llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+d,+zvfh,+v -target-abi=ilp32d \
; RUN:   -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFH,RV32ZVFH
; RUN: llc -mtriple=riscv64 -mattr=+d,+zvfh,+v -target-abi=lp64d \
; RUN:   -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFH,RV64ZVFH
; RUN: llc -mtriple=riscv32 -mattr=+d,+zvfhmin,+v -target-abi=ilp32d \
; RUN:   -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFHMIN,RV32ZVFHMIN
; RUN: llc -mtriple=riscv64 -mattr=+d,+zvfhmin,+v -target-abi=lp64d \
; RUN:   -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZVFHMIN,RV64ZVFHMIN

define <2 x half> @vp_round_v2f16(<2 x half> %va, <2 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v2f16:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
; ZVFH-NEXT:    vfabs.v v9, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, mf4, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v2f16:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v9
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v9, v8, v9, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT:    ret
  %v = call <2 x half> @llvm.vp.round.v2f16(<2 x half> %va, <2 x i1> %m, i32 %evl)
  ret <2 x half> %v
}

define <2 x half> @vp_round_v2f16_unmasked(<2 x half> %va, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v2f16_unmasked:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
; ZVFH-NEXT:    vfabs.v v9, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, mf4, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v2f16_unmasked:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v9
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v9, v8, v9, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT:    ret
  %v = call <2 x half> @llvm.vp.round.v2f16(<2 x half> %va, <2 x i1> splat (i1 true), i32 %evl)
  ret <2 x half> %v
}

define <4 x half> @vp_round_v4f16(<4 x half> %va, <4 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v4f16:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
; ZVFH-NEXT:    vfabs.v v9, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v4f16:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v9
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m1, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v9, v8, v9, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT:    ret
  %v = call <4 x half> @llvm.vp.round.v4f16(<4 x half> %va, <4 x i1> %m, i32 %evl)
  ret <4 x half> %v
}

define <4 x half> @vp_round_v4f16_unmasked(<4 x half> %va, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v4f16_unmasked:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
; ZVFH-NEXT:    vfabs.v v9, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v4f16_unmasked:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v9
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m1, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v9, v8, v9, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT:    ret
  %v = call <4 x half> @llvm.vp.round.v4f16(<4 x half> %va, <4 x i1> splat (i1 true), i32 %evl)
  ret <4 x half> %v
}

define <8 x half> @vp_round_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v8f16:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
; ZVFH-NEXT:    vfabs.v v9, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, m1, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v8f16:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v10
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v10, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v10, v8, v10, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT:    ret
  %v = call <8 x half> @llvm.vp.round.v8f16(<8 x half> %va, <8 x i1> %m, i32 %evl)
  ret <8 x half> %v
}

define <8 x half> @vp_round_v8f16_unmasked(<8 x half> %va, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v8f16_unmasked:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
; ZVFH-NEXT:    vfabs.v v9, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, m1, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v8f16_unmasked:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v10
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v10, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v10, v8, v10, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT:    ret
  %v = call <8 x half> @llvm.vp.round.v8f16(<8 x half> %va, <8 x i1> splat (i1 true), i32 %evl)
  ret <8 x half> %v
}

define <16 x half> @vp_round_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v16f16:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
; ZVFH-NEXT:    vfabs.v v10, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v10, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v16f16:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v12
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v12, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v12, v8, v12, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT:    ret
  %v = call <16 x half> @llvm.vp.round.v16f16(<16 x half> %va, <16 x i1> %m, i32 %evl)
  ret <16 x half> %v
}

define <16 x half> @vp_round_v16f16_unmasked(<16 x half> %va, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v16f16_unmasked:
; ZVFH:       # %bb.0:
; ZVFH-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
; ZVFH-NEXT:    vfabs.v v10, v8
; ZVFH-NEXT:    li a0, 25
; ZVFH-NEXT:    slli a0, a0, 10
; ZVFH-NEXT:    fmv.h.x fa5, a0
; ZVFH-NEXT:    vmflt.vf v0, v10, fa5
; ZVFH-NEXT:    fsrmi a0, 4
; ZVFH-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; ZVFH-NEXT:    fsrm a0
; ZVFH-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; ZVFH-NEXT:    ret
;
; ZVFHMIN-LABEL: vp_round_v16f16_unmasked:
; ZVFHMIN:       # %bb.0:
; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT:    lui a0, 307200
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT:    vfabs.v v8, v12
; ZVFHMIN-NEXT:    fmv.w.x fa5, a0
; ZVFHMIN-NEXT:    vmflt.vf v0, v8, fa5
; ZVFHMIN-NEXT:    fsrmi a0, 4
; ZVFHMIN-NEXT:    vfcvt.x.f.v v8, v12, v0.t
; ZVFHMIN-NEXT:    fsrm a0
; ZVFHMIN-NEXT:    vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT:    vfsgnj.vv v12, v8, v12, v0.t
; ZVFHMIN-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT:    vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT:    ret
  %v = call <16 x half> @llvm.vp.round.v16f16(<16 x half> %va, <16 x i1> splat (i1 true), i32 %evl)
  ret <16 x half> %v
}

