This is an archive of the discontinued LLVM Phabricator instance.

Differential D137766

[RISCV] Add basic cost model for vp float rounding instructions.
ClosedPublic

Authored by fakepaper56 on Nov 9 2022, 11:24 PM.

Details

Reviewers
craig.topper
reames
frasercrmck
rogfer01
Commits
rG8c8a6e148824: [RISCV] Add basic cost model for vp float rounding instructions.

Diff Detail

Event Timeline

fakepaper56 created this revision.Nov 9 2022, 11:24 PM
Herald added a project: Restricted Project. · View Herald TranscriptNov 9 2022, 11:24 PM
Herald added subscribers: sunshaoce, VincentWu, StephenFan and 27 others. · View Herald Transcript
fakepaper56 requested review of this revision.Nov 9 2022, 11:24 PM
Herald added a project: Restricted Project. · View Herald TranscriptNov 9 2022, 11:24 PM
Herald added subscribers: llvm-commits, • pcwang-thead, eopXD, MaskRay. · View Herald Transcript
Harbormaster completed remote builds in B197025: Diff 474451.Nov 10 2022, 12:57 AM
fakepaper56 updated this revision to Diff 477705.Nov 24 2022, 1:51 AM

Rebase and ping.

Harbormaster completed remote builds in B199360: Diff 477705.Nov 24 2022, 2:34 AM
reames accepted this revision.Dec 5 2022, 12:52 PM

LGTM

This revision is now accepted and ready to land.Dec 5 2022, 12:52 PM
fakepaper56 updated this revision to Diff 480304.Dec 5 2022, 6:17 PM

Use mapped isd opcode on isOperationCustom for each intrinsic id.

fakepaper56 requested review of this revision.Dec 5 2022, 6:21 PM
Harbormaster completed remote builds in B201266: Diff 480304.Dec 6 2022, 12:54 AM
craig.topper accepted this revision.Dec 6 2022, 10:06 PM

LGTM

This revision is now accepted and ready to land.Dec 6 2022, 10:06 PM
Closed by commit rG8c8a6e148824: [RISCV] Add basic cost model for vp float rounding instructions. (authored by fakepaper56). · Explain WhyDec 6 2022, 10:15 PM
This revision was automatically updated to reflect the committed changes.
fakepaper56 added a commit: rG8c8a6e148824: [RISCV] Add basic cost model for vp float rounding instructions..

Revision Contents

PathSize
llvm/
lib/
Target/
RISCV/
13 lines
test/
Analysis/
CostModel/
RISCV/
300 lines

Diff 474451

llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp

//===-- RISCVTargetTransformInfo.cpp - RISC-V specific TTI ----------------===// //===-- RISCVTargetTransformInfo.cpp - RISC-V specific TTI ----------------===//
Lint: Lint
Inline Actions

clang-format not found in user’s local PATH; not linting file.

Lint: Lint: clang-format not found in user’s local PATH; not linting file.
// //
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information. // See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// //
//===----------------------------------------------------------------------===// //===----------------------------------------------------------------------===//
▲ Show 20 Lines • Show All 523 Lines • ▼ Show 20 LinesRISCVTTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
case Intrinsic::vp_rint: { case Intrinsic::vp_rint: {
// RISC-V target uses at least 5 instructions to lower rounding intrinsics. // RISC-V target uses at least 5 instructions to lower rounding intrinsics.
unsigned Cost = 5; unsigned Cost = 5;
auto LT = getTypeLegalizationCost(RetTy); auto LT = getTypeLegalizationCost(RetTy);
if (TLI->isOperationCustom(ISD::VP_FRINT, LT.second)) if (TLI->isOperationCustom(ISD::VP_FRINT, LT.second))
return Cost * LT.first; return Cost * LT.first;
break; break;
} }
case Intrinsic::vp_ceil:
case Intrinsic::vp_floor:
case Intrinsic::vp_round:
case Intrinsic::vp_roundeven:
case Intrinsic::vp_roundtozero: {
// Rounding with static rounding mode needs two more instructions to
// swap/write FRM than vp_rint.
unsigned Cost = 7;
auto LT = getTypeLegalizationCost(RetTy);
if (TLI->isOperationCustom(ISD::VP_FRINT, LT.second))
return Cost * LT.first;
break;
}
} }
if (ST->hasVInstructions() && RetTy->isVectorTy()) { if (ST->hasVInstructions() && RetTy->isVectorTy()) {
auto LT = getTypeLegalizationCost(RetTy); auto LT = getTypeLegalizationCost(RetTy);
if (const auto *Entry = CostTableLookup(VectorIntrinsicCostTable, if (const auto *Entry = CostTableLookup(VectorIntrinsicCostTable,
ICA.getID(), LT.second)) ICA.getID(), LT.second))
return LT.first * Entry->Cost; return LT.first * Entry->Cost;
} }
▲ Show 20 Lines • Show All 488 Lines • Show Last 20 Lines

