[llvm] [VectorUtils] Trivially vectorize ldexp, [l]lround (PR #145545)
Ramkumar Ramachandra via llvm-commits
llvm-commits at lists.llvm.org
Wed Jun 25 04:25:07 PDT 2025
https://github.com/artagnon updated https://github.com/llvm/llvm-project/pull/145545
>From 6ce3e2ac96616ce2b3ed6b5b8b02cde1f50d152a Mon Sep 17 00:00:00 2001
From: Ramkumar Ramachandra <ramkumar.ramachandra at codasip.com>
Date: Tue, 24 Jun 2025 17:04:24 +0100
Subject: [PATCH 1/2] [VectorUtils] Trivially vectorize ldexp, [l]lround
---
llvm/lib/Analysis/VectorUtils.cpp | 6 +
.../Transforms/LoopVectorize/intrinsic.ll | 201 ++++++++++++++++++
2 files changed, 207 insertions(+)
diff --git a/llvm/lib/Analysis/VectorUtils.cpp b/llvm/lib/Analysis/VectorUtils.cpp
index 63fccee63c0ae..a391e92e84fc6 100644
--- a/llvm/lib/Analysis/VectorUtils.cpp
+++ b/llvm/lib/Analysis/VectorUtils.cpp
@@ -81,6 +81,7 @@ bool llvm::isTriviallyVectorizable(Intrinsic::ID ID) {
case Intrinsic::exp:
case Intrinsic::exp10:
case Intrinsic::exp2:
+ case Intrinsic::ldexp:
case Intrinsic::log:
case Intrinsic::log10:
case Intrinsic::log2:
@@ -108,6 +109,8 @@ bool llvm::isTriviallyVectorizable(Intrinsic::ID ID) {
case Intrinsic::canonicalize:
case Intrinsic::fptosi_sat:
case Intrinsic::fptoui_sat:
+ case Intrinsic::lround:
+ case Intrinsic::llround:
case Intrinsic::lrint:
case Intrinsic::llrint:
case Intrinsic::ucmp:
@@ -189,6 +192,8 @@ bool llvm::isVectorIntrinsicWithOverloadTypeAtArg(
switch (ID) {
case Intrinsic::fptosi_sat:
case Intrinsic::fptoui_sat:
+ case Intrinsic::lround:
+ case Intrinsic::llround:
case Intrinsic::lrint:
case Intrinsic::llrint:
case Intrinsic::vp_lrint:
@@ -203,6 +208,7 @@ bool llvm::isVectorIntrinsicWithOverloadTypeAtArg(
case Intrinsic::vp_is_fpclass:
return OpdIdx == 0;
case Intrinsic::powi:
+ case Intrinsic::ldexp:
return OpdIdx == -1 || OpdIdx == 1;
default:
return OpdIdx == -1;
diff --git a/llvm/test/Transforms/LoopVectorize/intrinsic.ll b/llvm/test/Transforms/LoopVectorize/intrinsic.ll
index 9c910d70807a1..10d83a456d0e2 100644
--- a/llvm/test/Transforms/LoopVectorize/intrinsic.ll
+++ b/llvm/test/Transforms/LoopVectorize/intrinsic.ll
@@ -324,6 +324,56 @@ for.end: ; preds = %for.body, %entry
declare double @llvm.exp2.f64(double)
+define void @ldexp_f32i32(i32 %n, ptr %y, ptr %x, i32 %exp) {
+; CHECK-LABEL: @ldexp_f32i32(
+; CHECK: llvm.ldexp.v4f32.v4i32
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds float, ptr %y, i32 %iv
+ %0 = load float, ptr %arrayidx, align 4
+ %call = tail call float @llvm.ldexp.f32.i32(float %0, i32 %exp)
+ %arrayidx2 = getelementptr inbounds float, ptr %x, i32 %iv
+ store float %call, ptr %arrayidx2, align 4
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare float @llvm.ldexp.f32.i32(float, i32)
+
+define void @ldexp_f64i32(i32 %n, ptr %y, ptr %x, i32 %exp) {
+; CHECK-LABEL: @ldexp_f64i32(
+; CHECK: llvm.ldexp.v4f64.v4i32
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds double, ptr %y, i32 %iv
+ %0 = load double, ptr %arrayidx, align 8
+ %call = tail call double @llvm.ldexp.f64.i32(double %0, i32 %exp)
+ %arrayidx2 = getelementptr inbounds double, ptr %x, i32 %iv
+ store double %call, ptr %arrayidx2, align 8
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare double @llvm.ldexp.f64.i32(double, i32)
+
define void @log_f32(i32 %n, ptr %y, ptr %x) {
; CHECK-LABEL: @log_f32(
; CHECK: llvm.log.v4f32
@@ -976,6 +1026,157 @@ for.end: ; preds = %for.body, %entry
declare double @llvm.roundeven.f64(double)
+
+define void @lround_i32f32(i32 %n, ptr %y, ptr %x) {
+; CHECK-LABEL: @lround_i32f32(
+; CHECK: llvm.lround.v4i32.v4f32
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds float, ptr %y, i32 %iv
+ %0 = load float, ptr %arrayidx, align 4
+ %call = tail call i32 @llvm.lround.i32.f32(float %0)
+ %arrayidx2 = getelementptr inbounds i32, ptr %x, i32 %iv
+ store i32 %call, ptr %arrayidx2, align 4
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare i32 @llvm.lround.i32.f32(float)
+
+define void @lround_i32f64(i32 %n, ptr %y, ptr %x) {
+; CHECK-LABEL: @lround_i32f64(
+; CHECK: llvm.lround.v4i32.v4f64
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds double, ptr %y, i32 %iv
+ %0 = load double, ptr %arrayidx, align 8
+ %call = tail call i32 @llvm.lround.i32.f64(double %0)
+ %arrayidx2 = getelementptr inbounds i32, ptr %x, i32 %iv
+ store i32 %call, ptr %arrayidx2, align 8
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare i32 @llvm.lround.i32.f64(double)
+
+define void @lround_i64f32(i32 %n, ptr %y, ptr %x) {
+; CHECK-LABEL: @lround_i64f32(
+; CHECK: llvm.lround.v4i64.v4f32
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds float, ptr %y, i32 %iv
