llvm/test/CodeGen/RISCV/rvv/select-bf16.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+d,+v,+zfbfmin,+zvfbfmin -target-abi=ilp32d \
; RUN:     -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFBFMIN
; RUN: llc -mtriple=riscv64 -mattr=+d,+v,+zfbfmin,+zvfbfmin -target-abi=lp64d \
; RUN:     -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFBFMIN
; RUN: llc -mtriple=riscv32 -mattr=+d,+v,+zfbfmin,+experimental-zvfbfa \
; RUN:     -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFBFA
; RUN: llc -mtriple=riscv64 -mattr=+d,+v,+zfbfmin,+experimental-zvfbfa \
; RUN:     -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFBFA

define <vscale x 1 x bfloat> @select_nxv1bf16(i1 zeroext %c, <vscale x 1 x bfloat> %a, <vscale x 1 x bfloat> %b) {
; CHECK-LABEL: select_nxv1bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli a1, zero, e8, mf8, ta, ma
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <vscale x 1 x bfloat> %a, <vscale x 1 x bfloat> %b
  ret <vscale x 1 x bfloat> %v
}

define <vscale x 1 x bfloat> @selectcc_nxv1bf16(bfloat %a, bfloat %b, <vscale x 1 x bfloat> %c, <vscale x 1 x bfloat> %d) {
; CHECK-LABEL: selectcc_nxv1bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    fcvt.s.bf16 fa5, fa1
; CHECK-NEXT:    fcvt.s.bf16 fa4, fa0
; CHECK-NEXT:    feq.s a0, fa4, fa5
; CHECK-NEXT:    vsetvli a1, zero, e8, mf8, ta, ma
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq bfloat %a, %b
  %v = select i1 %cmp, <vscale x 1 x bfloat> %c, <vscale x 1 x bfloat> %d
  ret <vscale x 1 x bfloat> %v
}

define <vscale x 2 x bfloat> @select_nxv2bf16(i1 zeroext %c, <vscale x 2 x bfloat> %a, <vscale x 2 x bfloat> %b) {
; CHECK-LABEL: select_nxv2bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <vscale x 2 x bfloat> %a, <vscale x 2 x bfloat> %b
  ret <vscale x 2 x bfloat> %v
}

define <vscale x 2 x bfloat> @selectcc_nxv2bf16(bfloat %a, bfloat %b, <vscale x 2 x bfloat> %c, <vscale x 2 x bfloat> %d) {
; CHECK-LABEL: selectcc_nxv2bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    fcvt.s.bf16 fa5, fa1
; CHECK-NEXT:    fcvt.s.bf16 fa4, fa0
; CHECK-NEXT:    feq.s a0, fa4, fa5
; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq bfloat %a, %b
  %v = select i1 %cmp, <vscale x 2 x bfloat> %c, <vscale x 2 x bfloat> %d
  ret <vscale x 2 x bfloat> %v
}

define <vscale x 4 x bfloat> @select_nxv4bf16(i1 zeroext %c, <vscale x 4 x bfloat> %a, <vscale x 4 x bfloat> %b) {
; CHECK-LABEL: select_nxv4bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <vscale x 4 x bfloat> %a, <vscale x 4 x bfloat> %b
  ret <vscale x 4 x bfloat> %v
}

define <vscale x 4 x bfloat> @selectcc_nxv4bf16(bfloat %a, bfloat %b, <vscale x 4 x bfloat> %c, <vscale x 4 x bfloat> %d) {
; CHECK-LABEL: selectcc_nxv4bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    fcvt.s.bf16 fa5, fa1
; CHECK-NEXT:    fcvt.s.bf16 fa4, fa0
; CHECK-NEXT:    feq.s a0, fa4, fa5
; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq bfloat %a, %b
  %v = select i1 %cmp, <vscale x 4 x bfloat> %c, <vscale x 4 x bfloat> %d
  ret <vscale x 4 x bfloat> %v
}

define <vscale x 8 x bfloat> @select_nxv8bf16(i1 zeroext %c, <vscale x 8 x bfloat> %a, <vscale x 8 x bfloat> %b) {
; CHECK-LABEL: select_nxv8bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <vscale x 8 x bfloat> %a, <vscale x 8 x bfloat> %b
  ret <vscale x 8 x bfloat> %v
}

define <vscale x 8 x bfloat> @selectcc_nxv8bf16(bfloat %a, bfloat %b, <vscale x 8 x bfloat> %c, <vscale x 8 x bfloat> %d) {
; CHECK-LABEL: selectcc_nxv8bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    fcvt.s.bf16 fa5, fa1
; CHECK-NEXT:    fcvt.s.bf16 fa4, fa0
; CHECK-NEXT:    feq.s a0, fa4, fa5
; CHECK-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq bfloat %a, %b
  %v = select i1 %cmp, <vscale x 8 x bfloat> %c, <vscale x 8 x bfloat> %d
  ret <vscale x 8 x bfloat> %v
}

define <vscale x 16 x bfloat> @select_nxv16bf16(i1 zeroext %c, <vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b) {
; CHECK-LABEL: select_nxv16bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
; CHECK-NEXT:    vmv.v.x v16, a0
; CHECK-NEXT:    vmsne.vi v0, v16, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b
  ret <vscale x 16 x bfloat> %v
}

