; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -march=ve -mattr=+vpu | FileCheck %s

declare <256 x i32> @llvm.vp.merge.v256i32(<256 x i1>, <256 x i32>, <256 x i32>, i32)

define fastcc <256 x i32> @test_vp_merge_v256i32_vv(<256 x i32> %i0, <256 x i32> %i1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256i32_vv:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s0, %s0, (32)0
; CHECK-NEXT:    lvl %s0
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %r0 = call <256 x i32> @llvm.vp.merge.v256i32(<256 x i1> %m, <256 x i32> %i0, <256 x i32> %i1, i32 %pivot)
  ret <256 x i32> %r0
}

define fastcc <256 x i32> @test_vp_merge_v256i32_vr(<256 x i32> %i0, i32 %s1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256i32_vr:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s1, %s1, (32)0
; CHECK-NEXT:    and %s0, %s0, (32)0
; CHECK-NEXT:    lea %s2, 256
; CHECK-NEXT:    lvl %s2
; CHECK-NEXT:    vbrd %v1, %s0
; CHECK-NEXT:    lvl %s1
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %xins = insertelement <256 x i32> undef, i32 %s1, i32 0
  %i1 = shufflevector <256 x i32> %xins, <256 x i32> undef, <256 x i32> zeroinitializer
  %r0 = call <256 x i32> @llvm.vp.merge.v256i32(<256 x i1> %m, <256 x i32> %i0, <256 x i32> %i1, i32 %pivot)
  ret <256 x i32> %r0
}

declare <256 x float> @llvm.vp.merge.v256f32(<256 x i1>, <256 x float>, <256 x float>, i32)

define fastcc <256 x float> @test_vp_merge_v256f32_vv(<256 x float> %i0, <256 x float> %i1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256f32_vv:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s0, %s0, (32)0
; CHECK-NEXT:    lvl %s0
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %r0 = call <256 x float> @llvm.vp.merge.v256f32(<256 x i1> %m, <256 x float> %i0, <256 x float> %i1, i32 %pivot)
  ret <256 x float> %r0
}

define fastcc <256 x float> @test_vp_merge_v256f32_vr(<256 x float> %i0, float %s1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256f32_vr:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s1, %s1, (32)0
; CHECK-NEXT:    lea %s2, 256
; CHECK-NEXT:    lvl %s2
; CHECK-NEXT:    vbrd %v1, %s0
; CHECK-NEXT:    lvl %s1
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %xins = insertelement <256 x float> undef, float %s1, i32 0
  %i1 = shufflevector <256 x float> %xins, <256 x float> undef, <256 x i32> zeroinitializer
  %r0 = call <256 x float> @llvm.vp.merge.v256f32(<256 x i1> %m, <256 x float> %i0, <256 x float> %i1, i32 %pivot)
  ret <256 x float> %r0
}

declare <256 x double> @llvm.vp.merge.v256f64(<256 x i1>, <256 x double>, <256 x double>, i32)

define fastcc <256 x double> @test_vp_merge_v256f64_vv(<256 x double> %i0, <256 x double> %i1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256f64_vv:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s0, %s0, (32)0
; CHECK-NEXT:    lvl %s0
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %r0 = call <256 x double> @llvm.vp.merge.v256f64(<256 x i1> %m, <256 x double> %i0, <256 x double> %i1, i32 %pivot)
  ret <256 x double> %r0
}

define fastcc <256 x double> @test_vp_merge_v256f64_vr(<256 x double> %i0, double %s1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256f64_vr:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s1, %s1, (32)0
; CHECK-NEXT:    lea %s2, 256
; CHECK-NEXT:    lvl %s2
; CHECK-NEXT:    vbrd %v1, %s0
; CHECK-NEXT:    lvl %s1
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %xins = insertelement <256 x double> undef, double %s1, i32 0
  %i1 = shufflevector <256 x double> %xins, <256 x double> undef, <256 x i32> zeroinitializer
  %r0 = call <256 x double> @llvm.vp.merge.v256f64(<256 x i1> %m, <256 x double> %i0, <256 x double> %i1, i32 %pivot)
  ret <256 x double> %r0
}

declare <256 x i64> @llvm.vp.merge.v256i64(<256 x i1>, <256 x i64>, <256 x i64>, i32)

define fastcc <256 x i64> @test_vp_merge_v256i64_vv(<256 x i64> %i0, <256 x i64> %i1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256i64_vv:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s0, %s0, (32)0
; CHECK-NEXT:    lvl %s0
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %r0 = call <256 x i64> @llvm.vp.merge.v256i64(<256 x i1> %m, <256 x i64> %i0, <256 x i64> %i1, i32 %pivot)
  ret <256 x i64> %r0
}

define fastcc <256 x i64> @test_vp_merge_v256i64_vr(<256 x i64> %i0, i64 %s1, <256 x i1> %m, i32 %pivot) {
; CHECK-LABEL: test_vp_merge_v256i64_vr:
; CHECK:       # %bb.0:
; CHECK-NEXT:    and %s1, %s1, (32)0
; CHECK-NEXT:    lea %s2, 256
; CHECK-NEXT:    lvl %s2
; CHECK-NEXT:    vbrd %v1, %s0
; CHECK-NEXT:    lvl %s1
; CHECK-NEXT:    vmrg %v1, %v1, %v0, %vm1
; CHECK-NEXT:    lea %s16, 256
; CHECK-NEXT:    lvl %s16
; CHECK-NEXT:    vor %v0, (0)1, %v1
; CHECK-NEXT:    b.l.t (, %s10)
  %xins = insertelement <256 x i64> undef, i64 %s1, i32 0
  %i1 = shufflevector <256 x i64> %xins, <256 x i64> undef, <256 x i32> zeroinitializer
  %r0 = call <256 x i64> @llvm.vp.merge.v256i64(<256 x i1> %m, <256 x i64> %i0, <256 x i64> %i1, i32 %pivot)
  ret <256 x i64> %r0
}