %zm3, <vscale x 16 x i8> %zm4)
declare { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") %pn, <vscale x 8 x i16> %zn1, <vscale x 8 x i16> %zn2, <vscale x 8 x i16> %zn3, <vscale x 8 x i16> %zn4, <vscale x 8 x i16> %zm1, <vscale x 8 x i16> %zm2, <vscale x 8 x i16> %zm3, <vscale x 8 x i16> %zm4)
declare { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") %pn, <vscale x 4 x i32> %zn1, <vscale x 4 x i32> %zn2, <vscale x 4 x i32> %zn3, <vscale x 4 x i32> %zn4, <vscale x 4 x i32> %zm1, <vscale x 4 x i32> %zm2, <vscale x 4 x i32> %zm3, <vscale x 4 x i32> %zm4)
declare { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") %pn, <vscale x 2 x i64> %zn1, <vscale x 2 x i64> %zn2, <vscale x 2 x i64> %zn3, <vscale x 2 x i64> %zn4, <vscale x 2 x i64> %zm1, <vscale x 2 x i64> %zm2, <vscale x 2 x i64> %zm3, <vscale x 2 x i64> %zm4)
declare { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.sel.x4.nxv8f16(target("aarch64.svcount") %pn, <vscale x 8 x half> %zn1, <vscale x 8 x half> %zn2, <vscale x 8 x half> %zn3, <vscale x 8 x half> %zn4, <vscale x 8 x half> %zm1, <vscale x 8 x half> %zm2, <vscale x 8 x half> %zm3, <vscale x 8 x half> %zm4)
declare { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.sel.x4.nxv8bf16(target("aarch64.svcount") %pn, <vscale x 8 x bfloat> %zn1, <vscale x 8 x bfloat> %zn2, <vscale x 8 x bfloat> %zn3, <vscale x 8 x bfloat> %zn4, <vscale x 8 x bfloat> %zm1, <vscale x 8 x bfloat> %zm2, <vscale x 8 x bfloat> %zm3, <vscale x 8 x bfloat> %zm4)
declare { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @llvm.aarch64.sve.sel.x4.nxv4f32(target("aarch64.svcount") %pn, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3, <vscale x 4 x float> %zn4, <vscale x 4 x float> %zm1, <vscale x 4 x float> %zm2, <vscale x 4 x float> %zm3, <vscale x 4 x float> %zm4)
declare { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @llvm.aarch64.sve.sel.x4.nxv2f64(target("aarch64.svcount") %pn, <vscale x 2 x double> %zn1, <vscale x 2 x double> %zn2, <vscale x 2 x double> %zn3, <vscale x 2 x double> %zn4, <vscale x 2 x double> %zm1, <vscale x 2 x double> %zm2, <vscale x 2 x double> %zm3, <vscale x 2 x double> %zm4)