; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=128 -verify-machineinstrs < %s \
; RUN:   | FileCheck %s
; RUN: llc -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=128 -verify-machineinstrs < %s \
; RUN:   | FileCheck %s

declare i1 @llvm.vp.reduce.and.v1i1(i1, <1 x i1>, <1 x i1>, i32)

define signext i1 @vpreduce_and_v1i1(i1 signext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v1i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT:    vmnot.m v9, v0
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v1i1(i1 %s, <1 x i1> %v, <1 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.or.v1i1(i1, <1 x i1>, <1 x i1>, i32)

define signext i1 @vpreduce_or_v1i1(i1 signext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_or_v1i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    snez a1, a1
; CHECK-NEXT:    or a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.or.v1i1(i1 %s, <1 x i1> %v, <1 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.xor.v1i1(i1, <1 x i1>, <1 x i1>, i32)

define signext i1 @vpreduce_xor_v1i1(i1 signext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_xor_v1i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    xor a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.xor.v1i1(i1 %s, <1 x i1> %v, <1 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.and.v2i1(i1, <2 x i1>, <2 x i1>, i32)

define signext i1 @vpreduce_and_v2i1(i1 signext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v2i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT:    vmnot.m v9, v0
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v2i1(i1 %s, <2 x i1> %v, <2 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.or.v2i1(i1, <2 x i1>, <2 x i1>, i32)

define signext i1 @vpreduce_or_v2i1(i1 signext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_or_v2i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    snez a1, a1
; CHECK-NEXT:    or a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.or.v2i1(i1 %s, <2 x i1> %v, <2 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.xor.v2i1(i1, <2 x i1>, <2 x i1>, i32)

define signext i1 @vpreduce_xor_v2i1(i1 signext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_xor_v2i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    xor a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.xor.v2i1(i1 %s, <2 x i1> %v, <2 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.and.v4i1(i1, <4 x i1>, <4 x i1>, i32)

define signext i1 @vpreduce_and_v4i1(i1 signext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v4i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT:    vmnot.m v9, v0
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v4i1(i1 %s, <4 x i1> %v, <4 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.or.v4i1(i1, <4 x i1>, <4 x i1>, i32)

define signext i1 @vpreduce_or_v4i1(i1 signext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_or_v4i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    snez a1, a1
; CHECK-NEXT:    or a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.or.v4i1(i1 %s, <4 x i1> %v, <4 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.xor.v4i1(i1, <4 x i1>, <4 x i1>, i32)

define signext i1 @vpreduce_xor_v4i1(i1 signext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_xor_v4i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    xor a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.xor.v4i1(i1 %s, <4 x i1> %v, <4 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.and.v8i1(i1, <8 x i1>, <8 x i1>, i32)

define signext i1 @vpreduce_and_v8i1(i1 signext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v8i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT:    vmnot.m v9, v0
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v8i1(i1 %s, <8 x i1> %v, <8 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.or.v8i1(i1, <8 x i1>, <8 x i1>, i32)

define signext i1 @vpreduce_or_v8i1(i1 signext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_or_v8i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    snez a1, a1
; CHECK-NEXT:    or a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.or.v8i1(i1 %s, <8 x i1> %v, <8 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.xor.v8i1(i1, <8 x i1>, <8 x i1>, i32)

define signext i1 @vpreduce_xor_v8i1(i1 signext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_xor_v8i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    xor a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.xor.v8i1(i1 %s, <8 x i1> %v, <8 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.and.v10i1(i1, <10 x i1>, <10 x i1>, i32)

define signext i1 @vpreduce_and_v10i1(i1 signext %s, <10 x i1> %v, <10 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v10i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT:    vmnot.m v9, v0
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v10i1(i1 %s, <10 x i1> %v, <10 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.and.v16i1(i1, <16 x i1>, <16 x i1>, i32)

define signext i1 @vpreduce_and_v16i1(i1 signext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v16i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT:    vmnot.m v9, v0
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v16i1(i1 %s, <16 x i1> %v, <16 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.and.v256i1(i1, <256 x i1>, <256 x i1>, i32)

define signext i1 @vpreduce_and_v256i1(i1 signext %s, <256 x i1> %v, <256 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_and_v256i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    addi a2, a1, -128
; CHECK-NEXT:    vmv1r.v v11, v0
; CHECK-NEXT:    li a3, 0
; CHECK-NEXT:    bltu a1, a2, .LBB14_2
; CHECK-NEXT:  # %bb.1:
; CHECK-NEXT:    mv a3, a2
; CHECK-NEXT:  .LBB14_2:
; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, mu
; CHECK-NEXT:    vmnot.m v8, v8
; CHECK-NEXT:    vmv1r.v v0, v10
; CHECK-NEXT:    vcpop.m a2, v8, v0.t
; CHECK-NEXT:    li a3, 128
; CHECK-NEXT:    seqz a2, a2
; CHECK-NEXT:    bltu a1, a3, .LBB14_4
; CHECK-NEXT:  # %bb.3:
; CHECK-NEXT:    li a1, 128
; CHECK-NEXT:  .LBB14_4:
; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, mu
; CHECK-NEXT:    vmnot.m v8, v11
; CHECK-NEXT:    vmv1r.v v0, v9
; CHECK-NEXT:    vcpop.m a1, v8, v0.t
; CHECK-NEXT:    seqz a1, a1
; CHECK-NEXT:    and a0, a1, a0
; CHECK-NEXT:    and a0, a2, a0
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.and.v256i1(i1 %s, <256 x i1> %v, <256 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.or.v16i1(i1, <16 x i1>, <16 x i1>, i32)

define signext i1 @vpreduce_or_v16i1(i1 signext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_or_v16i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    snez a1, a1
; CHECK-NEXT:    or a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.or.v16i1(i1 %s, <16 x i1> %v, <16 x i1> %m, i32 %evl)
  ret i1 %r
}

declare i1 @llvm.vp.reduce.xor.v16i1(i1, <16 x i1>, <16 x i1>, i32)

define signext i1 @vpreduce_xor_v16i1(i1 signext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_xor_v16i1:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmv1r.v v9, v0
; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT:    vmv1r.v v0, v8
; CHECK-NEXT:    vcpop.m a1, v9, v0.t
; CHECK-NEXT:    xor a0, a1, a0
; CHECK-NEXT:    andi a0, a0, 1
; CHECK-NEXT:    neg a0, a0
; CHECK-NEXT:    ret
  %r = call i1 @llvm.vp.reduce.xor.v16i1(i1 %s, <16 x i1> %v, <16 x i1> %m, i32 %evl)
  ret i1 %r
}