[llvm] ef7aed3 - [InstCombine] Do not fold 'and (sext (ashr X, Shift)), C' if Shift < 0
Daniel Bertalan via llvm-commits
llvm-commits at lists.llvm.org
Thu Jul 7 10:14:37 PDT 2022
Author: Daniel Bertalan
Date: 2022-07-07T19:13:55+02:00
New Revision: ef7aed3e112b47641ca2704187d3701765a831ce
URL: https://github.com/llvm/llvm-project/commit/ef7aed3e112b47641ca2704187d3701765a831ce
DIFF: https://github.com/llvm/llvm-project/commit/ef7aed3e112b47641ca2704187d3701765a831ce.diff
LOG: [InstCombine] Do not fold 'and (sext (ashr X, Shift)), C' if Shift < 0
The 'and (sext (ashr X, ShiftC)), C' --> 'lshr (sext X), ShiftC'
transformation would access out of bounds bits in APInt::getLowBitsSet
if the shift count was larger than X's bit width or if it was negative.
Fixes #56424
Added:
llvm/test/Transforms/InstCombine/pr56424.ll
Modified:
llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
Removed:
################################################################################
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp b/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
index d3578fe4a2ba9..a8f2cd79830ab 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
@@ -1808,7 +1808,8 @@ Instruction *InstCombinerImpl::visitAnd(BinaryOperator &I) {
unsigned Width = Ty->getScalarSizeInBits();
const APInt *ShiftC;
- if (match(Op0, m_OneUse(m_SExt(m_AShr(m_Value(X), m_APInt(ShiftC)))))) {
+ if (match(Op0, m_OneUse(m_SExt(m_AShr(m_Value(X), m_APInt(ShiftC))))) &&
+ ShiftC->ult(Width)) {
if (*C == APInt::getLowBitsSet(Width, Width - ShiftC->getZExtValue())) {
// We are clearing high bits that were potentially set by sext+ashr:
// and (sext (ashr X, ShiftC)), C --> lshr (sext X), ShiftC
diff --git a/llvm/test/Transforms/InstCombine/pr56424.ll b/llvm/test/Transforms/InstCombine/pr56424.ll
new file mode 100644
index 0000000000000..dda0b6d7e0be1
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/pr56424.ll
@@ -0,0 +1,27 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s --passes=instcombine -S | FileCheck %s
+
+; This would crash if we didn't check for a negative shift.
+; https://github.com/llvm/llvm-project/issues/56424
+define i64 @PR56424(i1 %cond, i32 %arg) {
+; CHECK-LABEL: @PR56424(
+; CHECK-NEXT: entry:
+; CHECK-NEXT: br i1 [[COND:%.*]], label [[IF_THEN:%.*]], label [[IF_END:%.*]]
+; CHECK: if.then:
+; CHECK-NEXT: br label [[IF_END]]
+; CHECK: if.end:
+; CHECK-NEXT: ret i64 0
+;
+entry:
+ br i1 %cond, label %if.then, label %if.end
+
+if.then:
+ %shr = ashr i32 %arg, -2
+ %sext = sext i32 %shr to i64
+ br label %if.end
+
+if.end:
+ %val = phi i64 [ %sext, %if.then ], [ 0, %entry ]
+ %and = and i64 -81, %val
+ ret i64 %and
+}
More information about the llvm-commits
mailing list