define <2 x float> @vp_round_v2f32(<2 x float> %va, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v2f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT:    vfabs.v v9, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v9, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <2 x float> @llvm.vp.round.v2f32(<2 x float> %va, <2 x i1> %m, i32 %evl)
  ret <2 x float> %v
}

define <2 x float> @vp_round_v2f32_unmasked(<2 x float> %va, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v2f32_unmasked:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT:    vfabs.v v9, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v9, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <2 x float> @llvm.vp.round.v2f32(<2 x float> %va, <2 x i1> splat (i1 true), i32 %evl)
  ret <2 x float> %v
}

define <4 x float> @vp_round_v4f32(<4 x float> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v4f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT:    vfabs.v v9, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v9, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <4 x float> @llvm.vp.round.v4f32(<4 x float> %va, <4 x i1> %m, i32 %evl)
  ret <4 x float> %v
}

define <4 x float> @vp_round_v4f32_unmasked(<4 x float> %va, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v4f32_unmasked:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT:    vfabs.v v9, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v9, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <4 x float> @llvm.vp.round.v4f32(<4 x float> %va, <4 x i1> splat (i1 true), i32 %evl)
  ret <4 x float> %v
}

define <8 x float> @vp_round_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v8f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
; CHECK-NEXT:    vfabs.v v10, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v10, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <8 x float> @llvm.vp.round.v8f32(<8 x float> %va, <8 x i1> %m, i32 %evl)
  ret <8 x float> %v
}

define <8 x float> @vp_round_v8f32_unmasked(<8 x float> %va, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v8f32_unmasked:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
; CHECK-NEXT:    vfabs.v v10, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v10, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <8 x float> @llvm.vp.round.v8f32(<8 x float> %va, <8 x i1> splat (i1 true), i32 %evl)
  ret <8 x float> %v
}

define <16 x float> @vp_round_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v16f32:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
; CHECK-NEXT:    vfabs.v v12, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v12, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <16 x float> @llvm.vp.round.v16f32(<16 x float> %va, <16 x i1> %m, i32 %evl)
  ret <16 x float> %v
}

define <16 x float> @vp_round_v16f32_unmasked(<16 x float> %va, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v16f32_unmasked:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
; CHECK-NEXT:    vfabs.v v12, v8
; CHECK-NEXT:    lui a0, 307200
; CHECK-NEXT:    fmv.w.x fa5, a0
; CHECK-NEXT:    vmflt.vf v0, v12, fa5
; CHECK-NEXT:    fsrmi a0, 4
; CHECK-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT:    fsrm a0
; CHECK-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT:    vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; CHECK-NEXT:    ret
  %v = call <16 x float> @llvm.vp.round.v16f32(<16 x float> %va, <16 x i1> splat (i1 true), i32 %evl)
  ret <16 x float> %v
}

define <2 x double> @vp_round_v2f64(<2 x double> %va, <2 x i1> %m, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v2f64:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI16_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI16_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v9, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v2f64:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v9, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v2f64:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI16_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI16_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v9, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v9, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v2f64:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v9, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v9, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <2 x double> @llvm.vp.round.v2f64(<2 x double> %va, <2 x i1> %m, i32 %evl)
  ret <2 x double> %v
}

define <2 x double> @vp_round_v2f64_unmasked(<2 x double> %va, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v2f64_unmasked:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI17_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI17_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v9, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v2f64_unmasked:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v9, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v9, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v2f64_unmasked:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI17_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI17_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v9, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v9, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v2f64_unmasked:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v9, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v9, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v9, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v9, v9, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m1, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v9, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <2 x double> @llvm.vp.round.v2f64(<2 x double> %va, <2 x i1> splat (i1 true), i32 %evl)
  ret <2 x double> %v
}

