[llvm] dbc681c - [VP][RISCV] Add vp.roundtozero and its RISC-V support
via llvm-commits
llvm-commits at lists.llvm.org
Fri Oct 7 02:15:32 PDT 2022
Author: eopXD
Date: 2022-10-07T02:15:23-07:00
New Revision: dbc681c98e14d495618ec42975e9f793d3a7a0b7
URL: https://github.com/llvm/llvm-project/commit/dbc681c98e14d495618ec42975e9f793d3a7a0b7
DIFF: https://github.com/llvm/llvm-project/commit/dbc681c98e14d495618ec42975e9f793d3a7a0b7.diff
LOG: [VP][RISCV] Add vp.roundtozero and its RISC-V support
The scalar instruction of this is `llvm.trunc`. However the naming of
ISD::VP_TRUNC is already taken by `trunc` of the LLVM IR. Naming this as
`vp.ftrunc` would likely cause confusion with `vp.fptrunc`. So adding
`vp.roundtozero` that will look similar to `vp.roundeven`.
Reviewed By: frasercrmck
Differential Revision: https://reviews.llvm.org/D135233
Added:
llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
Modified:
llvm/docs/LangRef.rst
llvm/include/llvm/IR/Intrinsics.td
llvm/include/llvm/IR/VPIntrinsics.def
llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
llvm/lib/Target/RISCV/RISCVISelLowering.cpp
llvm/unittests/IR/VPIntrinsicTest.cpp
Removed:
################################################################################
diff --git a/llvm/docs/LangRef.rst b/llvm/docs/LangRef.rst
index 4fe65ffe6db36..3f1930c0ba12d 100644
--- a/llvm/docs/LangRef.rst
+++ b/llvm/docs/LangRef.rst
@@ -14806,6 +14806,9 @@ Semantics:
This function returns the same values as the libm ``ceil`` functions
would, and handles error conditions in the same way.
+
+.. _int_llvm_trunc:
+
'``llvm.trunc.*``' Intrinsic
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -21664,6 +21667,53 @@ Examples:
%t = call <4 x float> @llvm.roundeven.v4f32(<4 x float> %a)
%also.r = select <4 x i1> %mask, <4 x float> %t, <4 x float> poison
+.. _int_vp_roundtozero:
+
+'``llvm.vp.roundtozero.*``' Intrinsics
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Syntax:
+"""""""
+This is an overloaded intrinsic.
+
+::
+
+ declare <16 x float> @llvm.vp.roundtozero.v16f32 (<16 x float> <op>, <16 x i1> <mask>, i32 <vector_length>)
+ declare <vscale x 4 x float> @llvm.vp.roundtozero.nxv4f32 (<vscale x 4 x float> <op>, <vscale x 4 x i1> <mask>, i32 <vector_length>)
+ declare <256 x double> @llvm.vp.roundtozero.v256f64 (<256 x double> <op>, <256 x i1> <mask>, i32 <vector_length>)
+
+Overview:
+"""""""""
+
+Predicated floating-point round-to-zero of a vector of floating-point values.
+
+
+Arguments:
+""""""""""
+
+The first operand and the result have the same vector of floating-point type.
+The second operand is the vector mask and has the same number of elements as the
+result vector type. The third operand is the explicit vector length of the
+operation.
+
+Semantics:
+""""""""""
+
+The '``llvm.vp.roundtozero``' intrinsic performs floating-point roundeven
+(:ref:`llvm.trunc <int_llvm_trunc>`) of the first vector operand on each enabled lane. The
+result on disabled lanes is a :ref:`poison value <poisonvalues>`.
