[LLVM] Maintain element type of @llvm.compiler.used/@llvm.used if it already exists

llvm/include/llvm/IR/GlobalValue.h

@@ -684,6 +684,18 @@ class GlobalValue : public Constant {
   LLVM_ABI bool canBeOmittedFromSymbolTable() const;
 };
 
+/// Adds global values to the llvm.used list.
+LLVM_ABI void appendToUsed(Module &M, ArrayRef<GlobalValue *> Values);
+
+/// Adds global values to the llvm.compiler.used list.
+LLVM_ABI void appendToCompilerUsed(Module &M, ArrayRef<GlobalValue *> Values);
+
+/// Removes global values from the llvm.used and llvm.compiler.used arrays. \p
+/// ShouldRemove should return true for any initializer field that should not be
+/// included in the replacement global.
+LLVM_ABI void removeFromUsedLists(Module &M,
+                                  function_ref<bool(Constant *)> ShouldRemove);
+
 } // end namespace llvm
 
 #endif // LLVM_IR_GLOBALVALUE_H

llvm/include/llvm/Transforms/Utils/ModuleUtils.h

@@ -96,18 +96,6 @@ getOrCreateSanitizerCtorAndInitFunctions(
 /// the list of public globals in the module.
 LLVM_ABI bool nameUnamedGlobals(Module &M);
 
-/// Adds global values to the llvm.used list.
-LLVM_ABI void appendToUsed(Module &M, ArrayRef<GlobalValue *> Values);
-
-/// Adds global values to the llvm.compiler.used list.
-LLVM_ABI void appendToCompilerUsed(Module &M, ArrayRef<GlobalValue *> Values);
-
-/// Removes global values from the llvm.used and llvm.compiler.used arrays. \p
-/// ShouldRemove should return true for any initializer field that should not be
-/// included in the replacement global.
-LLVM_ABI void removeFromUsedLists(Module &M,
-                                  function_ref<bool(Constant *)> ShouldRemove);
-
 /// Filter out potentially dead comdat functions where other entries keep the
 /// entire comdat group alive.
 ///

llvm/lib/Bitcode/Writer/BitcodeWriter.cpp

@@ -75,6 +75,7 @@
 #include "llvm/Support/SHA1.h"
 #include "llvm/Support/raw_ostream.h"
 #include "llvm/TargetParser/Triple.h"
+#include "llvm/Transforms/Utils/ModuleUtils.h"
 #include <algorithm>
 #include <cassert>
 #include <cstddef>
@@ -5861,28 +5862,19 @@ static const char *getSectionNameForCommandline(const Triple &T) {
   llvm_unreachable("Unimplemented ObjectFormatType");
 }
 
+static auto globalInUsedHasName(StringRef Name) {
+  return [Name](Constant *C) { return C->getName() == Name; };
+}
+
 void llvm::embedBitcodeInModule(llvm::Module &M, llvm::MemoryBufferRef Buf,
                                 bool EmbedBitcode, bool EmbedCmdline,
                                 const std::vector<uint8_t> &CmdArgs) {
-  // Save llvm.compiler.used and remove it.
-  SmallVector<Constant *, 2> UsedArray;
-  SmallVector<GlobalValue *, 4> UsedGlobals;
-  GlobalVariable *Used = collectUsedGlobalVariables(M, UsedGlobals, true);
-  Type *UsedElementType = Used ? Used->getValueType()->getArrayElementType()
-                               : PointerType::getUnqual(M.getContext());
-  for (auto *GV : UsedGlobals) {
-    if (GV->getName() != "llvm.embedded.module" &&
-        GV->getName() != "llvm.cmdline")
-      UsedArray.push_back(
-          ConstantExpr::getPointerBitCastOrAddrSpaceCast(GV, UsedElementType));
-  }
-  if (Used)
-    Used->eraseFromParent();
 