define <4 x double> @vp_round_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v4f64:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v10, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v10, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v4f64:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v10, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v10, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v4f64:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI18_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v10, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v10, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v4f64:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v10, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v10, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <4 x double> @llvm.vp.round.v4f64(<4 x double> %va, <4 x i1> %m, i32 %evl)
  ret <4 x double> %v
}

define <4 x double> @vp_round_v4f64_unmasked(<4 x double> %va, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v4f64_unmasked:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI19_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI19_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v10, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v10, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v4f64_unmasked:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v10, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v10, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v4f64_unmasked:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI19_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI19_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v10, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v10, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v4f64_unmasked:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v10, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v10, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v10, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v10, v10, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v10, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <4 x double> @llvm.vp.round.v4f64(<4 x double> %va, <4 x i1> splat (i1 true), i32 %evl)
  ret <4 x double> %v
}

define <8 x double> @vp_round_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v8f64:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v12, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v12, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v8f64:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v12, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v12, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v8f64:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI20_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v12, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v12, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v8f64:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v12, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v12, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <8 x double> @llvm.vp.round.v8f64(<8 x double> %va, <8 x i1> %m, i32 %evl)
  ret <8 x double> %v
}

define <8 x double> @vp_round_v8f64_unmasked(<8 x double> %va, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v8f64_unmasked:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI21_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI21_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v12, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v12, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v8f64_unmasked:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v12, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v12, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v8f64_unmasked:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI21_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI21_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v12, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v12, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v8f64_unmasked:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v12, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v12, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v12, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v12, v12, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v12, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <8 x double> @llvm.vp.round.v8f64(<8 x double> %va, <8 x i1> splat (i1 true), i32 %evl)
  ret <8 x double> %v
}

define <15 x double> @vp_round_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v15f64:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v16, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v15f64:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v16, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v15f64:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI22_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v15f64:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <15 x double> @llvm.vp.round.v15f64(<15 x double> %va, <15 x i1> %m, i32 %evl)
  ret <15 x double> %v
}

define <15 x double> @vp_round_v15f64_unmasked(<15 x double> %va, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v15f64_unmasked:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI23_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI23_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v16, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v15f64_unmasked:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v16, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v15f64_unmasked:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI23_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI23_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v15f64_unmasked:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <15 x double> @llvm.vp.round.v15f64(<15 x double> %va, <15 x i1> splat (i1 true), i32 %evl)
  ret <15 x double> %v
}

define <16 x double> @vp_round_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v16f64:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI24_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v16, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v16f64:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v16, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v16f64:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI24_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v16f64:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <16 x double> @llvm.vp.round.v16f64(<16 x double> %va, <16 x i1> %m, i32 %evl)
  ret <16 x double> %v
}

define <16 x double> @vp_round_v16f64_unmasked(<16 x double> %va, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v16f64_unmasked:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI25_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI25_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v16, v8
; RV32ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v16f64_unmasked:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v16, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v16f64_unmasked:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI25_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI25_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v16f64_unmasked:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v16, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v16, fa5
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v16, v16, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v16, v8, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <16 x double> @llvm.vp.round.v16f64(<16 x double> %va, <16 x i1> splat (i1 true), i32 %evl)
  ret <16 x double> %v
}

define <32 x double> @vp_round_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v32f64:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI26_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI26_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v24, v8
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vmflt.vf v0, v24, fa5
; RV32ZVFH-NEXT:    vfabs.v v24, v16
; RV32ZVFH-NEXT:    vmflt.vf v7, v24, fa5
; RV32ZVFH-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV32ZVFH-NEXT:    vmv1r.v v0, v7
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v32f64:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v24, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vmflt.vf v0, v24, fa5
; RV64ZVFH-NEXT:    vfabs.v v24, v16
; RV64ZVFH-NEXT:    vmflt.vf v7, v24, fa5
; RV64ZVFH-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV64ZVFH-NEXT:    vmv1r.v v0, v7
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v32f64:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI26_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI26_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v24, v8
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v24, fa5
; RV32ZVFHMIN-NEXT:    vfabs.v v24, v16
; RV32ZVFHMIN-NEXT:    vmflt.vf v7, v24, fa5
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV32ZVFHMIN-NEXT:    vmv1r.v v0, v7
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v32f64:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v24, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v24, fa5
; RV64ZVFHMIN-NEXT:    vfabs.v v24, v16
; RV64ZVFHMIN-NEXT:    vmflt.vf v7, v24, fa5
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV64ZVFHMIN-NEXT:    vmv1r.v v0, v7
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <32 x double> @llvm.vp.round.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
  ret <32 x double> %v
}