llvm/test/Analysis/CostModel/RISCV/fround.ll

Show First 20 Lines • Show All 310 Lines • ▼ Show 20 Lines;
call <16 x double> @llvm.roundeven.v16f64(<16 x double> undef) call <16 x double> @llvm.roundeven.v16f64(<16 x double> undef)
call <vscale x 1 x double> @llvm.roundeven.nvx1f64(<vscale x 1 x double> undef) call <vscale x 1 x double> @llvm.roundeven.nvx1f64(<vscale x 1 x double> undef)
call <vscale x 2 x double> @llvm.roundeven.nvx2f64(<vscale x 2 x double> undef) call <vscale x 2 x double> @llvm.roundeven.nvx2f64(<vscale x 2 x double> undef)
call <vscale x 4 x double> @llvm.roundeven.nvx5f64(<vscale x 4 x double> undef) call <vscale x 4 x double> @llvm.roundeven.nvx5f64(<vscale x 4 x double> undef)
call <vscale x 8 x double> @llvm.roundeven.nvx8f64(<vscale x 8 x double> undef) call <vscale x 8 x double> @llvm.roundeven.nvx8f64(<vscale x 8 x double> undef)
ret void ret void
} }
define void @vp_ceil() {
; CHECK-LABEL: 'vp_ceil'
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %1 = call <2 x float> @llvm.vp.ceil.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %2 = call <4 x float> @llvm.vp.ceil.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %3 = call <8 x float> @llvm.vp.ceil.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %4 = call <16 x float> @llvm.vp.ceil.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.ceil.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.ceil.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %7 = call <vscale x 4 x float> @llvm.vp.ceil.nxv4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %8 = call <vscale x 8 x float> @llvm.vp.ceil.nxv8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %9 = call <vscale x 16 x float> @llvm.vp.ceil.nxv16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %10 = call <2 x double> @llvm.vp.ceil.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %11 = call <4 x double> @llvm.vp.ceil.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %12 = call <8 x double> @llvm.vp.ceil.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %13 = call <16 x double> @llvm.vp.ceil.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %14 = call <vscale x 1 x double> @llvm.vp.ceil.nxv1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %15 = call <vscale x 2 x double> @llvm.vp.ceil.nxv2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %16 = call <vscale x 4 x double> @llvm.vp.ceil.nxv4f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %17 = call <vscale x 8 x double> @llvm.vp.ceil.nxv8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret void
;
call <2 x float> @llvm.vp.ceil.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
call <4 x float> @llvm.vp.ceil.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
call <8 x float> @llvm.vp.ceil.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
call <16 x float> @llvm.vp.ceil.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x float> @llvm.vp.ceil.nvx1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x float> @llvm.vp.ceil.nvx2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x float> @llvm.vp.ceil.nvx4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x float> @llvm.vp.ceil.nvx8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
call <vscale x 16 x float> @llvm.vp.ceil.nvx16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
call <2 x double> @llvm.vp.ceil.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
call <4 x double> @llvm.vp.ceil.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
call <8 x double> @llvm.vp.ceil.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
call <16 x double> @llvm.vp.ceil.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x double> @llvm.vp.ceil.nvx1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x double> @llvm.vp.ceil.nvx2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x double> @llvm.vp.ceil.nvx5f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x double> @llvm.vp.ceil.nvx8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
ret void
}
define void @vp_floor() {
; CHECK-LABEL: 'vp_floor'