+
+Examples:
+"""""""""
+
+.. code-block:: llvm
+
+ %r = call <4 x float> @llvm.vp.roundtozero.v4f32(<4 x float> %a, <4 x i1> %mask, i32 %evl)
+ ;; For all lanes below %evl, %r is lane-wise equivalent to %also.r
+
+ %t = call <4 x float> @llvm.trunc.v4f32(<4 x float> %a)
+ %also.r = select <4 x i1> %mask, <4 x float> %t, <4 x float> poison
+
.. _int_mload_mstore:
Masked Vector Load and Store Intrinsics
diff --git a/llvm/include/llvm/IR/Intrinsics.td b/llvm/include/llvm/IR/Intrinsics.td
index cad7485dfe71e..410a0eb6cecb3 100644
--- a/llvm/include/llvm/IR/Intrinsics.td
+++ b/llvm/include/llvm/IR/Intrinsics.td
@@ -1612,6 +1612,10 @@ let IntrProperties = [IntrNoMem, IntrNoSync, IntrWillReturn] in {
[ LLVMMatchType<0>,
LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>,
llvm_i32_ty]>;
+ def int_vp_roundtozero : DefaultAttrsIntrinsic<[ llvm_anyvector_ty ],
+ [ LLVMMatchType<0>,
+ LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>,
+ llvm_i32_ty]>;
// Casts
def int_vp_trunc : DefaultAttrsIntrinsic<[ llvm_anyvector_ty ],
diff --git a/llvm/include/llvm/IR/VPIntrinsics.def b/llvm/include/llvm/IR/VPIntrinsics.def
index fc1b4aa7ed286..3aadadc56b9cd 100644
--- a/llvm/include/llvm/IR/VPIntrinsics.def
+++ b/llvm/include/llvm/IR/VPIntrinsics.def
@@ -281,6 +281,10 @@ END_REGISTER_VP(vp_round, VP_FROUND)
BEGIN_REGISTER_VP(vp_roundeven, 1, 2, VP_FROUNDEVEN, -1)
END_REGISTER_VP(vp_roundeven, VP_FROUNDEVEN)
+// llvm.vp.roundtozero(x,mask,vlen)
+BEGIN_REGISTER_VP(vp_roundtozero, 1, 2, VP_FROUNDTOZERO, -1)
+END_REGISTER_VP(vp_roundtozero, VP_FROUNDTOZERO)
+
///// } Floating-Point Arithmetic
///// Type Casts {
diff --git a/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp b/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
index 545e9e99878f9..1d9004b194366 100644
--- a/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
+++ b/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp
@@ -1049,6 +1049,7 @@ void DAGTypeLegalizer::SplitVectorResult(SDNode *N, unsigned ResNo) {
case ISD::FSIN:
case ISD::FSQRT: case ISD::VP_SQRT:
case ISD::FTRUNC:
+ case ISD::VP_FROUNDTOZERO:
case ISD::SINT_TO_FP:
case ISD::VP_SINT_TO_FP:
case ISD::TRUNCATE:
@@ -4094,6 +4095,7 @@ void DAGTypeLegalizer::WidenVectorResult(SDNode *N, unsigned ResNo) {
case ISD::VP_FFLOOR:
case ISD::VP_FROUND:
case ISD::VP_FROUNDEVEN:
+ case ISD::VP_FROUNDTOZERO:
case ISD::FREEZE:
case ISD::ARITH_FENCE:
case ISD::FCANONICALIZE:
diff --git a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
index c37e78d0d3eb8..c0d389877a6b2 100644
--- a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
+++ b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
@@ -465,7 +465,7 @@ RISCVTargetLowering::RISCVTargetLowering(const TargetMachine &TM,
ISD::VP_SETCC, ISD::VP_FP_ROUND, ISD::VP_FP_EXTEND,
ISD::VP_SQRT, ISD::VP_FMINNUM, ISD::VP_FMAXNUM,
ISD::VP_FCEIL, ISD::VP_FFLOOR, ISD::VP_FROUND,
- ISD::VP_FROUNDEVEN, ISD::VP_FCOPYSIGN};
+ ISD::VP_FROUNDEVEN, ISD::VP_FCOPYSIGN, ISD::VP_FROUNDTOZERO};
static const unsigned IntegerVecReduceOps[] = {
ISD::VECREDUCE_ADD, ISD::VECREDUCE_AND, ISD::VECREDUCE_OR,
@@ -1965,7 +1965,9 @@ static RISCVFPRndMode::RoundingMode matchRoundingOp(unsigned Opc) {
case ISD::FROUNDEVEN:
case ISD::VP_FROUNDEVEN:
return RISCVFPRndMode::RNE;
- case ISD::FTRUNC: return RISCVFPRndMode::RTZ;
+ case ISD::FTRUNC:
+ case ISD::VP_FROUNDTOZERO:
+ return RISCVFPRndMode::RTZ;
case ISD::FFLOOR:
case ISD::VP_FFLOOR:
return RISCVFPRndMode::RDN;
@@ -1981,8 +1983,8 @@ static RISCVFPRndMode::RoundingMode matchRoundingOp(unsigned Opc) {
}
// Expand vector FTRUNC, FCEIL, FFLOOR, FROUND, VP_FCEIL, VP_FFLOOR, VP_FROUND
-// and VP_FROUNDEVEN by converting to the integer domain and back. Taking care
-// to avoid converting values that are nan or already correct.
+// VP_FROUNDEVEN and VP_FROUNDTOZERO by converting to the integer domain and
+// back. Taking care to avoid converting values that are nan or already correct.