define <32 x double> @vp_round_v32f64_unmasked(<32 x double> %va, i32 zeroext %evl) {
; RV32ZVFH-LABEL: vp_round_v32f64_unmasked:
; RV32ZVFH:       # %bb.0:
; RV32ZVFH-NEXT:    lui a0, %hi(.LCPI27_0)
; RV32ZVFH-NEXT:    fld fa5, %lo(.LCPI27_0)(a0)
; RV32ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfabs.v v24, v8
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vmflt.vf v0, v24, fa5
; RV32ZVFH-NEXT:    vfabs.v v24, v16
; RV32ZVFH-NEXT:    vmflt.vf v7, v24, fa5
; RV32ZVFH-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFH-NEXT:    fsrmi a0, 4
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV32ZVFH-NEXT:    vmv1r.v v0, v7
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV32ZVFH-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV32ZVFH-NEXT:    fsrm a0
; RV32ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFH-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV32ZVFH-NEXT:    ret
;
; RV64ZVFH-LABEL: vp_round_v32f64_unmasked:
; RV64ZVFH:       # %bb.0:
; RV64ZVFH-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfabs.v v24, v8
; RV64ZVFH-NEXT:    li a0, 1075
; RV64ZVFH-NEXT:    slli a0, a0, 52
; RV64ZVFH-NEXT:    fmv.d.x fa5, a0
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vmflt.vf v0, v24, fa5
; RV64ZVFH-NEXT:    vfabs.v v24, v16
; RV64ZVFH-NEXT:    vmflt.vf v7, v24, fa5
; RV64ZVFH-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFH-NEXT:    fsrmi a0, 4
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV64ZVFH-NEXT:    vmv1r.v v0, v7
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV64ZVFH-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV64ZVFH-NEXT:    fsrm a0
; RV64ZVFH-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFH-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFH-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV64ZVFH-NEXT:    ret
;
; RV32ZVFHMIN-LABEL: vp_round_v32f64_unmasked:
; RV32ZVFHMIN:       # %bb.0:
; RV32ZVFHMIN-NEXT:    lui a0, %hi(.LCPI27_0)
; RV32ZVFHMIN-NEXT:    fld fa5, %lo(.LCPI27_0)(a0)
; RV32ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfabs.v v24, v8
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vmflt.vf v0, v24, fa5
; RV32ZVFHMIN-NEXT:    vfabs.v v24, v16
; RV32ZVFHMIN-NEXT:    vmflt.vf v7, v24, fa5
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFHMIN-NEXT:    fsrmi a0, 4
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV32ZVFHMIN-NEXT:    vmv1r.v v0, v7
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV32ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV32ZVFHMIN-NEXT:    fsrm a0
; RV32ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV32ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV32ZVFHMIN-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV32ZVFHMIN-NEXT:    ret
;
; RV64ZVFHMIN-LABEL: vp_round_v32f64_unmasked:
; RV64ZVFHMIN:       # %bb.0:
; RV64ZVFHMIN-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfabs.v v24, v8
; RV64ZVFHMIN-NEXT:    li a0, 1075
; RV64ZVFHMIN-NEXT:    slli a0, a0, 52
; RV64ZVFHMIN-NEXT:    fmv.d.x fa5, a0
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vmflt.vf v0, v24, fa5
; RV64ZVFHMIN-NEXT:    vfabs.v v24, v16
; RV64ZVFHMIN-NEXT:    vmflt.vf v7, v24, fa5
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v8, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFHMIN-NEXT:    fsrmi a0, 4
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v8, v24, v8, v0.t
; RV64ZVFHMIN-NEXT:    vmv1r.v v0, v7
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
; RV64ZVFHMIN-NEXT:    vfcvt.x.f.v v24, v16, v0.t
; RV64ZVFHMIN-NEXT:    fsrm a0
; RV64ZVFHMIN-NEXT:    vfcvt.f.x.v v24, v24, v0.t
; RV64ZVFHMIN-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
; RV64ZVFHMIN-NEXT:    vfsgnj.vv v16, v24, v16, v0.t
; RV64ZVFHMIN-NEXT:    ret
  %v = call <32 x double> @llvm.vp.round.v32f64(<32 x double> %va, <32 x i1> splat (i1 true), i32 %evl)
  ret <32 x double> %v
}