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %1 = call <2 x float> @llvm.vp.floor.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %2 = call <4 x float> @llvm.vp.floor.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %3 = call <8 x float> @llvm.vp.floor.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %4 = call <16 x float> @llvm.vp.floor.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.floor.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.floor.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %7 = call <vscale x 4 x float> @llvm.vp.floor.nxv4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %8 = call <vscale x 8 x float> @llvm.vp.floor.nxv8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %9 = call <vscale x 16 x float> @llvm.vp.floor.nxv16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %10 = call <2 x double> @llvm.vp.floor.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %11 = call <4 x double> @llvm.vp.floor.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %12 = call <8 x double> @llvm.vp.floor.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %13 = call <16 x double> @llvm.vp.floor.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %14 = call <vscale x 1 x double> @llvm.vp.floor.nxv1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %15 = call <vscale x 2 x double> @llvm.vp.floor.nxv2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %16 = call <vscale x 4 x double> @llvm.vp.floor.nxv4f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %17 = call <vscale x 8 x double> @llvm.vp.floor.nxv8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret void
;
call <2 x float> @llvm.vp.floor.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
call <4 x float> @llvm.vp.floor.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
call <8 x float> @llvm.vp.floor.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
call <16 x float> @llvm.vp.floor.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x float> @llvm.vp.floor.nvx1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x float> @llvm.vp.floor.nvx2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x float> @llvm.vp.floor.nvx4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x float> @llvm.vp.floor.nvx8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
call <vscale x 16 x float> @llvm.vp.floor.nvx16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
call <2 x double> @llvm.vp.floor.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
call <4 x double> @llvm.vp.floor.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
call <8 x double> @llvm.vp.floor.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
call <16 x double> @llvm.vp.floor.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x double> @llvm.vp.floor.nvx1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x double> @llvm.vp.floor.nvx2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x double> @llvm.vp.floor.nvx5f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x double> @llvm.vp.floor.nvx8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
ret void
}
define void @vp_round() {
; CHECK-LABEL: 'vp_round'
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %1 = call <2 x float> @llvm.vp.round.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %2 = call <4 x float> @llvm.vp.round.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %3 = call <8 x float> @llvm.vp.round.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %4 = call <16 x float> @llvm.vp.round.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.round.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.round.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %7 = call <vscale x 4 x float> @llvm.vp.round.nxv4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %8 = call <vscale x 8 x float> @llvm.vp.round.nxv8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %9 = call <vscale x 16 x float> @llvm.vp.round.nxv16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %10 = call <2 x double> @llvm.vp.round.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %11 = call <4 x double> @llvm.vp.round.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %12 = call <8 x double> @llvm.vp.round.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %13 = call <16 x double> @llvm.vp.round.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %14 = call <vscale x 1 x double> @llvm.vp.round.nxv1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %15 = call <vscale x 2 x double> @llvm.vp.round.nxv2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %16 = call <vscale x 4 x double> @llvm.vp.round.nxv4f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %17 = call <vscale x 8 x double> @llvm.vp.round.nxv8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret void
;
call <2 x float> @llvm.vp.round.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
call <4 x float> @llvm.vp.round.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
call <8 x float> @llvm.vp.round.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
call <16 x float> @llvm.vp.round.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x float> @llvm.vp.round.nvx1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x float> @llvm.vp.round.nvx2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x float> @llvm.vp.round.nvx4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x float> @llvm.vp.round.nvx8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