+ %0 = load float, ptr %arrayidx, align 4
+ %call = tail call i64 @llvm.lround.i64.f32(float %0)
+ %arrayidx2 = getelementptr inbounds i64, ptr %x, i32 %iv
+ store i64 %call, ptr %arrayidx2, align 4
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare i64 @llvm.lround.i64.f32(float)
+
+define void @lround_i64f64(i32 %n, ptr %y, ptr %x) {
+; CHECK-LABEL: @lround_i64f64(
+; CHECK: llvm.lround.v4i64.v4f64
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds double, ptr %y, i32 %iv
+ %0 = load double, ptr %arrayidx, align 8
+ %call = tail call i64 @llvm.lround.i64.f64(double %0)
+ %arrayidx2 = getelementptr inbounds i64, ptr %x, i32 %iv
+ store i64 %call, ptr %arrayidx2, align 8
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare i64 @llvm.lround.i64.f64(double)
+
+define void @llround_i64f32(i32 %n, ptr %y, ptr %x) {
+; CHECK-LABEL: @llround_i64f32(
+; CHECK: llvm.llround.v4i64.v4f32
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds float, ptr %y, i32 %iv
+ %0 = load float, ptr %arrayidx, align 4
+ %call = tail call i64 @llvm.llround.i64.f32(float %0)
+ %arrayidx2 = getelementptr inbounds i64, ptr %x, i32 %iv
+ store i64 %call, ptr %arrayidx2, align 4
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare i64 @llvm.llround.i64.f32(float)
+
+define void @llround_i64f64(i32 %n, ptr %y, ptr %x) {
+; CHECK-LABEL: @llround_i64f64(
+; CHECK: llvm.llround.v4i64.v4f64
+; CHECK: ret void
+;
+entry:
+ br label %for.body
+
+for.body: ; preds = %entry, %for.body
+ %iv = phi i32 [ 0, %entry ], [ %iv.next, %for.body ]
+ %arrayidx = getelementptr inbounds double, ptr %y, i32 %iv
+ %0 = load double, ptr %arrayidx, align 8
+ %call = tail call i64 @llvm.llround.i64.f64(double %0)
+ %arrayidx2 = getelementptr inbounds i64, ptr %x, i32 %iv
+ store i64 %call, ptr %arrayidx2, align 8
+ %iv.next = add i32 %iv, 1
+ %exitcond = icmp eq i32 %iv.next, %n
+ br i1 %exitcond, label %for.end, label %for.body
+
+for.end: ; preds = %for.body, %entry
+ ret void
+}
+
+declare i64 @llvm.llround.i64.f64(double)
+
define void @fma_f32(i32 %n, ptr %y, ptr %x, ptr %z, ptr %w) {
; CHECK-LABEL: @fma_f32(
; CHECK: llvm.fma.v4f32
>From bff41722ecf9305715209998354349f36bfd066d Mon Sep 17 00:00:00 2001
From: Ramkumar Ramachandra <ramkumar.ramachandra at codasip.com>
Date: Wed, 25 Jun 2025 12:21:12 +0100
Subject: [PATCH 2/2] [CostModel] Add tests for ldexp, [l]lround
---
llvm/test/Analysis/CostModel/RISCV/exp.ll | 109 +++++++++++
llvm/test/Analysis/CostModel/RISCV/fround.ll | 194 +++++++++++++++++++
2 files changed, 303 insertions(+)
create mode 100644 llvm/test/Analysis/CostModel/RISCV/exp.ll
diff --git a/llvm/test/Analysis/CostModel/RISCV/exp.ll b/llvm/test/Analysis/CostModel/RISCV/exp.ll
new file mode 100644
index 0000000000000..fa7a9451f567a
--- /dev/null
+++ b/llvm/test/Analysis/CostModel/RISCV/exp.ll
@@ -0,0 +1,109 @@
+; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
+; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -S -mtriple=riscv64 -mattr=+v,+f,+d,+zvfh,+zvfbfmin %s | FileCheck %s
+
+define void @ldexp() {
+; CHECK-LABEL: 'ldexp'
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call bfloat @llvm.ldexp.bf16.i32(bfloat poison, i32 poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %2 = call <2 x bfloat> @llvm.ldexp.v2bf16.v2i32(<2 x bfloat> poison, <2 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %3 = call <4 x bfloat> @llvm.ldexp.v4bf16.v4i32(<4 x bfloat> poison, <4 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %4 = call <8 x bfloat> @llvm.ldexp.v8bf16.v8i32(<8 x bfloat> poison, <8 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %5 = call <16 x bfloat> @llvm.ldexp.v16bf16.v16i32(<16 x bfloat> poison, <16 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x bfloat> @llvm.ldexp.nxv1bf16.nxv1i32(<vscale x 1 x bfloat> poison, <vscale x 1 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x bfloat> @llvm.ldexp.nxv2bf16.nxv2i32(<vscale x 2 x bfloat> poison, <vscale x 2 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x bfloat> @llvm.ldexp.nxv4bf16.nxv4i32(<vscale x 4 x bfloat> poison, <vscale x 4 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x bfloat> @llvm.ldexp.nxv8bf16.nxv8i32(<vscale x 8 x bfloat> poison, <vscale x 8 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 16 x bfloat> @llvm.ldexp.nxv16bf16.nxv16i32(<vscale x 16 x bfloat> poison, <vscale x 16 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %11 = call float @llvm.ldexp.f32.i32(float poison, i32 poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %12 = call <2 x float> @llvm.ldexp.v2f32.v2i32(<2 x float> poison, <2 