Only promote args when function attributes are compatible

Summary:
Check to make sure that the caller and the callee have compatible
function arguments before promoting arguments.  This uses the same
TargetTransformInfo queries that are used to determine if attributes
are compatible for inlining.

The goal here is to avoid breaking ABI when a called function's ABI
depends on a target feature that is not enabled in the caller.

This is a very conservative fix for PR37358.  Ideally we would have a more
sophisticated check for ABI compatiblity rather than checking if the
attributes are compatible for inlining.

Reviewers: echristo, chandlerc, eli.friedman, craig.topper

Reviewed By: echristo, chandlerc

Subscribers: nikic, xbolva00, rkruppe, alexcrichton, llvm-commits

Differential Revision: https://reviews.llvm.org/D53554

llvm-svn: 351296

Author: tstellar

llvm/include/llvm/Analysis/TargetTransformInfo.h

@@ -934,6 +934,14 @@ class TargetTransformInfo {
   bool areInlineCompatible(const Function *Caller,
                            const Function *Callee) const;
 
+  /// \returns True if the caller and callee agree on how \p Args will be passed
+  /// to the callee.
+  /// \param[out] Args The list of compatible arguments.  The implementation may
+  /// filter out any incompatible args from this list.
+  bool areFunctionArgsABICompatible(const Function *Caller,
+                                    const Function *Callee,
+                                    SmallPtrSetImpl<Argument *> &Args) const;
+
   /// The type of load/store indexing.
   enum MemIndexedMode {
     MIM_Unindexed,  ///< No indexing.
@@ -1179,6 +1187,9 @@ class TargetTransformInfo::Concept {
       unsigned RemainingBytes, unsigned SrcAlign, unsigned DestAlign) const = 0;
   virtual bool areInlineCompatible(const Function *Caller,
                                    const Function *Callee) const = 0;
+  virtual bool
+  areFunctionArgsABICompatible(const Function *Caller, const Function *Callee,
+                               SmallPtrSetImpl<Argument *> &Args) const = 0;
   virtual bool isIndexedLoadLegal(MemIndexedMode Mode, Type *Ty) const = 0;
   virtual bool isIndexedStoreLegal(MemIndexedMode Mode,Type *Ty) const = 0;
   virtual unsigned getLoadStoreVecRegBitWidth(unsigned AddrSpace) const = 0;
@@ -1557,6 +1568,11 @@ class TargetTransformInfo::Model final : public TargetTransformInfo::Concept {
                            const Function *Callee) const override {
     return Impl.areInlineCompatible(Caller, Callee);
   }
+  bool areFunctionArgsABICompatible(
+      const Function *Caller, const Function *Callee,
+      SmallPtrSetImpl<Argument *> &Args) const override {
+    return Impl.areFunctionArgsABICompatible(Caller, Callee, Args);
+  }
   bool isIndexedLoadLegal(MemIndexedMode Mode, Type *Ty) const override {
     return Impl.isIndexedLoadLegal(Mode, Ty, getDataLayout());
   }

llvm/include/llvm/Analysis/TargetTransformInfoImpl.h

@@ -526,6 +526,14 @@ class TargetTransformInfoImplBase {
             Callee->getFnAttribute("target-features"));
   }
 
+  bool areFunctionArgsABICompatible(const Function *Caller, const Function *Callee,
+                                    SmallPtrSetImpl<Argument *> &Args) const {
+    return (Caller->getFnAttribute("target-cpu") ==
+            Callee->getFnAttribute("target-cpu")) &&
+           (Caller->getFnAttribute("target-features") ==
+            Callee->getFnAttribute("target-features"));
+  }
+
   bool isIndexedLoadLegal(TTI::MemIndexedMode Mode, Type *Ty,
                           const DataLayout &DL) const {
     return false;

llvm/lib/Analysis/TargetTransformInfo.cpp

@@ -625,6 +625,12 @@ bool TargetTransformInfo::areInlineCompatible(const Function *Caller,
   return TTIImpl->areInlineCompatible(Caller, Callee);
 }
 