call <vscale x 16 x float> @llvm.vp.round.nvx16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
call <2 x double> @llvm.vp.round.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
call <4 x double> @llvm.vp.round.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
call <8 x double> @llvm.vp.round.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
call <16 x double> @llvm.vp.round.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x double> @llvm.vp.round.nvx1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x double> @llvm.vp.round.nvx2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x double> @llvm.vp.round.nvx5f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x double> @llvm.vp.round.nvx8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
ret void
}
define void @vp_roundeven() {
; CHECK-LABEL: 'vp_roundeven'
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %1 = call <2 x float> @llvm.vp.roundeven.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %2 = call <4 x float> @llvm.vp.roundeven.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %3 = call <8 x float> @llvm.vp.roundeven.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %4 = call <16 x float> @llvm.vp.roundeven.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.roundeven.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.roundeven.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %7 = call <vscale x 4 x float> @llvm.vp.roundeven.nxv4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %8 = call <vscale x 8 x float> @llvm.vp.roundeven.nxv8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %9 = call <vscale x 16 x float> @llvm.vp.roundeven.nxv16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %10 = call <2 x double> @llvm.vp.roundeven.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %11 = call <4 x double> @llvm.vp.roundeven.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %12 = call <8 x double> @llvm.vp.roundeven.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %13 = call <16 x double> @llvm.vp.roundeven.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %14 = call <vscale x 1 x double> @llvm.vp.roundeven.nxv1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %15 = call <vscale x 2 x double> @llvm.vp.roundeven.nxv2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %16 = call <vscale x 4 x double> @llvm.vp.roundeven.nxv4f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %17 = call <vscale x 8 x double> @llvm.vp.roundeven.nxv8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret void
;
call <2 x float> @llvm.vp.roundeven.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
call <4 x float> @llvm.vp.roundeven.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
call <8 x float> @llvm.vp.roundeven.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
call <16 x float> @llvm.vp.roundeven.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x float> @llvm.vp.roundeven.nvx1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x float> @llvm.vp.roundeven.nvx2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x float> @llvm.vp.roundeven.nvx4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x float> @llvm.vp.roundeven.nvx8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
call <vscale x 16 x float> @llvm.vp.roundeven.nvx16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
call <2 x double> @llvm.vp.roundeven.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
call <4 x double> @llvm.vp.roundeven.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
call <8 x double> @llvm.vp.roundeven.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
call <16 x double> @llvm.vp.roundeven.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x double> @llvm.vp.roundeven.nvx1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x double> @llvm.vp.roundeven.nvx2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x double> @llvm.vp.roundeven.nvx5f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x double> @llvm.vp.roundeven.nvx8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
ret void
}
define void @vp_roundtozero() {
; CHECK-LABEL: 'vp_roundtozero'