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %13 = call <4 x float> @llvm.ldexp.v4f32.v4i32(<4 x float> poison, <4 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %14 = call <8 x float> @llvm.ldexp.v8f32.v8i32(<8 x float> poison, <8 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %15 = call <16 x float> @llvm.ldexp.v16f32.v16i32(<16 x float> poison, <16 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %16 = call <vscale x 1 x float> @llvm.ldexp.nxv1f32.nxv1i32(<vscale x 1 x float> poison, <vscale x 1 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %17 = call <vscale x 2 x float> @llvm.ldexp.nxv2f32.nxv2i32(<vscale x 2 x float> poison, <vscale x 2 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %18 = call <vscale x 4 x float> @llvm.ldexp.nxv4f32.nxv4i32(<vscale x 4 x float> poison, <vscale x 4 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %19 = call <vscale x 8 x float> @llvm.ldexp.nxv8f32.nxv8i32(<vscale x 8 x float> poison, <vscale x 8 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %20 = call <vscale x 16 x float> @llvm.ldexp.nxv16f32.nxv16i32(<vscale x 16 x float> poison, <vscale x 16 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %21 = call double @llvm.ldexp.f64.i32(double poison, i32 poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %22 = call <2 x double> @llvm.ldexp.v2f64.v2i32(<2 x double> poison, <2 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %23 = call <4 x double> @llvm.ldexp.v4f64.v4i32(<4 x double> poison, <4 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %24 = call <8 x double> @llvm.ldexp.v8f64.v8i32(<8 x double> poison, <8 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %25 = call <16 x double> @llvm.ldexp.v16f64.v16i32(<16 x double> poison, <16 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %26 = call <vscale x 1 x double> @llvm.ldexp.nxv1f64.nxv1i32(<vscale x 1 x double> poison, <vscale x 1 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %27 = call <vscale x 2 x double> @llvm.ldexp.nxv2f64.nxv2i32(<vscale x 2 x double> poison, <vscale x 2 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %28 = call <vscale x 4 x double> @llvm.ldexp.nxv4f64.nxv4i32(<vscale x 4 x double> poison, <vscale x 4 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %29 = call <vscale x 8 x double> @llvm.ldexp.nxv8f64.nxv8i32(<vscale x 8 x double> poison, <vscale x 8 x i32> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %30 = call <vscale x 16 x double> @llvm.ldexp.nxv16f64.nxv16i32(<vscale x 16 x double> poison, <vscale x 16 x i32> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %31 = call float @llvm.ldexp.f32.i64(float poison, i64 poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %32 = call <2 x float> @llvm.ldexp.v2f32.v2i64(<2 x float> poison, <2 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %33 = call <4 x float> @llvm.ldexp.v4f32.v4i64(<4 x float> poison, <4 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %34 = call <8 x float> @llvm.ldexp.v8f32.v8i64(<8 x float> poison, <8 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %35 = call <16 x float> @llvm.ldexp.v16f32.v16i64(<16 x float> poison, <16 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %36 = call <vscale x 1 x float> @llvm.ldexp.nxv1f32.nxv1i64(<vscale x 1 x float> poison, <vscale x 1 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %37 = call <vscale x 2 x float> @llvm.ldexp.nxv2f32.nxv2i64(<vscale x 2 x float> poison, <vscale x 2 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %38 = call <vscale x 4 x float> @llvm.ldexp.nxv4f32.nxv4i64(<vscale x 4 x float> poison, <vscale x 4 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %39 = call <vscale x 8 x float> @llvm.ldexp.nxv8f32.nxv8i64(<vscale x 8 x float> poison, <vscale x 8 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %40 = call <vscale x 16 x float> @llvm.ldexp.nxv16f32.nxv16i64(<vscale x 16 x float> poison, <vscale x 16 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %41 = call double @llvm.ldexp.f64.i64(double poison, i64 poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %42 = call <2 x double> @llvm.ldexp.v2f64.v2i64(<2 x double> poison, <2 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %43 = call <4 x double> @llvm.ldexp.v4f64.v4i64(<4 x double> poison, <4 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %44 = call <8 x double> @llvm.ldexp.v8f64.v8i64(<8 x double> poison, <8 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %45 = call <16 x double> @llvm.ldexp.v16f64.v16i64(<16 x double> poison, <16 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %46 = call <vscale