static SDValue
lowerFTRUNC_FCEIL_FFLOOR_FROUND(SDValue Op, SelectionDAG &DAG,
const RISCVSubtarget &Subtarget) {
@@ -2003,7 +2005,8 @@ lowerFTRUNC_FCEIL_FFLOOR_FROUND(SDValue Op, SelectionDAG &DAG,
bool IsVP = Op->getOpcode() == ISD::VP_FCEIL ||
Op->getOpcode() == ISD::VP_FFLOOR ||
Op->getOpcode() == ISD::VP_FROUND ||
- Op->getOpcode() == ISD::VP_FROUNDEVEN;
+ Op->getOpcode() == ISD::VP_FROUNDEVEN ||
+ Op->getOpcode() == ISD::VP_FROUNDTOZERO;
if (IsVP) {
Mask = Op.getOperand(1);
@@ -2053,7 +2056,8 @@ lowerFTRUNC_FCEIL_FFLOOR_FROUND(SDValue Op, SelectionDAG &DAG,
case ISD::FROUND:
case ISD::FROUNDEVEN:
case ISD::VP_FROUND:
- case ISD::VP_FROUNDEVEN: {
+ case ISD::VP_FROUNDEVEN:
+ case ISD::VP_FROUNDTOZERO: {
RISCVFPRndMode::RoundingMode FRM = matchRoundingOp(Op.getOpcode());
assert(FRM != RISCVFPRndMode::Invalid);
Truncated = DAG.getNode(RISCVISD::VFCVT_X_F_VL, DL, IntVT, Src,
@@ -3983,6 +3987,7 @@ SDValue RISCVTargetLowering::LowerOperation(SDValue Op,
case ISD::VP_FFLOOR:
case ISD::VP_FROUND:
case ISD::VP_FROUNDEVEN:
+ case ISD::VP_FROUNDTOZERO:
return lowerFTRUNC_FCEIL_FFLOOR_FROUND(Op, DAG, Subtarget);
}
}
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
new file mode 100644
index 0000000000000..6fa9e6c4f55d5
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
@@ -0,0 +1,807 @@
+; 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
+
+declare <2 x half> @llvm.vp.roundtozero.v2f16(<2 x half>, <2 x i1>, i32)
+
+define <2 x half> @vp_roundtozero_v2f16(<2 x half> %va, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v2f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI0_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI0_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
+; 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, e16, mf4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <2 x half> @llvm.vp.roundtozero.v2f16(<2 x half> %va, <2 x i1> %m, i32 %evl)
+ ret <2 x half> %v
+}
+
+define <2 x half> @vp_roundtozero_v2f16_unmasked(<2 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v2f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI1_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI1_0)(a1)
+; CHECK-NEXT: vsetivli zero, 2, e8, mf8, ta, ma
+; CHECK-NEXT: vmset.m v0
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
+; 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, e16, mf4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <2 x i1> poison, i1 true, i32 0
+ %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+ %v = call <2 x half> @llvm.vp.roundtozero.v2f16(<2 x half> %va, <2 x i1> %m, i32 %evl)
+ ret <2 x half> %v
+}
+
+declare <4 x half> @llvm.vp.roundtozero.v4f16(<4 x half>, <4 x i1>, i32)
+
+define <4 x half> @vp_roundtozero_v4f16(<4 x half> %va, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v4f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI2_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI2_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
+; 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, e16, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <4 x half> @llvm.vp.roundtozero.v4f16(<4 x half> %va, <4 x i1> %m, i32 %evl)
+ ret <4 x half> %v
+}
+
+define <4 x half> @vp_roundtozero_v4f16_unmasked(<4 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v4f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI3_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI3_0)(a1)
+; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; CHECK-NEXT: vmset.m v0
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
+; 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, e16, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <4 x i1> poison, i1 true, i32 0
+ %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+ %v = call <4 x half> @llvm.vp.roundtozero.v4f16(<4 x half> %va, <4 x i1> %m, i32 %evl)
+ ret <4 x half> %v
+}
+
+declare <8 x half> @llvm.vp.roundtozero.v8f16(<8 x half>, <8 x i1>, i32)
+
+define <8 x half> @vp_roundtozero_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v8f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI4_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI4_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
+; 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, e16, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <8 x half> @llvm.vp.roundtozero.v8f16(<8 x half> %va, <8 x i1> %m, i32 %evl)
+ ret <8 x half> %v
+}
+
+define <8 x half> @vp_roundtozero_v8f16_unmasked(<8 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v8f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI5_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI5_0)(a1)
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
+; CHECK-NEXT: vmset.m v0
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
+; 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, e16, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <8 x i1> poison, i1 true, i32 0
+ %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+ %v = call <8 x half> @llvm.vp.roundtozero.v8f16(<8 x half> %va, <8 x i1> %m, i32 %evl)
+ ret <8 x half> %v
+}
+
+declare <16 x half> @llvm.vp.roundtozero.v16f16(<16 x half>, <16 x i1>, i32)
+
+define <16 x half> @vp_roundtozero_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v16f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v10, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI6_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, e16, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <16 x half> @llvm.vp.roundtozero.v16f16(<16 x half> %va, <16 x i1> %m, i32 %evl)
+ ret <16 x half> %v
+}
+
+define <16 x half> @vp_roundtozero_v16f16_unmasked(<16 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v16f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI7_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI7_0)(a1)
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vmset.m v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, e16, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <16 x i1> poison, i1 true, i32 0
+ %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+ %v = call <16 x half> @llvm.vp.roundtozero.v16f16(<16 x half> %va, <16 x i1> %m, i32 %evl)
+ ret <16 x half> %v
+}
+
+declare <2 x float> @llvm.vp.roundtozero.v2f32(<2 x float>, <2 x i1>, i32)
+
+define <2 x float> @vp_roundtozero_v2f32(<2 x float> %va, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v2f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI8_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI8_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
+; 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.roundtozero.v2f32(<2 x float> %va, <2 x i1> %m, i32 %evl)
+ ret <2 x float> %v
+}
+
+define <2 x float> @vp_roundtozero_v2f32_unmasked(<2 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v2f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI9_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI9_0)(a1)
+; CHECK-NEXT: vsetivli zero, 2, e8, mf8, ta, ma
+; CHECK-NEXT: vmset.m v0
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
+; 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
+ %head = insertelement <2 x i1> poison, i1 true, i32 0
+ %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+ %v = call <2 x float> @llvm.vp.roundtozero.v2f32(<2 x float> %va, <2 x i1> %m, i32 %evl)
+ ret <2 x float> %v
+}
+
+declare <4 x float> @llvm.vp.roundtozero.v4f32(<4 x float>, <4 x i1>, i32)