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %1 = call <2 x float> @llvm.vp.roundtozero.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %2 = call <4 x float> @llvm.vp.roundtozero.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %3 = call <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %4 = call <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.roundtozero.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.roundtozero.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %7 = call <vscale x 4 x float> @llvm.vp.roundtozero.nxv4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %8 = call <vscale x 8 x float> @llvm.vp.roundtozero.nxv8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %9 = call <vscale x 16 x float> @llvm.vp.roundtozero.nxv16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %10 = call <2 x double> @llvm.vp.roundtozero.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %11 = call <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %12 = call <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %13 = call <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %14 = call <vscale x 1 x double> @llvm.vp.roundtozero.nxv1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %15 = call <vscale x 2 x double> @llvm.vp.roundtozero.nxv2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %16 = call <vscale x 4 x double> @llvm.vp.roundtozero.nxv4f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %17 = call <vscale x 8 x double> @llvm.vp.roundtozero.nxv8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret void
;
call <2 x float> @llvm.vp.roundtozero.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
call <4 x float> @llvm.vp.roundtozero.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
call <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
call <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x float> @llvm.vp.roundtozero.nvx1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x float> @llvm.vp.roundtozero.nvx2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x float> @llvm.vp.roundtozero.nvx4f32(<vscale x 4 x float> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x float> @llvm.vp.roundtozero.nvx8f32(<vscale x 8 x float> undef, <vscale x 8 x i1> undef, i32 undef)
call <vscale x 16 x float> @llvm.vp.roundtozero.nvx16f32(<vscale x 16 x float> undef, <vscale x 16 x i1> undef, i32 undef)
call <2 x double> @llvm.vp.roundtozero.v2f64(<2 x double> undef, <2 x i1> undef, i32 undef)
call <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double> undef, <4 x i1> undef, i32 undef)
call <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double> undef, <8 x i1> undef, i32 undef)
call <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double> undef, <16 x i1> undef, i32 undef)
call <vscale x 1 x double> @llvm.vp.roundtozero.nvx1f64(<vscale x 1 x double> undef, <vscale x 1 x i1> undef, i32 undef)
call <vscale x 2 x double> @llvm.vp.roundtozero.nvx2f64(<vscale x 2 x double> undef, <vscale x 2 x i1> undef, i32 undef)
call <vscale x 4 x double> @llvm.vp.roundtozero.nvx5f64(<vscale x 4 x double> undef, <vscale x 4 x i1> undef, i32 undef)
call <vscale x 8 x double> @llvm.vp.roundtozero.nvx8f64(<vscale x 8 x double> undef, <vscale x 8 x i1> undef, i32 undef)
ret void
}
define void @vp_rint() { define void @vp_rint() {
; CHECK-LABEL: 'vp_rint' ; CHECK-LABEL: 'vp_rint'
; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %1 = call <2 x float> @llvm.vp.rint.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef) ; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %1 = call <2 x float> @llvm.vp.rint.v2f32(<2 x float> undef, <2 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %2 = call <4 x float> @llvm.vp.rint.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef) ; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %2 = call <4 x float> @llvm.vp.rint.v4f32(<4 x float> undef, <4 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %3 = call <8 x float> @llvm.vp.rint.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef) ; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %3 = call <8 x float> @llvm.vp.rint.v8f32(<8 x float> undef, <8 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %4 = call <16 x float> @llvm.vp.rint.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef) ; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %4 = call <16 x float> @llvm.vp.rint.v16f32(<16 x float> undef, <16 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.rint.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef) ; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %5 = call <vscale x 1 x float> @llvm.vp.rint.nxv1f32(<vscale x 1 x float> undef, <vscale x 1 x i1> undef, i32 undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.rint.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef) ; CHECK-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %6 = call <vscale x 2 x float> @llvm.vp.rint.nxv2f32(<vscale x 2 x float> undef, <vscale x 2 x i1> undef, i32 undef)
▲ Show 20 Lines • Show All 165 Lines • ▼ Show 20 Lines
declare <4 x double> @llvm.roundeven.v4f64(<4 x double>) declare <4 x double> @llvm.roundeven.v4f64(<4 x double>)
declare <8 x double> @llvm.roundeven.v8f64(<8 x double>) declare <8 x double> @llvm.roundeven.v8f64(<8 x double>)
declare <16 x double> @llvm.roundeven.v16f64(<16 x double>) declare <16 x double> @llvm.roundeven.v16f64(<16 x double>)
