[llvm] [ConstantRange] Fix off by 1 bugs in UIToFP and SIToFP handling. (PR #86041)

Craig Topper via llvm-commits llvm-commits at lists.llvm.org
Thu Mar 21 08:25:27 PDT 2024


https://github.com/topperc updated https://github.com/llvm/llvm-project/pull/86041

>From 26ae69a78fbf234a70a0748dd11c511ce3dca5b6 Mon Sep 17 00:00:00 2001
From: Craig Topper <craig.topper at sifive.com>
Date: Wed, 20 Mar 2024 17:05:17 -0700
Subject: [PATCH 1/2] [Float2Int] Pre-commit test for ConstantRange bug. NFC

---
 llvm/test/Transforms/Float2Int/pr79158.ll | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)
 create mode 100644 llvm/test/Transforms/Float2Int/pr79158.ll

diff --git a/llvm/test/Transforms/Float2Int/pr79158.ll b/llvm/test/Transforms/Float2Int/pr79158.ll
new file mode 100644
index 00000000000000..19784efec5318b
--- /dev/null
+++ b/llvm/test/Transforms/Float2Int/pr79158.ll
@@ -0,0 +1,19 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
+; RUN: opt < %s -passes=float2int -S | FileCheck %s
+
+define i32 @pr79158(i32 %x) {
+; CHECK-LABEL: define i32 @pr79158(
+; CHECK-SAME: i32 [[X:%.*]]) {
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[CMP_I:%.*]] = icmp sgt i32 [[X]], 0
+; CHECK-NEXT:    [[TMP0:%.*]] = zext i1 [[CMP_I]] to i32
+; CHECK-NEXT:    [[MUL_I1:%.*]] = mul i32 [[TMP0]], 2147483647
+; CHECK-NEXT:    ret i32 [[MUL_I1]]
+;
+entry:
+  %cmp = icmp sgt i32 %x, 0
+  %conv = uitofp i1 %cmp to double
+  %mul = fmul double %conv, 0x41EFFFFFFFE00000
+  %conv1 = fptoui double %mul to i32
+  ret i32 %conv1
+}

>From 7a62d5f2ced559b8d0cc4ccde0dbec5bede26668 Mon Sep 17 00:00:00 2001
From: Craig Topper <craig.topper at sifive.com>
Date: Wed, 20 Mar 2024 16:36:44 -0700
Subject: [PATCH 2/2] [ConstantRange] Fix off by 1 bugs in UIToFP and SIToFP
 handling.

We were passing the min and max values of the range to the ConstantRange
constructor, but the constructor expects the upper bound to 1
more than the max value so we need to add 1.

We also need to use getNonEmpty so that passing 0, 0 to the constructor
creates a full range rather than an empty range. And passing smin, smax+1
doesn't cause an assertion.

I believe this fixes at least some of the reason #79158 was reverted.
---
 llvm/lib/IR/ConstantRange.cpp             |  4 ++--
 llvm/test/Transforms/Float2Int/pr79158.ll |  5 +++--
 llvm/unittests/IR/ConstantRangeTest.cpp   | 18 ++++++++++++++++++
 3 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/llvm/lib/IR/ConstantRange.cpp b/llvm/lib/IR/ConstantRange.cpp
index 3394a1ec8dc476..59e7a9f5eb1114 100644
--- a/llvm/lib/IR/ConstantRange.cpp
+++ b/llvm/lib/IR/ConstantRange.cpp
@@ -746,7 +746,7 @@ ConstantRange ConstantRange::castOp(Instruction::CastOps CastOp,
       Min = Min.zext(ResultBitWidth);
       Max = Max.zext(ResultBitWidth);
     }
-    return ConstantRange(std::move(Min), std::move(Max));
+    return getNonEmpty(std::move(Min), std::move(Max) + 1);
   }
   case Instruction::SIToFP: {
     // TODO: use input range if available
@@ -757,7 +757,7 @@ ConstantRange ConstantRange::castOp(Instruction::CastOps CastOp,
       SMin = SMin.sext(ResultBitWidth);
       SMax = SMax.sext(ResultBitWidth);
     }
-    return ConstantRange(std::move(SMin), std::move(SMax));
+    return getNonEmpty(std::move(SMin), std::move(SMax) + 1);
   }
   case Instruction::FPTrunc:
   case Instruction::FPExt:
diff --git a/llvm/test/Transforms/Float2Int/pr79158.ll b/llvm/test/Transforms/Float2Int/pr79158.ll
index 19784efec5318b..43c80a9dda5c10 100644
--- a/llvm/test/Transforms/Float2Int/pr79158.ll
+++ b/llvm/test/Transforms/Float2Int/pr79158.ll
@@ -6,8 +6,9 @@ define i32 @pr79158(i32 %x) {
 ; CHECK-SAME: i32 [[X:%.*]]) {
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[CMP_I:%.*]] = icmp sgt i32 [[X]], 0
-; CHECK-NEXT:    [[TMP0:%.*]] = zext i1 [[CMP_I]] to i32
-; CHECK-NEXT:    [[MUL_I1:%.*]] = mul i32 [[TMP0]], 2147483647
+; CHECK-NEXT:    [[TMP0:%.*]] = zext i1 [[CMP_I]] to i64
+; CHECK-NEXT:    [[MUL_I2:%.*]] = mul i64 [[TMP0]], 4294967295
+; CHECK-NEXT:    [[MUL_I1:%.*]] = trunc i64 [[MUL_I2]] to i32
 ; CHECK-NEXT:    ret i32 [[MUL_I1]]
 ;
 entry:
diff --git a/llvm/unittests/IR/ConstantRangeTest.cpp b/llvm/unittests/IR/ConstantRangeTest.cpp
index 34a162a5514e95..8ec120d70e99f6 100644
--- a/llvm/unittests/IR/ConstantRangeTest.cpp
+++ b/llvm/unittests/IR/ConstantRangeTest.cpp
@@ -2479,6 +2479,24 @@ TEST_F(ConstantRangeTest, castOps) {
   ConstantRange IntToPtr = A.castOp(Instruction::IntToPtr, 64);
   EXPECT_EQ(64u, IntToPtr.getBitWidth());
   EXPECT_TRUE(IntToPtr.isFullSet());
+
+  ConstantRange UIToFP = A.castOp(Instruction::UIToFP, 16);
+  EXPECT_EQ(16u, UIToFP.getBitWidth());
+  EXPECT_TRUE(UIToFP.isFullSet());
+
+  ConstantRange UIToFP2 = A.castOp(Instruction::UIToFP, 64);
+  ConstantRange B(APInt(64, 0), APInt(64, 65536));
+  EXPECT_EQ(64u, UIToFP2.getBitWidth());
+  EXPECT_EQ(B, UIToFP2);
+
+  ConstantRange SIToFP = A.castOp(Instruction::SIToFP, 16);
+  EXPECT_EQ(16u, SIToFP.getBitWidth());
+  EXPECT_TRUE(SIToFP.isFullSet());
+
+  ConstantRange SIToFP2 = A.castOp(Instruction::SIToFP, 64);
+  ConstantRange C(APInt(64, -32768), APInt(64, 32768));
+  EXPECT_EQ(64u, SIToFP2.getBitWidth());
+  EXPECT_EQ(C, SIToFP2);
 }
 
 TEST_F(ConstantRangeTest, binaryAnd) {



More information about the llvm-commits mailing list