+
+define <4 x float> @vp_roundtozero_v4f32(<4 x float> %va, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v4f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI10_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI10_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
+; 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.roundtozero.v4f32(<4 x float> %va, <4 x i1> %m, i32 %evl)
+ ret <4 x float> %v
+}
+
+define <4 x float> @vp_roundtozero_v4f32_unmasked(<4 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v4f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI11_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI11_0)(a1)
+; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; CHECK-NEXT: vmset.m v0
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
+; 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
+ %head = insertelement <4 x i1> poison, i1 true, i32 0
+ %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+ %v = call <4 x float> @llvm.vp.roundtozero.v4f32(<4 x float> %va, <4 x i1> %m, i32 %evl)
+ ret <4 x float> %v
+}
+
+declare <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float>, <8 x i1>, i32)
+
+define <8 x float> @vp_roundtozero_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v8f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v10, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI12_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI12_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float> %va, <8 x i1> %m, i32 %evl)
+ ret <8 x float> %v
+}
+
+define <8 x float> @vp_roundtozero_v8f32_unmasked(<8 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v8f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI13_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI13_0)(a1)
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
+; CHECK-NEXT: vmset.m v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <8 x i1> poison, i1 true, i32 0
+ %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+ %v = call <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float> %va, <8 x i1> %m, i32 %evl)
+ ret <8 x float> %v
+}
+
+declare <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float>, <16 x i1>, i32)
+
+define <16 x float> @vp_roundtozero_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v16f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v12, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI14_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI14_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float> %va, <16 x i1> %m, i32 %evl)
+ ret <16 x float> %v
+}
+
+define <16 x float> @vp_roundtozero_v16f32_unmasked(<16 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v16f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI15_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI15_0)(a1)
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vmset.m v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <16 x i1> poison, i1 true, i32 0
+ %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+ %v = call <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float> %va, <16 x i1> %m, i32 %evl)
+ ret <16 x float> %v
+}
+
+declare <2 x double> @llvm.vp.roundtozero.v2f64(<2 x double>, <2 x i1>, i32)
+
+define <2 x double> @vp_roundtozero_v2f64(<2 x double> %va, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v2f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI16_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI16_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, ma
+; 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, e64, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <2 x double> @llvm.vp.roundtozero.v2f64(<2 x double> %va, <2 x i1> %m, i32 %evl)
+ ret <2 x double> %v
+}
+
+define <2 x double> @vp_roundtozero_v2f64_unmasked(<2 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v2f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI17_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI17_0)(a1)
+; CHECK-NEXT: vsetivli zero, 2, e8, mf8, ta, ma
+; CHECK-NEXT: vmset.m v0
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, ma
+; 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, e64, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <2 x i1> poison, i1 true, i32 0
+ %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+ %v = call <2 x double> @llvm.vp.roundtozero.v2f64(<2 x double> %va, <2 x i1> %m, i32 %evl)
+ ret <2 x double> %v
+}
+
+declare <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double>, <4 x i1>, i32)
+
+define <4 x double> @vp_roundtozero_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v4f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v10, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, e64, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double> %va, <4 x i1> %m, i32 %evl)
+ ret <4 x double> %v
+}
+
+define <4 x double> @vp_roundtozero_v4f64_unmasked(<4 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v4f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI19_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI19_0)(a1)
+; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; CHECK-NEXT: vmset.m v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, e64, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <4 x i1> poison, i1 true, i32 0
+ %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+ %v = call <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double> %va, <4 x i1> %m, i32 %evl)
+ ret <4 x double> %v
+}
+
+declare <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double>, <8 x i1>, i32)
+
+define <8 x double> @vp_roundtozero_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v8f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v12, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double> %va, <8 x i1> %m, i32 %evl)
+ ret <8 x double> %v
+}
+
+define <8 x double> @vp_roundtozero_v8f64_unmasked(<8 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v8f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI21_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI21_0)(a1)
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
+; CHECK-NEXT: vmset.m v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <8 x i1> poison, i1 true, i32 0
+ %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+ %v = call <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double> %va, <8 x i1> %m, i32 %evl)
+ ret <8 x double> %v
+}
+
+declare <15 x double> @llvm.vp.roundtozero.v15f64(<15 x double>, <15 x i1>, i32)
+
+define <15 x double> @vp_roundtozero_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v15f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v16, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <15 x double> @llvm.vp.roundtozero.v15f64(<15 x double> %va, <15 x i1> %m, i32 %evl)
+ ret <15 x double> %v
+}
+
+define <15 x double> @vp_roundtozero_v15f64_unmasked(<15 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v15f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI23_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI23_0)(a1)
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vmset.m v16
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <15 x i1> poison, i1 true, i32 0
+ %m = shufflevector <15 x i1> %head, <15 x i1> poison, <15 x i32> zeroinitializer
+ %v = call <15 x double> @llvm.vp.roundtozero.v15f64(<15 x double> %va, <15 x i1> %m, i32 %evl)
+ ret <15 x double> %v
+}
+