declare <vscale x 1 x double> @llvm.roundeven.nvx1f64(<vscale x 1 x double>) declare <vscale x 1 x double> @llvm.roundeven.nvx1f64(<vscale x 1 x double>)
declare <vscale x 2 x double> @llvm.roundeven.nvx2f64(<vscale x 2 x double>) declare <vscale x 2 x double> @llvm.roundeven.nvx2f64(<vscale x 2 x double>)
declare <vscale x 4 x double> @llvm.roundeven.nvx5f64(<vscale x 4 x double>) declare <vscale x 4 x double> @llvm.roundeven.nvx5f64(<vscale x 4 x double>)
declare <vscale x 8 x double> @llvm.roundeven.nvx8f64(<vscale x 8 x double>) declare <vscale x 8 x double> @llvm.roundeven.nvx8f64(<vscale x 8 x double>)
declare <2 x float> @llvm.vp.ceil.v2f32(<2 x float>, <2 x i1>, i32)
declare <4 x float> @llvm.vp.ceil.v4f32(<4 x float>, <4 x i1>, i32)
declare <8 x float> @llvm.vp.ceil.v8f32(<8 x float>, <8 x i1>, i32)
declare <16 x float> @llvm.vp.ceil.v16f32(<16 x float>, <16 x i1>, i32)
declare <vscale x 1 x float> @llvm.vp.ceil.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x float> @llvm.vp.ceil.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x float> @llvm.vp.ceil.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x float> @llvm.vp.ceil.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
declare <vscale x 16 x float> @llvm.vp.ceil.nvx16f32(<vscale x 16 x float>, <vscale x 16 x i1>, i32)
declare double @llvm.vp.ceil.f64(double)
declare <2 x double> @llvm.vp.ceil.v2f64(<2 x double>, <2 x i1>, i32)
declare <4 x double> @llvm.vp.ceil.v4f64(<4 x double>, <4 x i1>, i32)
declare <8 x double> @llvm.vp.ceil.v8f64(<8 x double>, <8 x i1>, i32)
declare <16 x double> @llvm.vp.ceil.v16f64(<16 x double>, <16 x i1>, i32)
declare <vscale x 1 x double> @llvm.vp.ceil.nvx1f64(<vscale x 1 x double>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x double> @llvm.vp.ceil.nvx2f64(<vscale x 2 x double>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x double> @llvm.vp.ceil.nvx5f64(<vscale x 4 x double>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x double> @llvm.vp.ceil.nvx8f64(<vscale x 8 x double>, <vscale x 8 x i1>, i32)
declare <2 x float> @llvm.vp.floor.v2f32(<2 x float>, <2 x i1>, i32)
declare <4 x float> @llvm.vp.floor.v4f32(<4 x float>, <4 x i1>, i32)
declare <8 x float> @llvm.vp.floor.v8f32(<8 x float>, <8 x i1>, i32)
declare <16 x float> @llvm.vp.floor.v16f32(<16 x float>, <16 x i1>, i32)
declare <vscale x 1 x float> @llvm.vp.floor.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x float> @llvm.vp.floor.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x float> @llvm.vp.floor.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x float> @llvm.vp.floor.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
declare <vscale x 16 x float> @llvm.vp.floor.nvx16f32(<vscale x 16 x float>, <vscale x 16 x i1>, i32)
declare double @llvm.vp.floor.f64(double)
declare <2 x double> @llvm.vp.floor.v2f64(<2 x double>, <2 x i1>, i32)
declare <4 x double> @llvm.vp.floor.v4f64(<4 x double>, <4 x i1>, i32)
declare <8 x double> @llvm.vp.floor.v8f64(<8 x double>, <8 x i1>, i32)
declare <16 x double> @llvm.vp.floor.v16f64(<16 x double>, <16 x i1>, i32)
declare <vscale x 1 x double> @llvm.vp.floor.nvx1f64(<vscale x 1 x double>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x double> @llvm.vp.floor.nvx2f64(<vscale x 2 x double>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x double> @llvm.vp.floor.nvx5f64(<vscale x 4 x double>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x double> @llvm.vp.floor.nvx8f64(<vscale x 8 x double>, <vscale x 8 x i1>, i32)
declare <2 x float> @llvm.vp.round.v2f32(<2 x float>, <2 x i1>, i32)
declare <4 x float> @llvm.vp.round.v4f32(<4 x float>, <4 x i1>, i32)
declare <8 x float> @llvm.vp.round.v8f32(<8 x float>, <8 x i1>, i32)
declare <16 x float> @llvm.vp.round.v16f32(<16 x float>, <16 x i1>, i32)
declare <vscale x 1 x float> @llvm.vp.round.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x float> @llvm.vp.round.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x float> @llvm.vp.round.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x float> @llvm.vp.round.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
declare <vscale x 16 x float> @llvm.vp.round.nvx16f32(<vscale x 16 x float>, <vscale x 16 x i1>, i32)
declare double @llvm.vp.round.f64(double)
declare <2 x double> @llvm.vp.round.v2f64(<2 x double>, <2 x i1>, i32)
declare <4 x double> @llvm.vp.round.v4f64(<4 x double>, <4 x i1>, i32)
declare <8 x double> @llvm.vp.round.v8f64(<8 x double>, <8 x i1>, i32)
declare <16 x double> @llvm.vp.round.v16f64(<16 x double>, <16 x i1>, i32)
