Compiler projects using llvm
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+d,+zfh,+experimental-zvfh,+v -target-abi=ilp32d -riscv-v-vector-bits-min=128 \
; RUN:   -verify-machineinstrs < %s | FileCheck %s
; RUN: llc -mtriple=riscv64 -mattr=+d,+zfh,+experimental-zvfh,+v -target-abi=lp64d -riscv-v-vector-bits-min=128 \
; RUN:   -verify-machineinstrs < %s | FileCheck %s

define <2 x half> @select_v2f16(i1 zeroext %c, <2 x half> %a, <2 x half> %b) {
; CHECK-LABEL: select_v2f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <2 x half> %a, <2 x half> %b
  ret <2 x half> %v
}

define <2 x half> @selectcc_v2f16(half %a, half %b, <2 x half> %c, <2 x half> %d) {
; CHECK-LABEL: selectcc_v2f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.h a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq half %a, %b
  %v = select i1 %cmp, <2 x half> %c, <2 x half> %d
  ret <2 x half> %v
}

define <4 x half> @select_v4f16(i1 zeroext %c, <4 x half> %a, <4 x half> %b) {
; CHECK-LABEL: select_v4f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <4 x half> %a, <4 x half> %b
  ret <4 x half> %v
}

define <4 x half> @selectcc_v4f16(half %a, half %b, <4 x half> %c, <4 x half> %d) {
; CHECK-LABEL: selectcc_v4f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.h a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq half %a, %b
  %v = select i1 %cmp, <4 x half> %c, <4 x half> %d
  ret <4 x half> %v
}

define <8 x half> @select_v8f16(i1 zeroext %c, <8 x half> %a, <8 x half> %b) {
; CHECK-LABEL: select_v8f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <8 x half> %a, <8 x half> %b
  ret <8 x half> %v
}

define <8 x half> @selectcc_v8f16(half %a, half %b, <8 x half> %c, <8 x half> %d) {
; CHECK-LABEL: selectcc_v8f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.h a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq half %a, %b
  %v = select i1 %cmp, <8 x half> %c, <8 x half> %d
  ret <8 x half> %v
}

define <16 x half> @select_v16f16(i1 zeroext %c, <16 x half> %a, <16 x half> %b) {
; CHECK-LABEL: select_v16f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, mu
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <16 x half> %a, <16 x half> %b
  ret <16 x half> %v
}

define <16 x half> @selectcc_v16f16(half %a, half %b, <16 x half> %c, <16 x half> %d) {
; CHECK-LABEL: selectcc_v16f16:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.h a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, mu
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq half %a, %b
  %v = select i1 %cmp, <16 x half> %c, <16 x half> %d
  ret <16 x half> %v
}

define <2 x float> @select_v2f32(i1 zeroext %c, <2 x float> %a, <2 x float> %b) {
; CHECK-LABEL: select_v2f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <2 x float> %a, <2 x float> %b
  ret <2 x float> %v
}

define <2 x float> @selectcc_v2f32(float %a, float %b, <2 x float> %c, <2 x float> %d) {
; CHECK-LABEL: selectcc_v2f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.s a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq float %a, %b
  %v = select i1 %cmp, <2 x float> %c, <2 x float> %d
  ret <2 x float> %v
}

define <4 x float> @select_v4f32(i1 zeroext %c, <4 x float> %a, <4 x float> %b) {
; CHECK-LABEL: select_v4f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <4 x float> %a, <4 x float> %b
  ret <4 x float> %v
}

define <4 x float> @selectcc_v4f32(float %a, float %b, <4 x float> %c, <4 x float> %d) {
; CHECK-LABEL: selectcc_v4f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.s a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq float %a, %b
  %v = select i1 %cmp, <4 x float> %c, <4 x float> %d
  ret <4 x float> %v
}

define <8 x float> @select_v8f32(i1 zeroext %c, <8 x float> %a, <8 x float> %b) {
; CHECK-LABEL: select_v8f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <8 x float> %a, <8 x float> %b
  ret <8 x float> %v
}