   // Embed the bitcode for the llvm module.
   std::string Data;
   ArrayRef<uint8_t> ModuleData;
   Triple T(M.getTargetTriple());
+  SmallVector<GlobalValue *, 2> NewGlobals;
 
   if (EmbedBitcode) {
     if (Buf.getBufferSize() == 0 ||
@@ -5901,23 +5893,23 @@ void llvm::embedBitcodeInModule(llvm::Module &M, llvm::MemoryBufferRef Buf,
   }
   llvm::Constant *ModuleConstant =
       llvm::ConstantDataArray::get(M.getContext(), ModuleData);
-  llvm::GlobalVariable *GV = new llvm::GlobalVariable(
+  llvm::GlobalVariable *EmbeddedModule = new llvm::GlobalVariable(
       M, ModuleConstant->getType(), true, llvm::GlobalValue::PrivateLinkage,
       ModuleConstant);
-  GV->setSection(getSectionNameForBitcode(T));
+  EmbeddedModule->setSection(getSectionNameForBitcode(T));
   // Set alignment to 1 to prevent padding between two contributions from input
   // sections after linking.
-  GV->setAlignment(Align(1));
-  UsedArray.push_back(
-      ConstantExpr::getPointerBitCastOrAddrSpaceCast(GV, UsedElementType));
+  EmbeddedModule->setAlignment(Align(1));
+  NewGlobals.push_back(EmbeddedModule);
   if (llvm::GlobalVariable *Old =
           M.getGlobalVariable("llvm.embedded.module", true)) {
+    removeFromUsedLists(M, globalInUsedHasName("llvm.embedded.module"));
     assert(Old->hasZeroLiveUses() &&
            "llvm.embedded.module can only be used once in llvm.compiler.used");
-    GV->takeName(Old);
+    EmbeddedModule->takeName(Old);
     Old->eraseFromParent();
   } else {
-    GV->setName("llvm.embedded.module");
+    EmbeddedModule->setName("llvm.embedded.module");
   }
 
   // Skip if only bitcode needs to be embedded.
@@ -5927,30 +5919,21 @@ void llvm::embedBitcodeInModule(llvm::Module &M, llvm::MemoryBufferRef Buf,
                               CmdArgs.size());
     llvm::Constant *CmdConstant =
         llvm::ConstantDataArray::get(M.getContext(), CmdData);
-    GV = new llvm::GlobalVariable(M, CmdConstant->getType(), true,
-                                  llvm::GlobalValue::PrivateLinkage,
-                                  CmdConstant);
-    GV->setSection(getSectionNameForCommandline(T));
-    GV->setAlignment(Align(1));
-    UsedArray.push_back(
-        ConstantExpr::getPointerBitCastOrAddrSpaceCast(GV, UsedElementType));
+    GlobalVariable *CmdLine = new llvm::GlobalVariable(
+        M, CmdConstant->getType(), true, llvm::GlobalValue::PrivateLinkage,
+        CmdConstant);
+    CmdLine->setSection(getSectionNameForCommandline(T));
+    CmdLine->setAlignment(Align(1));
     if (llvm::GlobalVariable *Old = M.getGlobalVariable("llvm.cmdline", true)) {
+      removeFromUsedLists(M, globalInUsedHasName("llvm.cmdline"));
       assert(Old->hasZeroLiveUses() &&
              "llvm.cmdline can only be used once in llvm.compiler.used");
-      GV->takeName(Old);
+      CmdLine->takeName(Old);
       Old->eraseFromParent();
     } else {
-      GV->setName("llvm.cmdline");
+      CmdLine->setName("llvm.cmdline");
     }
+    NewGlobals.push_back(CmdLine);
+    appendToCompilerUsed(M, NewGlobals);
   }
-
-  if (UsedArray.empty())
-    return;
-
-  // Recreate llvm.compiler.used.
-  ArrayType *ATy = ArrayType::get(UsedElementType, UsedArray.size());
-  auto *NewUsed = new GlobalVariable(
-      M, ATy, false, llvm::GlobalValue::AppendingLinkage,
-      llvm::ConstantArray::get(ATy, UsedArray), "llvm.compiler.used");
-  NewUsed->setSection("llvm.metadata");
 }

llvm/lib/CodeGen/JMCInstrumenter.cpp

@@ -36,7 +36,6 @@
 #include "llvm/Pass.h"
 #include "llvm/Support/DJB.h"
 #include "llvm/Support/Path.h"
-#include "llvm/Transforms/Utils/ModuleUtils.h"
 
 using namespace llvm;
 