declare <vscale x 1 x double> @llvm.vp.round.nvx1f64(<vscale x 1 x double>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x double> @llvm.vp.round.nvx2f64(<vscale x 2 x double>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x double> @llvm.vp.round.nvx5f64(<vscale x 4 x double>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x double> @llvm.vp.round.nvx8f64(<vscale x 8 x double>, <vscale x 8 x i1>, i32)
declare <2 x float> @llvm.vp.roundeven.v2f32(<2 x float>, <2 x i1>, i32)
declare <4 x float> @llvm.vp.roundeven.v4f32(<4 x float>, <4 x i1>, i32)
declare <8 x float> @llvm.vp.roundeven.v8f32(<8 x float>, <8 x i1>, i32)
declare <16 x float> @llvm.vp.roundeven.v16f32(<16 x float>, <16 x i1>, i32)
declare <vscale x 1 x float> @llvm.vp.roundeven.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x float> @llvm.vp.roundeven.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x float> @llvm.vp.roundeven.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x float> @llvm.vp.roundeven.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
declare <vscale x 16 x float> @llvm.vp.roundeven.nvx16f32(<vscale x 16 x float>, <vscale x 16 x i1>, i32)
declare double @llvm.vp.roundeven.f64(double)
declare <2 x double> @llvm.vp.roundeven.v2f64(<2 x double>, <2 x i1>, i32)
declare <4 x double> @llvm.vp.roundeven.v4f64(<4 x double>, <4 x i1>, i32)
declare <8 x double> @llvm.vp.roundeven.v8f64(<8 x double>, <8 x i1>, i32)
declare <16 x double> @llvm.vp.roundeven.v16f64(<16 x double>, <16 x i1>, i32)
declare <vscale x 1 x double> @llvm.vp.roundeven.nvx1f64(<vscale x 1 x double>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x double> @llvm.vp.roundeven.nvx2f64(<vscale x 2 x double>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x double> @llvm.vp.roundeven.nvx5f64(<vscale x 4 x double>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x double> @llvm.vp.roundeven.nvx8f64(<vscale x 8 x double>, <vscale x 8 x i1>, i32)
declare <2 x float> @llvm.vp.roundtozero.v2f32(<2 x float>, <2 x i1>, i32)
declare <4 x float> @llvm.vp.roundtozero.v4f32(<4 x float>, <4 x i1>, i32)
declare <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float>, <8 x i1>, i32)
declare <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float>, <16 x i1>, i32)
declare <vscale x 1 x float> @llvm.vp.roundtozero.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x float> @llvm.vp.roundtozero.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x float> @llvm.vp.roundtozero.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x float> @llvm.vp.roundtozero.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
declare <vscale x 16 x float> @llvm.vp.roundtozero.nvx16f32(<vscale x 16 x float>, <vscale x 16 x i1>, i32)
declare double @llvm.vp.roundtozero.f64(double)
declare <2 x double> @llvm.vp.roundtozero.v2f64(<2 x double>, <2 x i1>, i32)
declare <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double>, <4 x i1>, i32)
declare <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double>, <8 x i1>, i32)
declare <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double>, <16 x i1>, i32)
declare <vscale x 1 x double> @llvm.vp.roundtozero.nvx1f64(<vscale x 1 x double>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x double> @llvm.vp.roundtozero.nvx2f64(<vscale x 2 x double>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x double> @llvm.vp.roundtozero.nvx5f64(<vscale x 4 x double>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x double> @llvm.vp.roundtozero.nvx8f64(<vscale x 8 x double>, <vscale x 8 x i1>, i32)
declare <2 x float> @llvm.vp.rint.v2f32(<2 x float>, <2 x i1>, i32) declare <2 x float> @llvm.vp.rint.v2f32(<2 x float>, <2 x i1>, i32)
declare <4 x float> @llvm.vp.rint.v4f32(<4 x float>, <4 x i1>, i32) declare <4 x float> @llvm.vp.rint.v4f32(<4 x float>, <4 x i1>, i32)
declare <8 x float> @llvm.vp.rint.v8f32(<8 x float>, <8 x i1>, i32) declare <8 x float> @llvm.vp.rint.v8f32(<8 x float>, <8 x i1>, i32)
declare <16 x float> @llvm.vp.rint.v16f32(<16 x float>, <16 x i1>, i32) declare <16 x float> @llvm.vp.rint.v16f32(<16 x float>, <16 x i1>, i32)
declare <vscale x 1 x float> @llvm.vp.rint.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32) declare <vscale x 1 x float> @llvm.vp.rint.nvx1f32(<vscale x 1 x float>, <vscale x 1 x i1>, i32)
declare <vscale x 2 x float> @llvm.vp.rint.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32) declare <vscale x 2 x float> @llvm.vp.rint.nvx2f32(<vscale x 2 x float>, <vscale x 2 x i1>, i32)
declare <vscale x 4 x float> @llvm.vp.rint.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32) declare <vscale x 4 x float> @llvm.vp.rint.nvx4f32(<vscale x 4 x float>, <vscale x 4 x i1>, i32)
declare <vscale x 8 x float> @llvm.vp.rint.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32) declare <vscale x 8 x float> @llvm.vp.rint.nvx8f32(<vscale x 8 x float>, <vscale x 8 x i1>, i32)
Show All 10 Lines