+declare <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double>, <16 x i1>, i32)
+
+define <16 x double> @vp_roundtozero_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v16f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v16, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double> %va, <16 x i1> %m, i32 %evl)
+ ret <16 x double> %v
+}
+
+define <16 x double> @vp_roundtozero_v16f64_unmasked(<16 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v16f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI25_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI25_0)(a1)
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vmset.m v16
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <16 x i1> poison, i1 true, i32 0
+ %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+ %v = call <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double> %va, <16 x i1> %m, i32 %evl)
+ ret <16 x double> %v
+}
+
+declare <32 x double> @llvm.vp.roundtozero.v32f64(<32 x double>, <32 x i1>, i32)
+
+define <32 x double> @vp_roundtozero_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v32f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: sub sp, sp, a1
+; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: li a1, 0
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: addi a2, a0, -16
+; CHECK-NEXT: vslidedown.vi v25, v0, 2
+; CHECK-NEXT: bltu a0, a2, .LBB26_2
+; CHECK-NEXT: # %bb.1:
+; CHECK-NEXT: mv a1, a2
+; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v25, v8, ft0, v0.t
+; CHECK-NEXT: fsrmi a1, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: fsrm a1
+; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: bltu a0, a1, .LBB26_4
+; CHECK-NEXT: # %bb.3:
+; CHECK-NEXT: li a0, 16
+; CHECK-NEXT: .LBB26_4:
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8re8.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v24, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8re8.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
+; CHECK-NEXT: ret
+ %v = call <32 x double> @llvm.vp.roundtozero.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
+ ret <32 x double> %v
+}
+
+define <32 x double> @vp_roundtozero_v32f64_unmasked(<32 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_v32f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: sub sp, sp, a1
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: li a1, 0
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: addi a2, a0, -16
+; CHECK-NEXT: vmset.m v24
+; CHECK-NEXT: bltu a0, a2, .LBB27_2
+; CHECK-NEXT: # %bb.1:
+; CHECK-NEXT: mv a1, a2
+; CHECK-NEXT: .LBB27_2:
+; CHECK-NEXT: lui a2, %hi(.LCPI27_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI27_0)(a2)
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmv1r.v v25, v24
+; CHECK-NEXT: vmflt.vf v25, v8, ft0, v0.t
+; CHECK-NEXT: fsrmi a1, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: fsrm a1
+; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: bltu a0, a1, .LBB27_4
+; CHECK-NEXT: # %bb.3:
+; CHECK-NEXT: li a0, 16
+; CHECK-NEXT: .LBB27_4:
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8re8.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v24, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8re8.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
+; CHECK-NEXT: ret
+ %head = insertelement <32 x i1> poison, i1 true, i32 0
+ %m = shufflevector <32 x i1> %head, <32 x i1> poison, <32 x i32> zeroinitializer
+ %v = call <32 x double> @llvm.vp.roundtozero.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
+ ret <32 x double> %v
+}
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
new file mode 100644
index 0000000000000..8a8feb37fb156
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
@@ -0,0 +1,848 @@
+; 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 \
+; RUN: -verify-machineinstrs < %s | FileCheck %s
+; RUN: llc -mtriple=riscv64 -mattr=+d,+zfh,+experimental-zvfh,+v -target-abi=lp64d \
+; RUN: -verify-machineinstrs < %s | FileCheck %s
+
+declare <vscale x 1 x half> @llvm.vp.roundtozero.nxv1f16(<vscale x 1 x half>, <vscale x 1 x i1>, i32)
+
+define <vscale x 1 x half> @vp_roundtozero_nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv1f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI0_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI0_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
+; 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, e16, mf4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 1 x half> @llvm.vp.roundtozero.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x i1> %m, i32 %evl)
+ ret <vscale x 1 x half> %v
+}
+
+define <vscale x 1 x half> @vp_roundtozero_nxv1f16_unmasked(<vscale x 1 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv1f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI1_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI1_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e16, mf4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %v = call <vscale x 1 x half> @llvm.vp.roundtozero.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x i1> %m, i32 %evl)
+ ret <vscale x 1 x half> %v
+}
+
+declare <vscale x 2 x half> @llvm.vp.roundtozero.nxv2f16(<vscale x 2 x half>, <vscale x 2 x i1>, i32)
+
+define <vscale x 2 x half> @vp_roundtozero_nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv2f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI2_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI2_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
+; 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, e16, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 2 x half> @llvm.vp.roundtozero.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x i1> %m, i32 %evl)
+ ret <vscale x 2 x half> %v
+}
+
+define <vscale x 2 x half> @vp_roundtozero_nxv2f16_unmasked(<vscale x 2 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv2f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI3_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI3_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e16, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %v = call <vscale x 2 x half> @llvm.vp.roundtozero.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x i1> %m, i32 %evl)
+ ret <vscale x 2 x half> %v
+}
+
+declare <vscale x 4 x half> @llvm.vp.roundtozero.nxv4f16(<vscale x 4 x half>, <vscale x 4 x i1>, i32)
+
+define <vscale x 4 x half> @vp_roundtozero_nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv4f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI4_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI4_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
+; 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, e16, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 4 x half> @llvm.vp.roundtozero.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x i1> %m, i32 %evl)
+ ret <vscale x 4 x half> %v
+}
+
+define <vscale x 4 x half> @vp_roundtozero_nxv4f16_unmasked(<vscale x 4 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv4f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI5_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI5_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e16, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %v = call <vscale x 4 x half> @llvm.vp.roundtozero.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x i1> %m, i32 %evl)
+ ret <vscale x 4 x half> %v
+}
+