x 1 x double> @llvm.ldexp.nxv1f64.nxv1i64(<vscale x 1 x double> poison, <vscale x 1 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %47 = call <vscale x 2 x double> @llvm.ldexp.nxv2f64.nxv2i64(<vscale x 2 x double> poison, <vscale x 2 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %48 = call <vscale x 4 x double> @llvm.ldexp.nxv4f64.nxv4i64(<vscale x 4 x double> poison, <vscale x 4 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %49 = call <vscale x 8 x double> @llvm.ldexp.nxv8f64.nxv8i64(<vscale x 8 x double> poison, <vscale x 8 x i64> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %50 = call <vscale x 16 x double> @llvm.ldexp.nxv16f64.nxv16i64(<vscale x 16 x double> poison, <vscale x 16 x i64> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+;
+ call bfloat @llvm.ldexp.bf16.i32(bfloat poison, i32 poison)
+ call <2 x bfloat> @llvm.ldexp.v2bf16.v2i32(<2 x bfloat> poison, <2 x i32> poison)
+ call <4 x bfloat> @llvm.ldexp.v4bf16.v4i32(<4 x bfloat> poison, <4 x i32> poison)
+ call <8 x bfloat> @llvm.ldexp.v8bf16.v8i32(<8 x bfloat> poison, <8 x i32> poison)
+ call <16 x bfloat> @llvm.ldexp.v16bf16.v16i32(<16 x bfloat> poison, <16 x i32> poison)
+ call <vscale x 1 x bfloat> @llvm.ldexp.nxv1bf16.nxv1i32(<vscale x 1 x bfloat> poison, <vscale x 1 x i32> poison)
+ call <vscale x 2 x bfloat> @llvm.ldexp.nxv2bf16.nxv2i32(<vscale x 2 x bfloat> poison, <vscale x 2 x i32> poison)
+ call <vscale x 4 x bfloat> @llvm.ldexp.nxv4bf16.nxv4i32(<vscale x 4 x bfloat> poison, <vscale x 4 x i32> poison)
+ call <vscale x 8 x bfloat> @llvm.ldexp.nxv8bf16.nxv8i32(<vscale x 8 x bfloat> poison, <vscale x 8 x i32> poison)
+ call <vscale x 16 x bfloat> @llvm.ldexp.nxv16bf16.nxv16i32(<vscale x 16 x bfloat> poison, <vscale x 16 x i32> poison)
+ call float @llvm.ldexp.f32.i32(float poison, i32 poison)
+ call <2 x float> @llvm.ldexp.v2f32.v2i32(<2 x float> poison, <2 x i32> poison)
+ call <4 x float> @llvm.ldexp.v4f32.v4i32(<4 x float> poison, <4 x i32> poison)
+ call <8 x float> @llvm.ldexp.v8f32.v8i32(<8 x float> poison, <8 x i32> poison)
+ call <16 x float> @llvm.ldexp.v16f32.v16i32(<16 x float> poison, <16 x i32> poison)
+ call <vscale x 1 x float> @llvm.ldexp.nxv1f32.nxv1i32(<vscale x 1 x float> poison, <vscale x 1 x i32> poison)
+ call <vscale x 2 x float> @llvm.ldexp.nxv2f32.nxv2i32(<vscale x 2 x float> poison, <vscale x 2 x i32> poison)
+ call <vscale x 4 x float> @llvm.ldexp.nxv4f32.nxv4i32(<vscale x 4 x float> poison, <vscale x 4 x i32> poison)
+ call <vscale x 8 x float> @llvm.ldexp.nxv8f32.nxv8i32(<vscale x 8 x float> poison, <vscale x 8 x i32> poison)
+ call <vscale x 16 x float> @llvm.ldexp.nxv16f32.nxv16i32(<vscale x 16 x float> poison, <vscale x 16 x i32> poison)
+ call double @llvm.ldexp.f64.i32(double poison, i32 poison)
+ call <2 x double> @llvm.ldexp.v2f64.v2i32(<2 x double> poison, <2 x i32> poison)
+ call <4 x double> @llvm.ldexp.v4f64.v4i32(<4 x double> poison, <4 x i32> poison)
+ call <8 x double> @llvm.ldexp.v8f64.v8i32(<8 x double> poison, <8 x i32> poison)
+ call <16 x double> @llvm.ldexp.v16f64.v16i32(<16 x double> poison, <16 x i32> poison)
+ call <vscale x 1 x double> @llvm.ldexp.nxv1f64.nxv1i32(<vscale x 1 x double> poison, <vscale x 1 x i32> poison)
+ call <vscale x 2 x double> @llvm.ldexp.nxv2f64.nxv2i32(<vscale x 2 x double> poison, <vscale x 2 x i32> poison)
+ call <vscale x 4 x double> @llvm.ldexp.nxv4f64.nxv4i32(<vscale x 4 x double> poison, <vscale x 4 x i32> poison)
+ call <vscale x 8 x double> @llvm.ldexp.nxv8f64.nxv8i32(<vscale x 8 x double> poison, <vscale x 8 x i32> poison)
+ call <vscale x 16 x double> @llvm.ldexp.nxv16f64.nxv16i32(<vscale x 16 x double> poison, <vscale x 16 x i32> poison)
+ call float @llvm.ldexp.f32.i64(float poison, i64 poison)
+ call <2 x float> @llvm.ldexp.v2f32.v2i64(<2 x float> poison, <2 x i64> poison)
+ call <4 x float> @llvm.ldexp.v4f32.v4i64(<4 x float> poison, <4 x i64> poison)
+ call <8 x float> @llvm.ldexp.v8f32.v8i64(<8 x float> poison, <8 x i64> poison)
+ call <16 x float> @llvm.ldexp.v16f32.v16i64(<16 x float> poison, <16 x i64> poison)
+ call <vscale x 1 x float> @llvm.ldexp.nxv1f32.nxv1i64(<vscale x 1 x float> poison, <vscale x 1 x i64> poison)
+ call <vscale x 2 x float> @llvm.ldexp.nxv2f32.nxv2i64(<vscale x 2 x float> poison, <vscale x 2 x i64> poison)
+ call <vscale x 4 x float> @llvm.ldexp.nxv4f32.nxv4i64(<vscale x 4 x float> poison, <vscale x 4 x i64> poison)
+ call <vscale x 8 x float> @llvm.ldexp.nxv8f32.nxv8i64(<vscale x 8 x float> poison, <vscale x 8 x i64> poison)
+ call <vscale x 16 x float> @llvm.ldexp.nxv16f32.nxv16i64(<vscale x 16 x float> poison, <vscale x 16 x i64> poison)