+bool TargetTransformInfo::areFunctionArgsABICompatible(
+    const Function *Caller, const Function *Callee,
+    SmallPtrSetImpl<Argument *> &Args) const {
+  return TTIImpl->areFunctionArgsABICompatible(Caller, Callee, Args);
+}
+
 bool TargetTransformInfo::isIndexedLoadLegal(MemIndexedMode Mode,
                                              Type *Ty) const {
   return TTIImpl->isIndexedLoadLegal(Mode, Ty);

llvm/lib/Transforms/IPO/ArgumentPromotion.cpp

@@ -49,6 +49,7 @@
 #include "llvm/Analysis/Loads.h"
 #include "llvm/Analysis/MemoryLocation.h"
 #include "llvm/Analysis/TargetLibraryInfo.h"
+#include "llvm/Analysis/TargetTransformInfo.h"
 #include "llvm/IR/Argument.h"
 #include "llvm/IR/Attributes.h"
 #include "llvm/IR/BasicBlock.h"
@@ -809,6 +810,21 @@ static bool canPaddingBeAccessed(Argument *arg) {
   return false;
 }
 
+static bool areFunctionArgsABICompatible(
+    const Function &F, const TargetTransformInfo &TTI,
+    SmallPtrSetImpl<Argument *> &ArgsToPromote,
+    SmallPtrSetImpl<Argument *> &ByValArgsToTransform) {
+  for (const Use &U : F.uses()) {
+    CallSite CS(U.getUser());
+    const Function *Caller = CS.getCaller();
+    const Function *Callee = CS.getCalledFunction();
+    if (!TTI.areFunctionArgsABICompatible(Caller, Callee, ArgsToPromote) ||
+        !TTI.areFunctionArgsABICompatible(Caller, Callee, ByValArgsToTransform))
+      return false;
+  }
+  return true;
+}
+
 /// PromoteArguments - This method checks the specified function to see if there
 /// are any promotable arguments and if it is safe to promote the function (for
 /// example, all callers are direct).  If safe to promote some arguments, it
@@ -817,7 +833,8 @@ static Function *
 promoteArguments(Function *F, function_ref<AAResults &(Function &F)> AARGetter,
                  unsigned MaxElements,
                  Optional<function_ref<void(CallSite OldCS, CallSite NewCS)>>
-                     ReplaceCallSite) {
+                     ReplaceCallSite,
+                 const TargetTransformInfo &TTI) {
   // Don't perform argument promotion for naked functions; otherwise we can end
   // up removing parameters that are seemingly 'not used' as they are referred
   // to in the assembly.
@@ -846,7 +863,7 @@ promoteArguments(Function *F, function_ref<AAResults &(Function &F)> AARGetter,
 
   // Second check: make sure that all callers are direct callers.  We can't
   // transform functions that have indirect callers.  Also see if the function
-  // is self-recursive.
+  // is self-recursive and check that target features are compatible.
   bool isSelfRecursive = false;
   for (Use &U : F->uses()) {
     CallSite CS(U.getUser());
@@ -955,6 +972,10 @@ promoteArguments(Function *F, function_ref<AAResults &(Function &F)> AARGetter,
   if (ArgsToPromote.empty() && ByValArgsToTransform.empty())
     return nullptr;
 
+  if (!areFunctionArgsABICompatible(*F, TTI, ArgsToPromote,
+                                    ByValArgsToTransform))
+    return nullptr;
+
   return doPromotion(F, ArgsToPromote, ByValArgsToTransform, ReplaceCallSite);
 }
 
@@ -980,7 +1001,9 @@ PreservedAnalyses ArgumentPromotionPass::run(LazyCallGraph::SCC &C,
         return FAM.getResult<AAManager>(F);
       };
 
-      Function *NewF = promoteArguments(&OldF, AARGetter, MaxElements, None);
+      const TargetTransformInfo &TTI = FAM.getResult<TargetIRAnalysis>(OldF);
+      Function *NewF =
+          promoteArguments(&OldF, AARGetter, MaxElements, None, TTI);
       if (!NewF)
         continue;
       LocalChange = true;
@@ -1018,6 +1041,7 @@ struct ArgPromotion : public CallGraphSCCPass {
   void getAnalysisUsage(AnalysisUsage &AU) const override {
     AU.addRequired<AssumptionCacheTracker>();
     AU.addRequired<TargetLibraryInfoWrapperPass>();
+    AU.addRequired<TargetTransformInfoWrapperPass>();
     getAAResultsAnalysisUsage(AU);
     CallGraphSCCPass::getAnalysisUsage(AU);
   }
@@ -1043,6 +1067,7 @@ INITIALIZE_PASS_BEGIN(ArgPromotion, "argpromotion",
 INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
 INITIALIZE_PASS_DEPENDENCY(CallGraphWrapperPass)
 INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
 INITIALIZE_PASS_END(ArgPromotion, "argpromotion",
                     "Promote 'by reference' arguments to scalars", false, false)
 