+declare <vscale x 8 x half> @llvm.vp.roundtozero.nxv8f16(<vscale x 8 x half>, <vscale x 8 x i1>, i32)
+
+define <vscale x 8 x half> @vp_roundtozero_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv8f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v10, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI6_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, e16, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 8 x half> @llvm.vp.roundtozero.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 %evl)
+ ret <vscale x 8 x half> %v
+}
+
+define <vscale x 8 x half> @vp_roundtozero_nxv8f16_unmasked(<vscale x 8 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv8f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI7_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI7_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vfabs.v v10, v8
+; CHECK-NEXT: vmflt.vf v0, v10, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e16, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %v = call <vscale x 8 x half> @llvm.vp.roundtozero.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 %evl)
+ ret <vscale x 8 x half> %v
+}
+
+declare <vscale x 16 x half> @llvm.vp.roundtozero.nxv16f16(<vscale x 16 x half>, <vscale x 16 x i1>, i32)
+
+define <vscale x 16 x half> @vp_roundtozero_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv16f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v12, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI8_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI8_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 16 x half> @llvm.vp.roundtozero.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 %evl)
+ ret <vscale x 16 x half> %v
+}
+
+define <vscale x 16 x half> @vp_roundtozero_nxv16f16_unmasked(<vscale x 16 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv16f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI9_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI9_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8
+; CHECK-NEXT: vmflt.vf v0, v12, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e16, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %v = call <vscale x 16 x half> @llvm.vp.roundtozero.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 %evl)
+ ret <vscale x 16 x half> %v
+}
+
+declare <vscale x 32 x half> @llvm.vp.roundtozero.nxv32f16(<vscale x 32 x half>, <vscale x 32 x i1>, i32)
+
+define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv32f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v16, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI10_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI10_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 32 x half> @llvm.vp.roundtozero.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
+ ret <vscale x 32 x half> %v
+}
+
+define <vscale x 32 x half> @vp_roundtozero_nxv32f16_unmasked(<vscale x 32 x half> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv32f16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI11_0)
+; CHECK-NEXT: flh ft0, %lo(.LCPI11_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8
+; CHECK-NEXT: vmflt.vf v0, v16, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 32 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 32 x i1> %head, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %v = call <vscale x 32 x half> @llvm.vp.roundtozero.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
+ ret <vscale x 32 x half> %v
+}
+
+declare <vscale x 1 x float> @llvm.vp.roundtozero.nxv1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
+
+define <vscale x 1 x float> @vp_roundtozero_nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv1f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI12_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI12_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
+; 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 <vscale x 1 x float> @llvm.vp.roundtozero.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x i1> %m, i32 %evl)
+ ret <vscale x 1 x float> %v
+}
+
+define <vscale x 1 x float> @vp_roundtozero_nxv1f32_unmasked(<vscale x 1 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv1f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI13_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI13_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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
+ %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %v = call <vscale x 1 x float> @llvm.vp.roundtozero.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x i1> %m, i32 %evl)
+ ret <vscale x 1 x float> %v
+}
+
+declare <vscale x 2 x float> @llvm.vp.roundtozero.nxv2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
+
+define <vscale x 2 x float> @vp_roundtozero_nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv2f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI14_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI14_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
+; 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 <vscale x 2 x float> @llvm.vp.roundtozero.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x i1> %m, i32 %evl)
+ ret <vscale x 2 x float> %v
+}
+
+define <vscale x 2 x float> @vp_roundtozero_nxv2f32_unmasked(<vscale x 2 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv2f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI15_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI15_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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
+ %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %v = call <vscale x 2 x float> @llvm.vp.roundtozero.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x i1> %m, i32 %evl)
+ ret <vscale x 2 x float> %v
+}
+
+declare <vscale x 4 x float> @llvm.vp.roundtozero.nxv4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
+
+define <vscale x 4 x float> @vp_roundtozero_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv4f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v10, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI16_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI16_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 4 x float> @llvm.vp.roundtozero.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 %evl)
+ ret <vscale x 4 x float> %v
+}
+
+define <vscale x 4 x float> @vp_roundtozero_nxv4f32_unmasked(<vscale x 4 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv4f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI17_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI17_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vfabs.v v10, v8
+; CHECK-NEXT: vmflt.vf v0, v10, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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
+ %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %v = call <vscale x 4 x float> @llvm.vp.roundtozero.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 %evl)
+ ret <vscale x 4 x float> %v
+}
+
+declare <vscale x 8 x float> @llvm.vp.roundtozero.nxv8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
+
+define <vscale x 8 x float> @vp_roundtozero_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv8f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v12, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 8 x float> @llvm.vp.roundtozero.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 %evl)
+ ret <vscale x 8 x float> %v
+}
+