+ call double @llvm.ldexp.f64.i64(double poison, i64 poison)
+ call <2 x double> @llvm.ldexp.v2f64.v2i64(<2 x double> poison, <2 x i64> poison)
+ call <4 x double> @llvm.ldexp.v4f64.v4i64(<4 x double> poison, <4 x i64> poison)
+ call <8 x double> @llvm.ldexp.v8f64.v8i64(<8 x double> poison, <8 x i64> poison)
+ call <16 x double> @llvm.ldexp.v16f64.v16i64(<16 x double> poison, <16 x i64> poison)
+ call <vscale x 1 x double> @llvm.ldexp.nxv1f64.nxv1i64(<vscale x 1 x double> poison, <vscale x 1 x i64> poison)
+ call <vscale x 2 x double> @llvm.ldexp.nxv2f64.nxv2i64(<vscale x 2 x double> poison, <vscale x 2 x i64> poison)
+ call <vscale x 4 x double> @llvm.ldexp.nxv4f64.nxv4i64(<vscale x 4 x double> poison, <vscale x 4 x i64> poison)
+ call <vscale x 8 x double> @llvm.ldexp.nxv8f64.nxv8i64(<vscale x 8 x double> poison, <vscale x 8 x i64> poison)
+ call <vscale x 16 x double> @llvm.ldexp.nxv16f64.nxv16i64(<vscale x 16 x double> poison, <vscale x 16 x i64> poison)
+ ret void
+}
diff --git a/llvm/test/Analysis/CostModel/RISCV/fround.ll b/llvm/test/Analysis/CostModel/RISCV/fround.ll
index a0818d487d151..94c09389cadec 100644
--- a/llvm/test/Analysis/CostModel/RISCV/fround.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/fround.ll
@@ -606,6 +606,200 @@ define void @llrint_fp16() {
ret void
}
+define void @lround() {
+; CHECK-LABEL: 'lround'
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call i32 @llvm.lround.i32.bf16(bfloat poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %2 = call <2 x i32> @llvm.lround.v2i32.v2bf16(<2 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %3 = call <4 x i32> @llvm.lround.v4i32.v4bf16(<4 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %4 = call <8 x i32> @llvm.lround.v8i32.v8bf16(<8 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %5 = call <16 x i32> @llvm.lround.v16i32.v16bf16(<16 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x i32> @llvm.lround.nxv1i32.nxv1bf16(<vscale x 1 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x i32> @llvm.lround.nxv2i32.nxv2bf16(<vscale x 2 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x i32> @llvm.lround.nxv4i32.nxv4bf16(<vscale x 4 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x i32> @llvm.lround.nxv8i32.nxv8bf16(<vscale x 8 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 16 x i32> @llvm.lround.nxv16i32.nxv16bf16(<vscale x 16 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %11 = call i32 @llvm.lround.i32.f32(float poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %12 = call <2 x i32> @llvm.lround.v2i32.v2f32(<2 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %13 = call <4 x i32> @llvm.lround.v4i32.v4f32(<4 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %14 = call <8 x i32> @llvm.lround.v8i32.v8f32(<8 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %15 = call <16 x i32> @llvm.lround.v16i32.v16f32(<16 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %16 = call <vscale x 1 x i32> @llvm.lround.nxv1i32.nxv1f32(<vscale x 1 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %17 = call <vscale x 2 x i32> @llvm.lround.nxv2i32.nxv2f32(<vscale x 2 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %18 = call <vscale x 4 x i32> @llvm.lround.nxv4i32.nxv4f32(<vscale x 4 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %19 = call <vscale x 8 x i32> @llvm.lround.nxv8i32.nxv8f32(<vscale x 8 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %20 = call <vscale x 16 x i32> @llvm.lround.nxv16i32.nxv16f32(<vscale x 16 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %21 = call i32 @llvm.lround.i32.f64(double poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %22 = call <2 x i32> @llvm.lround.v2i32.v2f64(<2 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %23 = call <4 x i32> @llvm.lround.v4i32.v4f64(<4 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %24 = call <8 x i32> @llvm.lround.v8i32.v8f64(<8 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %25 = call <16 x i32> @llvm.lround.v16i32.v16f64(<16 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %26 = call <vscale x 1 x i32> @llvm.lround.nxv1i32.nxv1f64(<vscale x 1 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %27 = call <vscale x 2 x i32> @llvm.lround.nxv2i32.nxv2f64(<vscale x 2 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %28 = call <vscale x 4 x i32> @llvm.lround.nxv4i32.nxv4f64(<vscale x 4 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %29 = call <vscale x 8 x i32> @llvm.lround.nxv8i32.nxv8f64(<vscale x 8 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %30 = call <vscale x 16 x