From 0bf75d85419e87d3cb0352f7ebc8ef252869e9f6 Mon Sep 17 00:00:00 2001 From: "Wang, Xin10" Date: Thu, 30 Mar 2023 10:33:57 +0800 Subject: [PATCH] Handle the unexpected inputs for pass HardwareLoops For a function TryConvertLoop in pass HardwareLoops, wrong input arguments will lead to crash. There will be 3 cases. In line 342, compiler want to get something from HWLoopInfo.CountType, which depends on if argument Bitwidth is given, if not, will crash. In Function isHardwareLoopCandidate, it dereference CountType too. In Function InsertLoopDec, it dereference LoopDecrement. They all could lead to crash. This patch add condition to this pass, when we meet unexpected inputs then skip the pass. Reviewed By: samparker, fhahn Differential Revision: https://reviews.llvm.org/D146277 --- llvm/include/llvm/Analysis/TargetTransformInfo.h | 2 +- llvm/lib/Analysis/TargetTransformInfo.cpp | 8 +++++ .../Transforms/HardwareLoops/unexpected-inputs.ll | 37 ++++++++++++++++++++++ 3 files changed, 46 insertions(+), 1 deletion(-) create mode 100644 llvm/test/Transforms/HardwareLoops/unexpected-inputs.ll diff --git a/llvm/include/llvm/Analysis/TargetTransformInfo.h b/llvm/include/llvm/Analysis/TargetTransformInfo.h index 3ce5f5c..c3e50ad 100644 --- a/llvm/include/llvm/Analysis/TargetTransformInfo.h +++ b/llvm/include/llvm/Analysis/TargetTransformInfo.h @@ -95,7 +95,7 @@ struct MemIntrinsicInfo { /// Attributes of a target dependent hardware loop. struct HardwareLoopInfo { HardwareLoopInfo() = delete; - HardwareLoopInfo(Loop *L) : L(L) {} + HardwareLoopInfo(Loop *L); Loop *L = nullptr; BasicBlock *ExitBlock = nullptr; BranchInst *ExitBranch = nullptr; diff --git a/llvm/lib/Analysis/TargetTransformInfo.cpp b/llvm/lib/Analysis/TargetTransformInfo.cpp index b536577..cdeffcb 100644 --- a/llvm/lib/Analysis/TargetTransformInfo.cpp +++ b/llvm/lib/Analysis/TargetTransformInfo.cpp @@ -108,6 +108,14 @@ IntrinsicCostAttributes::IntrinsicCostAttributes(Intrinsic::ID Id, Type *RTy, Arguments.insert(Arguments.begin(), Args.begin(), Args.end()); } +HardwareLoopInfo::HardwareLoopInfo(Loop *L) : L(L) { + // Match default options: + // - hardware-loop-counter-bitwidth = 32 + // - hardware-loop-decrement = 1 + CountType = Type::getInt32Ty(L->getHeader()->getContext()); + LoopDecrement = ConstantInt::get(CountType, 1); +} + bool HardwareLoopInfo::isHardwareLoopCandidate(ScalarEvolution &SE, LoopInfo &LI, DominatorTree &DT, bool ForceNestedLoop, diff --git a/llvm/test/Transforms/HardwareLoops/unexpected-inputs.ll b/llvm/test/Transforms/HardwareLoops/unexpected-inputs.ll new file mode 100644 index 0000000..86d0213 --- /dev/null +++ b/llvm/test/Transforms/HardwareLoops/unexpected-inputs.ll @@ -0,0 +1,37 @@ +; RUN: opt -passes='hardware-loops' -S %s -o - | FileCheck %s --check-prefix=CHECK +; RUN: opt -passes='hardware-loops' -S %s -o - | FileCheck %s --check-prefix=CHECK +; RUN: opt -passes='hardware-loops' -S %s -o - | FileCheck %s --check-prefix=CHECK + +define void @while_lt(i32 %i, i32 %N, ptr nocapture %A) { +; CHECK-LABEL: @while_lt( +; CHECK-NEXT: entry: +; CHECK-NEXT: [[CMP4:%.*]] = icmp ult i32 [[I:%.*]], [[N:%.*]] +; CHECK-NEXT: br i1 [[CMP4]], label [[WHILE_BODY_PREHEADER:%.*]], label [[WHILE_END:%.*]] +; CHECK: while.body.preheader: +; CHECK-NEXT: [[TMP0:%.*]] = sub i32 [[N]], [[I]] +; CHECK-NEXT: call void @llvm.set.loop.iterations.i32(i32 [[TMP0]]) +; CHECK-NEXT: br label [[WHILE_BODY:%.*]] +; CHECK: while.body: +; CHECK-NEXT: [[I_ADDR_05:%.*]] = phi i32 [ [[INC:%.*]], [[WHILE_BODY]] ], [ [[I]], [[WHILE_BODY_PREHEADER]] ] +; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[A:%.*]], i32 [[I_ADDR_05]] +; CHECK-NEXT: store i32 [[I_ADDR_05]], ptr [[ARRAYIDX]], align 4 +; CHECK-NEXT: [[INC]] = add nuw i32 [[I_ADDR_05]], 1 +; CHECK-NEXT: [[TMP1:%.*]] = call i1 @llvm.loop.decrement.i32(i32 1) +; CHECK-NEXT: br i1 [[TMP1]], label [[WHILE_BODY]], label [[WHILE_END]] +; CHECK: while.end: +; CHECK-NEXT: ret void +entry: + %cmp4 = icmp ult i32 %i, %N + br i1 %cmp4, label %while.body, label %while.end + +while.body: + %i.addr.05 = phi i32 [ %inc, %while.body ], [ %i, %entry ] + %arrayidx = getelementptr inbounds i32, ptr %A, i32 %i.addr.05 + store i32 %i.addr.05, ptr %arrayidx, align 4 + %inc = add nuw i32 %i.addr.05, 1 + %exitcond = icmp eq i32 %inc, %N + br i1 %exitcond, label %while.end, label %while.body + +while.end: + ret void +} -- 2.7.4