+define <vscale x 8 x float> @vp_roundtozero_nxv8f32_unmasked(<vscale x 8 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv8f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI19_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI19_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8
+; CHECK-NEXT: vmflt.vf v0, v12, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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
+ %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %v = call <vscale x 8 x float> @llvm.vp.roundtozero.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 %evl)
+ ret <vscale x 8 x float> %v
+}
+
+declare <vscale x 16 x float> @llvm.vp.roundtozero.nxv16f32(<vscale x 16 x float>, <vscale x 16 x i1>, i32)
+
+define <vscale x 16 x float> @vp_roundtozero_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv16f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v16, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 16 x float> @llvm.vp.roundtozero.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 %evl)
+ ret <vscale x 16 x float> %v
+}
+
+define <vscale x 16 x float> @vp_roundtozero_nxv16f32_unmasked(<vscale x 16 x float> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv16f32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI21_0)
+; CHECK-NEXT: flw ft0, %lo(.LCPI21_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8
+; CHECK-NEXT: vmflt.vf v0, v16, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %v = call <vscale x 16 x float> @llvm.vp.roundtozero.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 %evl)
+ ret <vscale x 16 x float> %v
+}
+
+declare <vscale x 1 x double> @llvm.vp.roundtozero.nxv1f64(<vscale x 1 x double>, <vscale x 1 x i1>, i32)
+
+define <vscale x 1 x double> @vp_roundtozero_nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv1f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
+; CHECK-NEXT: vmflt.vf v0, v9, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, ma
+; 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, e64, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 1 x double> @llvm.vp.roundtozero.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x i1> %m, i32 %evl)
+ ret <vscale x 1 x double> %v
+}
+
+define <vscale x 1 x double> @vp_roundtozero_nxv1f64_unmasked(<vscale x 1 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv1f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI23_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI23_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e64, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %v = call <vscale x 1 x double> @llvm.vp.roundtozero.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x i1> %m, i32 %evl)
+ ret <vscale x 1 x double> %v
+}
+
+declare <vscale x 2 x double> @llvm.vp.roundtozero.nxv2f64(<vscale x 2 x double>, <vscale x 2 x i1>, i32)
+
+define <vscale x 2 x double> @vp_roundtozero_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv2f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v10, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
+; CHECK-NEXT: vmflt.vf v10, v12, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v10
+; 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, e64, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 2 x double> @llvm.vp.roundtozero.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 %evl)
+ ret <vscale x 2 x double> %v
+}
+
+define <vscale x 2 x double> @vp_roundtozero_nxv2f64_unmasked(<vscale x 2 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv2f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI25_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI25_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vfabs.v v10, v8
+; CHECK-NEXT: vmflt.vf v0, v10, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e64, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %v = call <vscale x 2 x double> @llvm.vp.roundtozero.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 %evl)
+ ret <vscale x 2 x double> %v
+}
+
+declare <vscale x 4 x double> @llvm.vp.roundtozero.nxv4f64(<vscale x 4 x double>, <vscale x 4 x i1>, i32)
+
+define <vscale x 4 x double> @vp_roundtozero_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv4f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v12, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vmflt.vf v12, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 4 x double> @llvm.vp.roundtozero.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 %evl)
+ ret <vscale x 4 x double> %v
+}
+
+define <vscale x 4 x double> @vp_roundtozero_nxv4f64_unmasked(<vscale x 4 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv4f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI27_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI27_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vfabs.v v12, v8
+; CHECK-NEXT: vmflt.vf v0, v12, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; 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, e64, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %v = call <vscale x 4 x double> @llvm.vp.roundtozero.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 %evl)
+ ret <vscale x 4 x double> %v
+}
+
+declare <vscale x 7 x double> @llvm.vp.roundtozero.nxv7f64(<vscale x 7 x double>, <vscale x 7 x i1>, i32)
+
+define <vscale x 7 x double> @vp_roundtozero_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv7f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v16, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 7 x double> @llvm.vp.roundtozero.nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 %evl)
+ ret <vscale x 7 x double> %v
+}
+
+define <vscale x 7 x double> @vp_roundtozero_nxv7f64_unmasked(<vscale x 7 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv7f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI29_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI29_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8
+; CHECK-NEXT: vmflt.vf v0, v16, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 7 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 7 x i1> %head, <vscale x 7 x i1> poison, <vscale x 7 x i32> zeroinitializer
+ %v = call <vscale x 7 x double> @llvm.vp.roundtozero.nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 %evl)
+ ret <vscale x 7 x double> %v
+}
+
+declare <vscale x 8 x double> @llvm.vp.roundtozero.nxv8f64(<vscale x 8 x double>, <vscale x 8 x i1>, i32)
+
+define <vscale x 8 x double> @vp_roundtozero_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv8f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vmv1r.v v16, v0
+; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v16, v24, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: ret
+ %v = call <vscale x 8 x double> @llvm.vp.roundtozero.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 %evl)
+ ret <vscale x 8 x double> %v
+}
+
+define <vscale x 8 x double> @vp_roundtozero_nxv8f64_unmasked(<vscale x 8 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv8f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: lui a1, %hi(.LCPI31_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI31_0)(a1)
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v16, v8
+; CHECK-NEXT: vmflt.vf v0, v16, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %v = call <vscale x 8 x double> @llvm.vp.roundtozero.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 %evl)
+ ret <vscale x 8 x double> %v
+}
+
+; Test splitting.