i32> @llvm.lround.nxv16i32.nxv16f64(<vscale x 16 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %31 = call i64 @llvm.lround.i64.bf16(bfloat poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %32 = call <2 x i64> @llvm.lround.v2i64.v2bf16(<2 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %33 = call <4 x i64> @llvm.lround.v4i64.v4bf16(<4 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %34 = call <8 x i64> @llvm.lround.v8i64.v8bf16(<8 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %35 = call <16 x i64> @llvm.lround.v16i64.v16bf16(<16 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %36 = call <vscale x 1 x i64> @llvm.lround.nxv1i64.nxv1bf16(<vscale x 1 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %37 = call <vscale x 2 x i64> @llvm.lround.nxv2i64.nxv2bf16(<vscale x 2 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %38 = call <vscale x 4 x i64> @llvm.lround.nxv4i64.nxv4bf16(<vscale x 4 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %39 = call <vscale x 8 x i64> @llvm.lround.nxv8i64.nxv8bf16(<vscale x 8 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %40 = call <vscale x 16 x i64> @llvm.lround.nxv16i64.nxv16bf16(<vscale x 16 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %41 = call i64 @llvm.lround.i64.f32(float poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %42 = call <2 x i64> @llvm.lround.v2i64.v2f32(<2 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %43 = call <4 x i64> @llvm.lround.v4i64.v4f32(<4 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %44 = call <8 x i64> @llvm.lround.v8i64.v8f32(<8 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %45 = call <16 x i64> @llvm.lround.v16i64.v16f32(<16 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %46 = call <vscale x 1 x i64> @llvm.lround.nxv1i64.nxv1f32(<vscale x 1 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %47 = call <vscale x 2 x i64> @llvm.lround.nxv2i64.nxv2f32(<vscale x 2 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %48 = call <vscale x 4 x i64> @llvm.lround.nxv4i64.nxv4f32(<vscale x 4 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %49 = call <vscale x 8 x i64> @llvm.lround.nxv8i64.nxv8f32(<vscale x 8 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %50 = call <vscale x 16 x i64> @llvm.lround.nxv16i64.nxv16f32(<vscale x 16 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %51 = call i64 @llvm.lround.i64.f64(double poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %52 = call <2 x i64> @llvm.lround.v2i64.v2f64(<2 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %53 = call <4 x i64> @llvm.lround.v4i64.v4f64(<4 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %54 = call <8 x i64> @llvm.lround.v8i64.v8f64(<8 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %55 = call <16 x i64> @llvm.lround.v16i64.v16f64(<16 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %56 = call <vscale x 1 x i64> @llvm.lround.nxv1i64.nxv1f64(<vscale x 1 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %57 = call <vscale x 2 x i64> @llvm.lround.nxv2i64.nxv2f64(<vscale x 2 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %58 = call <vscale x 4 x i64> @llvm.lround.nxv4i64.nxv4f64(<vscale x 4 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %59 = call <vscale x 8 x i64> @llvm.lround.nxv8i64.nxv8f64(<vscale x 8 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %60 = call <vscale x 16 x i64> @llvm.lround.nxv16i64.nxv16f64(<vscale x 16 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+;
+ call i32 @llvm.lround.i32.bf16(bfloat poison)
+ call <2 x i32> @llvm.lround.v2i32.v2bf16(<2 x bfloat> poison)
+ call <4 x i32> @llvm.lround.v4i32.v4bf16(<4 x bfloat> poison)
+ call <8 x i32> @llvm.lround.v8i32.v8bf16(<8 x bfloat> poison)
+ call <16 x i32> @llvm.lround.v16i32.v16bf16(<16 x bfloat> poison)
+ call <vscale x 1 x i32> @llvm.lround.nxv1i32.nxv1bf16(<vscale x 1 x bfloat> poison)
+ call <vscale x 2 x i32> @llvm.lround.nxv2i32.nxv2bf16(<vscale x 2 x bfloat> poison)
+ call <vscale x 4 x i32> @llvm.lround.nxv4i32.nxv4bf16(<vscale x 4 x bfloat> poison)
+ call <vscale x 8 x i32> @llvm.lround.nxv8i32.nxv8bf16(<vscale x 8 x bfloat> poison)
+ call <vscale x 16 x i32> @llvm.lround.nxv16i32.nxv16bf16(<vscale x 16 x bfloat> poison)
+ call i32 @llvm.lround.i32.f32(float poison)
+ call <2 x i32> @llvm.lround.v2i32.v2f32(<2 x float> poison)
+ call <4 x i32> @llvm.lround.v4i32.v4f32(<4 x float> poison)
+ call <8 x i32> @llvm.lround.v8i32.v8f32(<8 x float> poison)
+ call <16 x i32> @llvm.lround.v16i32.v16f32(<16 x float> poison)
+ call <vscale x 1 x i32> @llvm.lround.nxv1i32.nxv1f32(<vscale x 1 x float> poison)
