llvm/test/CodeGen/AArch64/sme2-fp8-intrinsics-cvt.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sme2,+fp8 -verify-machineinstrs -force-streaming < %s | FileCheck %s

; FCVT / FCVTN / BFCVT

define <vscale x 16 x i8> @fcvt_x2(<vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1) {
; CHECK-LABEL: fcvt_x2:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
; CHECK-NEXT:    fcvt z0.b, { z0.h, z1.h }
; CHECK-NEXT:    ret
  %res = call <vscale x 16 x i8> @llvm.aarch64.sve.fp8.cvt.x2.nxv8f16(<vscale x 8 x half> %zn0, <vscale x 8 x half> %zn1)
  ret <vscale x 16 x i8> %res
}

define <vscale x 16 x i8> @fcvt_x4(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3) {
; CHECK-LABEL: fcvt_x4:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $z3 killed $z3 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    // kill: def $z2 killed $z2 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    fcvt z0.b, { z0.s - z3.s }
; CHECK-NEXT:    ret
  %res = call <vscale x 16 x i8> @llvm.aarch64.sve.fp8.cvt.x4(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1,
                                                              <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3)
  ret <vscale x 16 x i8> %res
}

define <vscale x 16 x i8> @fcvtn(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1, <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3) {
; CHECK-LABEL: fcvtn:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $z3 killed $z3 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    // kill: def $z2 killed $z2 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
; CHECK-NEXT:    fcvtn z0.b, { z0.s - z3.s }
; CHECK-NEXT:    ret
  %res = call <vscale x 16 x i8> @llvm.aarch64.sve.fp8.cvtn.x4(<vscale x 4 x float> %zn0, <vscale x 4 x float> %zn1,
                                                               <vscale x 4 x float> %zn2, <vscale x 4 x float> %zn3)
  ret <vscale x 16 x i8> %res
}

define <vscale x 16 x i8> @bfcvt(<vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1) {
; CHECK-LABEL: bfcvt:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
; CHECK-NEXT:    bfcvt z0.b, { z0.h, z1.h }
; CHECK-NEXT:    ret
  %res = call <vscale x 16 x i8> @llvm.aarch64.sve.fp8.cvt.x2.nxv8bf16(<vscale x 8 x bfloat> %zn0, <vscale x 8 x bfloat> %zn1)
  ret <vscale x 16 x i8> %res
}

; F1CVT / F2CVT

define { <vscale x 8 x half>, <vscale x 8 x half> } @f1cvt(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: f1cvt:
; CHECK:       // %bb.0:
; CHECK-NEXT:    f1cvt { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.fp8.cvt1.x2.nxv8f16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x half>, <vscale x 8 x half> } %res
}

define { <vscale x 8 x half>, <vscale x 8 x half> } @f2cvt(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: f2cvt:
; CHECK:       // %bb.0:
; CHECK-NEXT:    f2cvt { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.fp8.cvt2.x2.nxv8f16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x half>, <vscale x 8 x half> } %res
}

; BF1CVT / BF2CVT

define { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @bf1cvt(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: bf1cvt:
; CHECK:       // %bb.0:
; CHECK-NEXT:    bf1cvt { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.fp8.cvt1.x2.nxv8bf16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } %res
}

define { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @bf2cvt(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: bf2cvt:
; CHECK:       // %bb.0:
; CHECK-NEXT:    bf2cvt { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.fp8.cvt2.x2.nxv8bf16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } %res
}

; F1CVTL / F2CVTL

define { <vscale x 8 x half>, <vscale x 8 x half> } @f1cvtl(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: f1cvtl:
; CHECK:       // %bb.0:
; CHECK-NEXT:    f1cvtl { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.fp8.cvtl1.x2.nxv8f16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x half>, <vscale x 8 x half> } %res
}

define { <vscale x 8 x half>, <vscale x 8 x half> } @f2cvtl(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: f2cvtl:
; CHECK:       // %bb.0:
; CHECK-NEXT:    f2cvtl { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.fp8.cvtl2.x2.nxvbf16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x half>, <vscale x 8 x half> } %res
}

; BF1CVTL / BF2CVTL

define { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @bf1cvtl(<vscale x 16 x i8> %zm) {
; CHECK-LABEL: bf1cvtl:
; CHECK:       // %bb.0:
; CHECK-NEXT:    bf1cvtl { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.fp8.cvtl1.x2.nxv8bf16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } %res
}

define { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @bf2cvtl( <vscale x 16 x i8> %zm) {
; CHECK-LABEL: bf2cvtl:
; CHECK:       // %bb.0:
; CHECK-NEXT:    bf2cvtl { z0.h, z1.h }, z0.b
; CHECK-NEXT:    ret
  %res = call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.fp8.cvtl2.x2.nxv8bf16(<vscale x 16 x i8> %zm)
  ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } %res
}