define <8 x float> @selectcc_v8f32(float %a, float %b, <8 x float> %c, <8 x float> %d) {
; CHECK-LABEL: selectcc_v8f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.s a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq float %a, %b
  %v = select i1 %cmp, <8 x float> %c, <8 x float> %d
  ret <8 x float> %v
}

define <16 x float> @select_v16f32(i1 zeroext %c, <16 x float> %a, <16 x float> %b) {
; CHECK-LABEL: select_v16f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, mu
; CHECK-NEXT:    vmv.v.x v16, a0
; CHECK-NEXT:    vmsne.vi v0, v16, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <16 x float> %a, <16 x float> %b
  ret <16 x float> %v
}

define <16 x float> @selectcc_v16f32(float %a, float %b, <16 x float> %c, <16 x float> %d) {
; CHECK-LABEL: selectcc_v16f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.s a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, mu
; CHECK-NEXT:    vmv.v.x v16, a0
; CHECK-NEXT:    vmsne.vi v0, v16, 0
; CHECK-NEXT:    vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq float %a, %b
  %v = select i1 %cmp, <16 x float> %c, <16 x float> %d
  ret <16 x float> %v
}

define <2 x double> @select_v2f64(i1 zeroext %c, <2 x double> %a, <2 x double> %b) {
; CHECK-LABEL: select_v2f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <2 x double> %a, <2 x double> %b
  ret <2 x double> %v
}

define <2 x double> @selectcc_v2f64(double %a, double %b, <2 x double> %c, <2 x double> %d) {
; CHECK-LABEL: selectcc_v2f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.d a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, mu
; CHECK-NEXT:    vmv.v.x v10, a0
; CHECK-NEXT:    vmsne.vi v0, v10, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v9, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq double %a, %b
  %v = select i1 %cmp, <2 x double> %c, <2 x double> %d
  ret <2 x double> %v
}

define <4 x double> @select_v4f64(i1 zeroext %c, <4 x double> %a, <4 x double> %b) {
; CHECK-LABEL: select_v4f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <4 x double> %a, <4 x double> %b
  ret <4 x double> %v
}

define <4 x double> @selectcc_v4f64(double %a, double %b, <4 x double> %c, <4 x double> %d) {
; CHECK-LABEL: selectcc_v4f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.d a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT:    vmv.v.x v12, a0
; CHECK-NEXT:    vmsne.vi v0, v12, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v10, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq double %a, %b
  %v = select i1 %cmp, <4 x double> %c, <4 x double> %d
  ret <4 x double> %v
}

define <8 x double> @select_v8f64(i1 zeroext %c, <8 x double> %a, <8 x double> %b) {
; CHECK-LABEL: select_v8f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT:    vmv.v.x v16, a0
; CHECK-NEXT:    vmsne.vi v0, v16, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <8 x double> %a, <8 x double> %b
  ret <8 x double> %v
}

define <8 x double> @selectcc_v8f64(double %a, double %b, <8 x double> %c, <8 x double> %d) {
; CHECK-LABEL: selectcc_v8f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.d a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT:    vmv.v.x v16, a0
; CHECK-NEXT:    vmsne.vi v0, v16, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v12, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq double %a, %b
  %v = select i1 %cmp, <8 x double> %c, <8 x double> %d
  ret <8 x double> %v
}

define <16 x double> @select_v16f64(i1 zeroext %c, <16 x double> %a, <16 x double> %b) {
; CHECK-LABEL: select_v16f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, mu
; CHECK-NEXT:    vmv.v.x v24, a0
; CHECK-NEXT:    vmsne.vi v0, v24, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT:    ret
  %v = select i1 %c, <16 x double> %a, <16 x double> %b
  ret <16 x double> %v
}

define <16 x double> @selectcc_v16f64(double %a, double %b, <16 x double> %c, <16 x double> %d) {
; CHECK-LABEL: selectcc_v16f64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    feq.d a0, fa0, fa1
; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, mu
; CHECK-NEXT:    vmv.v.x v24, a0
; CHECK-NEXT:    vmsne.vi v0, v24, 0
; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT:    vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT:    ret
  %cmp = fcmp oeq double %a, %b
  %v = select i1 %cmp, <16 x double> %c, <16 x double> %d
  ret <16 x double> %v
}