+declare <vscale x 16 x double> @llvm.vp.roundtozero.nxv16f64(<vscale x 16 x double>, <vscale x 16 x i1>, i32)
+
+define <vscale x 16 x double> @vp_roundtozero_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv16f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: sub sp, sp, a1
+; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: li a2, 0
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: srli a4, a1, 3
+; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; CHECK-NEXT: sub a3, a0, a1
+; CHECK-NEXT: vslidedown.vx v25, v0, a4
+; CHECK-NEXT: bltu a0, a3, .LBB32_2
+; CHECK-NEXT: # %bb.1:
+; CHECK-NEXT: mv a2, a3
+; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v25, v8, ft0, v0.t
+; CHECK-NEXT: fsrmi a2, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: fsrm a2
+; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: bltu a0, a1, .LBB32_4
+; CHECK-NEXT: # %bb.3:
+; CHECK-NEXT: mv a0, a1
+; CHECK-NEXT: .LBB32_4:
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8re8.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vmflt.vf v24, v16, ft0, v0.t
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8re8.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
+; CHECK-NEXT: ret
+ %v = call <vscale x 16 x double> @llvm.vp.roundtozero.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
+ ret <vscale x 16 x double> %v
+}
+
+define <vscale x 16 x double> @vp_roundtozero_nxv16f64_unmasked(<vscale x 16 x double> %va, i32 zeroext %evl) {
+; CHECK-LABEL: vp_roundtozero_nxv16f64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: mv a2, a0
+; CHECK-NEXT: bltu a0, a1, .LBB33_2
+; CHECK-NEXT: # %bb.1:
+; CHECK-NEXT: mv a2, a1
+; CHECK-NEXT: .LBB33_2:
+; CHECK-NEXT: lui a3, %hi(.LCPI33_0)
+; CHECK-NEXT: fld ft0, %lo(.LCPI33_0)(a3)
+; CHECK-NEXT: li a3, 0
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v8
+; CHECK-NEXT: vmflt.vf v0, v24, ft0
+; CHECK-NEXT: fsrmi a2, 1
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: fsrm a2
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: sub a1, a0, a1
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: bltu a0, a1, .LBB33_4
+; CHECK-NEXT: # %bb.3:
+; CHECK-NEXT: mv a3, a1
+; CHECK-NEXT: .LBB33_4:
+; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v16
+; CHECK-NEXT: vmflt.vf v0, v24, ft0
+; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: fsrm a0
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
+; CHECK-NEXT: ret
+ %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+ %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %v = call <vscale x 16 x double> @llvm.vp.roundtozero.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
+ ret <vscale x 16 x double> %v
+}
diff --git a/llvm/unittests/IR/VPIntrinsicTest.cpp b/llvm/unittests/IR/VPIntrinsicTest.cpp
index be3a315ca53ff..91a9a7a60b964 100644
--- a/llvm/unittests/IR/VPIntrinsicTest.cpp
+++ b/llvm/unittests/IR/VPIntrinsicTest.cpp
@@ -59,6 +59,9 @@ class VPIntrinsicTest : public testing::Test {
Str << " declare <8 x float> @llvm.vp.roundeven.v8f32(<8 x float>, <8 x "
"i1>, "
"i32)";
+ Str << " declare <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float>, <8 x "
+ "i1>, "
+ "i32)";
Str << " declare <8 x float> @llvm.vp.ceil.v8f32(<8 x float>, <8 x i1>, "
"i32)";
Str << " declare <8 x float> @llvm.vp.fneg.v8f32(<8 x float>, <8 x i1>, "
More information about the llvm-commits
mailing list