@@ -1079,8 +1104,10 @@ bool ArgPromotion::runOnSCC(CallGraphSCC &SCC) {
         CallerNode->replaceCallEdge(OldCS, NewCS, NewCalleeNode);
       };
 
+      const TargetTransformInfo &TTI =
+          getAnalysis<TargetTransformInfoWrapperPass>().getTTI(*OldF);
       if (Function *NewF = promoteArguments(OldF, AARGetter, MaxElements,
-                                            {ReplaceCallSite})) {
+                                            {ReplaceCallSite}, TTI)) {
         LocalChange = true;
 
         // Update the call graph for the newly promoted function.

llvm/test/Transforms/ArgumentPromotion/X86/attributes.ll

@@ -0,0 +1,53 @@
+; RUN: opt -S -argpromotion < %s | FileCheck %s
+; RUN: opt -S -passes=argpromotion < %s | FileCheck %s
+; Test that we only promote arguments when the caller/callee have compatible
+; function attrubtes.
+
+target triple = "x86_64-unknown-linux-gnu"
+
+; CHECK-LABEL: @no_promote_avx2(<4 x i64>* %arg, <4 x i64>* readonly %arg1)
+define internal fastcc void @no_promote_avx2(<4 x i64>* %arg, <4 x i64>* readonly %arg1) #0 {
+bb:
+  %tmp = load <4 x i64>, <4 x i64>* %arg1
+  store <4 x i64> %tmp, <4 x i64>* %arg
+  ret void
+}
+
+define void @no_promote(<4 x i64>* %arg) #1 {
+bb:
+  %tmp = alloca <4 x i64>, align 32
+  %tmp2 = alloca <4 x i64>, align 32
+  %tmp3 = bitcast <4 x i64>* %tmp to i8*
+  call void @llvm.memset.p0i8.i64(i8* align 32 %tmp3, i8 0, i64 32, i1 false)
+  call fastcc void @no_promote_avx2(<4 x i64>* %tmp2, <4 x i64>* %tmp)
+  %tmp4 = load <4 x i64>, <4 x i64>* %tmp2, align 32
+  store <4 x i64> %tmp4, <4 x i64>* %arg, align 2
+  ret void
+}
+
+; CHECK-LABEL: @promote_avx2(<4 x i64>* %arg, <4 x i64> %
+define internal fastcc void @promote_avx2(<4 x i64>* %arg, <4 x i64>* readonly %arg1) #0 {
+bb:
+  %tmp = load <4 x i64>, <4 x i64>* %arg1
+  store <4 x i64> %tmp, <4 x i64>* %arg
+  ret void
+}
+
+define void @promote(<4 x i64>* %arg) #0 {
+bb:
+  %tmp = alloca <4 x i64>, align 32
+  %tmp2 = alloca <4 x i64>, align 32
+  %tmp3 = bitcast <4 x i64>* %tmp to i8*
+  call void @llvm.memset.p0i8.i64(i8* align 32 %tmp3, i8 0, i64 32, i1 false)
+  call fastcc void @promote_avx2(<4 x i64>* %tmp2, <4 x i64>* %tmp)
+  %tmp4 = load <4 x i64>, <4 x i64>* %tmp2, align 32
+  store <4 x i64> %tmp4, <4 x i64>* %arg, align 2
+  ret void
+}
+
+; Function Attrs: argmemonly nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture writeonly, i8, i64, i1) #2
+
+attributes #0 = { inlinehint norecurse nounwind uwtable "target-features"="+avx2" }
+attributes #1 = { nounwind uwtable }
+attributes #2 = { argmemonly nounwind }