diff --git a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp --- a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp +++ b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp @@ -96,19 +96,20 @@ /// A helper routine of InstCombiner::visitMul(). /// -/// If C is a scalar/vector of known powers of 2, then this function returns -/// a new scalar/vector obtained from logBase2 of C. +/// If C is a scalar/fixed width vector of known powers of 2, then this +/// function returns a new scalar/fixed width vector obtained from logBase2 +/// of C. /// Return a null pointer otherwise. static Constant *getLogBase2(Type *Ty, Constant *C) { const APInt *IVal; if (match(C, m_APInt(IVal)) && IVal->isPowerOf2()) return ConstantInt::get(Ty, IVal->logBase2()); - if (!Ty->isVectorTy()) + if (!isa(Ty)) return nullptr; SmallVector Elts; - for (unsigned I = 0, E = cast(Ty)->getNumElements(); I != E; + for (unsigned I = 0, E = cast(Ty)->getNumElements(); I != E; ++I) { Constant *Elt = C->getAggregateElement(I); if (!Elt) diff --git a/llvm/test/Transforms/InstCombine/udiv-pow2-vscale.ll b/llvm/test/Transforms/InstCombine/udiv-pow2-vscale.ll new file mode 100644 --- /dev/null +++ b/llvm/test/Transforms/InstCombine/udiv-pow2-vscale.ll @@ -0,0 +1,27 @@ +; RUN: opt -instcombine -S < %s | FileCheck %s + +; This vscale udiv with a power-of-2 spalt on the rhs should not crash opt + +; CHECK: define @udiv_pow2_vscale() +define @udiv_pow2_vscale( %lhs) { + %splatter = insertelement undef, i32 2, i32 0 + %rhs = shufflevector %splatter, + undef, + zeroinitializer + %res = udiv %lhs, %rhs + ret %res +} + +; This fixed width udiv with a power-of-2 splat on the rhs should also not +; crash, and instcombine should eliminate the udiv + +; CHECK-LABEL: define @udiv_pow2_fixed() +; CHECK-NOT: udiv +define <2 x i32> @udiv_pow2_fixed(<2 x i32> %lhs) { + %splatter = insertelement <2 x i32> undef, i32 2, i32 0 + %rhs = shufflevector <2 x i32> %splatter, + <2 x i32> undef, + <2 x i32> zeroinitializer + %res = udiv <2 x i32> %lhs, %rhs + ret <2 x i32> %res +}