llvm/lib/IR/Globals.cpp

@@ -12,6 +12,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "LLVMContextImpl.h"
+#include "llvm/ADT/SetVector.h"
 #include "llvm/IR/ConstantRange.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DerivedTypes.h"
@@ -671,3 +672,80 @@ void GlobalIFunc::applyAlongResolverPath(
   DenseSet<const GlobalAlias *> Aliases;
   findBaseObject(getResolver(), Aliases, Op);
 }
+
+static void collectUsedGlobals(GlobalVariable *GV,
+                               SmallSetVector<Constant *, 16> &Init) {
+  if (!GV || !GV->hasInitializer())
+    return;
+
+  auto *CA = cast<ConstantArray>(GV->getInitializer());
+  for (Use &Op : CA->operands())
+    Init.insert(cast<Constant>(Op));
+}
+
+static void appendToUsedList(Module &M, StringRef Name,
+                             ArrayRef<GlobalValue *> Values) {
+  GlobalVariable *GV = M.getGlobalVariable(Name);
+
+  SmallSetVector<Constant *, 16> Init;
+  collectUsedGlobals(GV, Init);
+  Type *ArrayEltTy = GV ? GV->getValueType()->getArrayElementType()
+                        : PointerType::getUnqual(M.getContext());
+  if (GV)
+    GV->eraseFromParent();
+
+  for (auto *V : Values)
+    Init.insert(ConstantExpr::getPointerBitCastOrAddrSpaceCast(V, ArrayEltTy));
+
+  if (Init.empty())
+    return;
+
+  ArrayType *ATy = ArrayType::get(ArrayEltTy, Init.size());
+  GV = new GlobalVariable(M, ATy, false, GlobalValue::AppendingLinkage,
+                          ConstantArray::get(ATy, Init.getArrayRef()), Name);
+  GV->setSection("llvm.metadata");
+}
+
+void llvm::appendToUsed(Module &M, ArrayRef<GlobalValue *> Values) {
+  appendToUsedList(M, "llvm.used", Values);
+}
+
+void llvm::appendToCompilerUsed(Module &M, ArrayRef<GlobalValue *> Values) {
+  appendToUsedList(M, "llvm.compiler.used", Values);
+}
+
+static void removeFromUsedList(Module &M, StringRef Name,
+                               function_ref<bool(Constant *)> ShouldRemove) {
+  GlobalVariable *GV = M.getNamedGlobal(Name);
+  if (!GV)
+    return;
+
+  SmallSetVector<Constant *, 16> Init;
+  collectUsedGlobals(GV, Init);
+
+  Type *ArrayEltTy = cast<ArrayType>(GV->getValueType())->getElementType();
+
+  SmallVector<Constant *, 16> NewInit;
+  for (Constant *MaybeRemoved : Init) {
+    if (!ShouldRemove(MaybeRemoved->stripPointerCasts()))
+      NewInit.push_back(MaybeRemoved);
+  }
+
+  if (!NewInit.empty()) {
+    ArrayType *ATy = ArrayType::get(ArrayEltTy, NewInit.size());
+    GlobalVariable *NewGV =
+        new GlobalVariable(M, ATy, false, GlobalValue::AppendingLinkage,
+                           ConstantArray::get(ATy, NewInit), "", GV,
+                           GV->getThreadLocalMode(), GV->getAddressSpace());
+    NewGV->setSection(GV->getSection());
+    NewGV->takeName(GV);
+  }
+
+  GV->eraseFromParent();
+}
+
+void llvm::removeFromUsedLists(Module &M,
+                               function_ref<bool(Constant *)> ShouldRemove) {
+  removeFromUsedList(M, "llvm.used", ShouldRemove);
+  removeFromUsedList(M, "llvm.compiler.used", ShouldRemove);
+}

llvm/lib/Target/AMDGPU/AMDGPUCtorDtorLowering.cpp

@@ -19,7 +19,6 @@
 #include "llvm/IR/Module.h"
 #include "llvm/IR/Value.h"
 #include "llvm/Pass.h"
-#include "llvm/Transforms/Utils/ModuleUtils.h"
 
 using namespace llvm;
 

llvm/lib/Target/NVPTX/NVPTXCtorDtorLowering.cpp

@@ -24,7 +24,6 @@
 #include "llvm/Pass.h"
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Support/MD5.h"
-#include "llvm/Transforms/Utils/ModuleUtils.h"
 
 using namespace llvm;
 

llvm/lib/Transforms/Utils/ModuleUtils.cpp

@@ -123,82 +123,6 @@ void llvm::transformGlobalDtors(Module &M, const GlobalCtorTransformFn &Fn) {
   transformGlobalArray("llvm.global_dtors", M, Fn);
 }
 