+ call <vscale x 2 x i32> @llvm.lround.nxv2i32.nxv2f32(<vscale x 2 x float> poison)
+ call <vscale x 4 x i32> @llvm.lround.nxv4i32.nxv4f32(<vscale x 4 x float> poison)
+ call <vscale x 8 x i32> @llvm.lround.nxv8i32.nxv8f32(<vscale x 8 x float> poison)
+ call <vscale x 16 x i32> @llvm.lround.nxv16i32.nxv16f32(<vscale x 16 x float> poison)
+ call i32 @llvm.lround.i32.f64(double poison)
+ call <2 x i32> @llvm.lround.v2i32.v2f64(<2 x double> poison)
+ call <4 x i32> @llvm.lround.v4i32.v4f64(<4 x double> poison)
+ call <8 x i32> @llvm.lround.v8i32.v8f64(<8 x double> poison)
+ call <16 x i32> @llvm.lround.v16i32.v16f64(<16 x double> poison)
+ call <vscale x 1 x i32> @llvm.lround.nxv1i32.nxv1f64(<vscale x 1 x double> poison)
+ call <vscale x 2 x i32> @llvm.lround.nxv2i32.nxv2f64(<vscale x 2 x double> poison)
+ call <vscale x 4 x i32> @llvm.lround.nxv4i32.nxv4f64(<vscale x 4 x double> poison)
+ call <vscale x 8 x i32> @llvm.lround.nxv8i32.nxv8f64(<vscale x 8 x double> poison)
+ call <vscale x 16 x i32> @llvm.lround.nxv16i32.nxv16f64(<vscale x 16 x double> poison)
+ call i64 @llvm.lround.i64.bf16(bfloat poison)
+ call <2 x i64> @llvm.lround.v2i64.v2bf16(<2 x bfloat> poison)
+ call <4 x i64> @llvm.lround.v4i64.v4bf16(<4 x bfloat> poison)
+ call <8 x i64> @llvm.lround.v8i64.v8bf16(<8 x bfloat> poison)
+ call <16 x i64> @llvm.lround.v16i64.v16bf16(<16 x bfloat> poison)
+ call <vscale x 1 x i64> @llvm.lround.nxv1i64.nxv1bf16(<vscale x 1 x bfloat> poison)
+ call <vscale x 2 x i64> @llvm.lround.nxv2i64.nxv2bf16(<vscale x 2 x bfloat> poison)
+ call <vscale x 4 x i64> @llvm.lround.nxv4i64.nxv4bf16(<vscale x 4 x bfloat> poison)
+ call <vscale x 8 x i64> @llvm.lround.nxv8i64.nxv8bf16(<vscale x 8 x bfloat> poison)
+ call <vscale x 16 x i64> @llvm.lround.nxv16i64.nxv16bf16(<vscale x 16 x bfloat> poison)
+ call i64 @llvm.lround.i64.f32(float poison)
+ call <2 x i64> @llvm.lround.v2i64.v2f32(<2 x float> poison)
+ call <4 x i64> @llvm.lround.v4i64.v4f32(<4 x float> poison)
+ call <8 x i64> @llvm.lround.v8i64.v8f32(<8 x float> poison)
+ call <16 x i64> @llvm.lround.v16i64.v16f32(<16 x float> poison)
+ call <vscale x 1 x i64> @llvm.lround.nxv1i64.nxv1f32(<vscale x 1 x float> poison)
+ call <vscale x 2 x i64> @llvm.lround.nxv2i64.nxv2f32(<vscale x 2 x float> poison)
+ call <vscale x 4 x i64> @llvm.lround.nxv4i64.nxv4f32(<vscale x 4 x float> poison)
+ call <vscale x 8 x i64> @llvm.lround.nxv8i64.nxv8f32(<vscale x 8 x float> poison)
+ call <vscale x 16 x i64> @llvm.lround.nxv16i64.nxv16f32(<vscale x 16 x float> poison)
+ call i64 @llvm.lround.i64.f64(double poison)
+ call <2 x i64> @llvm.lround.v2i64.v2f64(<2 x double> poison)
+ call <4 x i64> @llvm.lround.v4i64.v4f64(<4 x double> poison)
+ call <8 x i64> @llvm.lround.v8i64.v8f64(<8 x double> poison)
+ call <16 x i64> @llvm.lround.v16i64.v16f64(<16 x double> poison)
+ call <vscale x 1 x i64> @llvm.lround.nxv1i64.nxv1f64(<vscale x 1 x double> poison)
+ call <vscale x 2 x i64> @llvm.lround.nxv2i64.nxv2f64(<vscale x 2 x double> poison)
+ call <vscale x 4 x i64> @llvm.lround.nxv4i64.nxv4f64(<vscale x 4 x double> poison)
+ call <vscale x 8 x i64> @llvm.lround.nxv8i64.nxv8f64(<vscale x 8 x double> poison)
+ call <vscale x 16 x i64> @llvm.lround.nxv16i64.nxv16f64(<vscale x 16 x double> poison)
+ ret void
+}
+
+define void @llround() {
+; CHECK-LABEL: 'llround'
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call i64 @llvm.llround.i64.bf16(bfloat poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %2 = call <2 x i64> @llvm.llround.v2i64.v2bf16(<2 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %3 = call <4 x i64> @llvm.llround.v4i64.v4bf16(<4 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %4 = call <8 x i64> @llvm.llround.v8i64.v8bf16(<8 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %5 = call <16 x i64> @llvm.llround.v16i64.v16bf16(<16 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x i64> @llvm.llround.nxv1i64.nxv1bf16(<vscale x 1 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x i64> @llvm.llround.nxv2i64.nxv2bf16(<vscale x 2 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x i64> @llvm.llround.nxv4i64.nxv4bf16(<vscale x 4 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x i64> @llvm.llround.nxv8i64.nxv8bf16(<vscale x 8 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 16 x i64> @llvm.llround.nxv16i64.nxv16bf16(<vscale x 16 x bfloat> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %11 = call i64 @llvm.llround.i64.f32(float poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %12 = call <2 x i64> @llvm.llround.v2i64.v2f32(<2 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %13 = call <4 x i64> @llvm.llround.v4i64.v4f32(<4 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %14 = call <8 x i64> @llvm.llround.v8i64.v8f32(<8 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %15 = call <16 x i64> @llvm.llround.v16i64.v16f32(<16 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %16 = call <vscale x 1 x i64> @llvm.llround.nxv1i64.nxv1f32(<vscale x 1 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %17 = call <vscale x 2 x i64> @llvm.llround.nxv2i64.nxv2f32(<vscale x 2 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %18 = call <vscale x 4 x i64> @llvm.llround.nxv4i64.nxv4f32(<vscale x 4 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %19 = call <vscale x 8 x i64> @llvm.llround.nxv8i64.nxv8f32(<vscale x 8 x float> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %20 = call <vscale x 16 x i64> @llvm.llround.nxv16i64.nxv16f32(<vscale x 16 x float> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %21 = call i64 @llvm.llround.i64.f64(double poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %22 = call <2 x i64> @llvm.llround.v2i64.v2f64(<2 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %23 = call <4 x i64> @llvm.llround.v4i64.v4f64(<4 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %24 = call <8 x i64> @llvm.llround.v8i64.v8f64(<8 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 47 for instruction: %25 = call <16 x i64> @llvm.llround.v16i64.v16f64(<16 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %26 = call <vscale x 1 x i64> @llvm.llround.nxv1i64.nxv1f64(<vscale x 1 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %27 = call <vscale x 2 x i64> @llvm.llround.nxv2i64.nxv2f64(<vscale x 2 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %28 = call <vscale x 4 x i64> @llvm.llround.nxv4i64.nxv4f64(<vscale x 4 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %29 = call <vscale x 8 x i64> @llvm.llround.nxv8i64.nxv8f64(<vscale x 8 x double> poison)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %30 = call <vscale x 16 x i64> @llvm.llround.nxv16i64.nxv16f64(<vscale x 16 x double> poison)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+;
+ call i64 @llvm.llround.i64.bf16(bfloat poison)
+ call <2 x i64> @llvm.llround.v2i64.v2bf16(<2 x bfloat> poison)
+ call <4 x i64> @llvm.llround.v4i64.v4bf16(<4 x bfloat> poison)
+ call <8 x i64> @llvm.llround.v8i64.v8bf16(<8 x bfloat> poison)
+ call <16 x i64> @llvm.llround.v16i64.v16bf16(<16 x bfloat> poison)
+ call <vscale x 1 x i64> @llvm.llround.nxv1i64.nxv1bf16(<vscale x 1 x bfloat> poison)
+ call <vscale x 2 x i64> @llvm.llround.nxv2i64.nxv2bf16(<vscale x 2 x bfloat> poison)
+ call <vscale x 4 x i64> @llvm.llround.nxv4i64.nxv4bf16(<vscale x 4 x bfloat> poison)
+ call <vscale x 8 x i64> @llvm.llround.nxv8i64.nxv8bf16(<vscale x 8 x bfloat> poison)
+ call <vscale x 16 x i64> @llvm.llround.nxv16i64.nxv16bf16(<vscale x 16 x bfloat> poison)
+ call i64 @llvm.llround.i64.f32(float poison)
+ call <2 x i64> @llvm.llround.v2i64.v2f32(<2 x float> poison)
+ call <4 x i64> @llvm.llround.v4i64.v4f32(<4 x float> poison)
+ call <8 x i64> @llvm.llround.v8i64.v8f32(<8 x float> poison)
+ call <16 x i64> @llvm.llround.v16i64.v16f32(<16 x float> poison)
+ call <vscale x 1 x i64> @llvm.llround.nxv1i64.nxv1f32(<vscale x 1 x float> poison)
+ call <vscale x 2 x i64> @llvm.llround.nxv2i64.nxv2f32(<vscale x 2 x float> poison)
+ call <vscale x 4 x i64> @llvm.llround.nxv4i64.nxv4f32(<vscale x 4 x float> poison)
+ call <vscale x 8 x i64> @llvm.llround.nxv8i64.nxv8f32(<vscale x 8 x float> poison)
+ call <vscale x 16 x i64> @llvm.llround.nxv16i64.nxv16f32(<vscale x 16 x float> poison)
+ call i64 @llvm.llround.i64.f64(double poison)
+ call <2 x i64> @llvm.llround.v2i64.v2f64(<2 x double> poison)
+ call <4 x i64> @llvm.llround.v4i64.v4f64(<4 x double> poison)
+ call <8 x i64> @llvm.llround.v8i64.v8f64(<8 x double> poison)
+ call <16 x i64> @llvm.llround.v16i64.v16f64(<16 x double> poison)
+ call <vscale x 1 x i64> @llvm.llround.nxv1i64.nxv1f64(<vscale x 1 x double> poison)
+ call <vscale x 2 x i64> @llvm.llround.nxv2i64.nxv2f64(<vscale x 2 x double> poison)
+ call <vscale x 4 x i64> @llvm.llround.nxv4i64.nxv4f64(<vscale x 4 x double> poison)
+ call <vscale x 8 x i64> @llvm.llround.nxv8i64.nxv8f64(<vscale x 8 x double> poison)
+ call <vscale x 16 x i64> @llvm.llround.nxv16i64.nxv16f64(<vscale x 16 x double> poison)
+ ret void
+}
+
define void @nearbyint() {
; CHECK-LABEL: 'nearbyint'
; CHECK-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %1 = call bfloat @llvm.nearbyint.bf16(bfloat undef)
More information about the llvm-commits
mailing list