define <vscale x 16 x bfloat> @selectcc_nxv16bf16(bfloat %a, bfloat %b, <vscale x 16 x bfloat> %c, <vscale x 16 x bfloat> %d) {
; CHECK-LABEL: selectcc_nxv16bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    fcvt.s.bf16 fa5, fa1
; CHECK-NEXT:    fcvt.s.bf16 fa4, fa0
; CHECK-NEXT:    feq.s a0, fa4, fa5
; CHECK-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
; CHECK-NEXT:    vmv.v.x v16, a0
; CHECK-NEXT:    vmsne.vi v0, v16, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq bfloat %a, %b
  %v = select i1 %cmp, <vscale x 16 x bfloat> %c, <vscale x 16 x bfloat> %d
  ret <vscale x 16 x bfloat> %v
}

define <vscale x 32 x bfloat> @select_nxv32bf16(i1 zeroext %c, <vscale x 32 x bfloat> %a, <vscale x 32 x bfloat> %b) {
; CHECK-LABEL: select_nxv32bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
; CHECK-NEXT:    vmv.v.x v24, a0
; CHECK-NEXT:    vmsne.vi v0, v24, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <vscale x 32 x bfloat> %a, <vscale x 32 x bfloat> %b
  ret <vscale x 32 x bfloat> %v
}

define <vscale x 32 x bfloat> @selectcc_nxv32bf16(bfloat %a, bfloat %b, <vscale x 32 x bfloat> %c, <vscale x 32 x bfloat> %d) {
; CHECK-LABEL: selectcc_nxv32bf16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    fcvt.s.bf16 fa5, fa1
; CHECK-NEXT:    fcvt.s.bf16 fa4, fa0
; CHECK-NEXT:    feq.s a0, fa4, fa5
; CHECK-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
; CHECK-NEXT:    vmv.v.x v24, a0
; CHECK-NEXT:    vmsne.vi v0, v24, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT:    vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq bfloat %a, %b
  %v = select i1 %cmp, <vscale x 32 x bfloat> %c, <vscale x 32 x bfloat> %d
  ret <vscale x 32 x bfloat> %v
}

define <vscale x 4 x bfloat> @vselect_vf_nxv4bf16(bfloat %a, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %cc) {
; ZVFBFMIN-LABEL: vselect_vf_nxv4bf16:
; ZVFBFMIN:       # %bb.0:
; ZVFBFMIN-NEXT:    fmv.x.h a0, fa0
; ZVFBFMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
; ZVFBFMIN-NEXT:    vmv.v.x v9, a0
; ZVFBFMIN-NEXT:    vmerge.vvm v8, v8, v9, v0
; ZVFBFMIN-NEXT:    ret
;
; ZVFBFA-LABEL: vselect_vf_nxv4bf16:
; ZVFBFA:       # %bb.0:
; ZVFBFA-NEXT:    vsetvli a0, zero, e16alt, m1, ta, ma
; ZVFBFA-NEXT:    vfmerge.vfm v8, v8, fa0, v0
; ZVFBFA-NEXT:    ret
  %insert = insertelement <vscale x 4 x bfloat> poison, bfloat %a, i32 0
  %va = shufflevector <vscale x 4 x bfloat> %insert, <vscale x 4 x bfloat> poison, <vscale x 4 x i32> zeroinitializer
  %vsel = select <vscale x 4 x i1> %cc, <vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb
  ret <vscale x 4 x bfloat> %vsel
}

define <vscale x 8 x bfloat> @vselect_vf_nxv8bf16(bfloat %a, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %cc) {
; ZVFBFMIN-LABEL: vselect_vf_nxv8bf16:
; ZVFBFMIN:       # %bb.0:
; ZVFBFMIN-NEXT:    fmv.x.h a0, fa0
; ZVFBFMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
; ZVFBFMIN-NEXT:    vmv.v.x v10, a0
; ZVFBFMIN-NEXT:    vmerge.vvm v8, v8, v10, v0
; ZVFBFMIN-NEXT:    ret
;
; ZVFBFA-LABEL: vselect_vf_nxv8bf16:
; ZVFBFA:       # %bb.0:
; ZVFBFA-NEXT:    vsetvli a0, zero, e16alt, m2, ta, ma
; ZVFBFA-NEXT:    vfmerge.vfm v8, v8, fa0, v0
; ZVFBFA-NEXT:    ret
  %insert = insertelement <vscale x 8 x bfloat> poison, bfloat %a, i32 0
  %va = shufflevector <vscale x 8 x bfloat> %insert, <vscale x 8 x bfloat> poison, <vscale x 8 x i32> zeroinitializer
  %vsel = select <vscale x 8 x i1> %cc, <vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb
  ret <vscale x 8 x bfloat> %vsel
}

define <vscale x 4 x bfloat> @vselect_vfpzero_nxv4bf16(<vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %cc) {
; ZVFBFMIN-LABEL: vselect_vfpzero_nxv4bf16:
; ZVFBFMIN:       # %bb.0:
; ZVFBFMIN-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
; ZVFBFMIN-NEXT:    vmv.v.i v9, 0
; ZVFBFMIN-NEXT:    vmerge.vvm v8, v8, v9, v0
; ZVFBFMIN-NEXT:    ret
;
; ZVFBFA-LABEL: vselect_vfpzero_nxv4bf16:
; ZVFBFA:       # %bb.0:
; ZVFBFA-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
; ZVFBFA-NEXT:    vmerge.vim v8, v8, 0, v0
; ZVFBFA-NEXT:    ret
  %vsel = select <vscale x 4 x i1> %cc, <vscale x 4 x bfloat> splat (bfloat 0.0), <vscale x 4 x bfloat> %vb
  ret <vscale x 4 x bfloat> %vsel
}