-static void collectUsedGlobals(GlobalVariable *GV,
-                               SmallSetVector<Constant *, 16> &Init) {
-  if (!GV || !GV->hasInitializer())
-    return;
-
-  auto *CA = cast<ConstantArray>(GV->getInitializer());
-  for (Use &Op : CA->operands())
-    Init.insert(cast<Constant>(Op));
-}
-
-static void appendToUsedList(Module &M, StringRef Name, ArrayRef<GlobalValue *> Values) {
-  GlobalVariable *GV = M.getGlobalVariable(Name);
-
-  SmallSetVector<Constant *, 16> Init;
-  collectUsedGlobals(GV, Init);
-  if (GV)
-    GV->eraseFromParent();
-
-  Type *ArrayEltTy = llvm::PointerType::getUnqual(M.getContext());
-  for (auto *V : Values)
-    Init.insert(ConstantExpr::getPointerBitCastOrAddrSpaceCast(V, ArrayEltTy));
-
-  if (Init.empty())
-    return;
-
-  ArrayType *ATy = ArrayType::get(ArrayEltTy, Init.size());
-  GV = new llvm::GlobalVariable(M, ATy, false, GlobalValue::AppendingLinkage,
-                                ConstantArray::get(ATy, Init.getArrayRef()),
-                                Name);
-  GV->setSection("llvm.metadata");
-}
-
-void llvm::appendToUsed(Module &M, ArrayRef<GlobalValue *> Values) {
-  appendToUsedList(M, "llvm.used", Values);
-}
-
-void llvm::appendToCompilerUsed(Module &M, ArrayRef<GlobalValue *> Values) {
-  appendToUsedList(M, "llvm.compiler.used", Values);
-}
-
-static void removeFromUsedList(Module &M, StringRef Name,
-                               function_ref<bool(Constant *)> ShouldRemove) {
-  GlobalVariable *GV = M.getNamedGlobal(Name);
-  if (!GV)
-    return;
-
-  SmallSetVector<Constant *, 16> Init;
-  collectUsedGlobals(GV, Init);
-
-  Type *ArrayEltTy = cast<ArrayType>(GV->getValueType())->getElementType();
-
-  SmallVector<Constant *, 16> NewInit;
-  for (Constant *MaybeRemoved : Init) {
-    if (!ShouldRemove(MaybeRemoved->stripPointerCasts()))
-      NewInit.push_back(MaybeRemoved);
-  }
-
-  if (!NewInit.empty()) {
-    ArrayType *ATy = ArrayType::get(ArrayEltTy, NewInit.size());
-    GlobalVariable *NewGV =
-        new GlobalVariable(M, ATy, false, GlobalValue::AppendingLinkage,
-                           ConstantArray::get(ATy, NewInit), "", GV,
-                           GV->getThreadLocalMode(), GV->getAddressSpace());
-    NewGV->setSection(GV->getSection());
-    NewGV->takeName(GV);
-  }
-
-  GV->eraseFromParent();
-}
-
-void llvm::removeFromUsedLists(Module &M,
-                               function_ref<bool(Constant *)> ShouldRemove) {
-  removeFromUsedList(M, "llvm.used", ShouldRemove);
-  removeFromUsedList(M, "llvm.compiler.used", ShouldRemove);
-}
-
 void llvm::setKCFIType(Module &M, Function &F, StringRef MangledType) {
   if (!M.getModuleFlag("kcfi"))
     return;

llvm/lib/Transforms/Utils/SampleProfileLoaderBaseUtil.cpp

@@ -14,7 +14,6 @@
 #include "llvm/Analysis/ProfileSummaryInfo.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/Module.h"
-#include "llvm/Transforms/Utils/ModuleUtils.h"
 
 namespace llvm {
 

llvm/unittests/IR/CMakeLists.txt

@@ -49,6 +49,7 @@ add_llvm_unittest(IRTests
   TypesTest.cpp
   UseTest.cpp
   UserTest.cpp
+  UsedGlobalTest.cpp
   ValueHandleTest.cpp
   ValueMapTest.cpp
   ValueTest.cpp