[Alignment][NFC] Use Align with CreateElementUnorderedAtomicMemCpy

[gchatelet]

Summary

This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790

[gchatelet]

• remove wrongly added file

[merge_guards_bot]

Unit tests: pass. 62017 tests passed, 0 failed and 783 were skipped.

clang-tidy: unknown.

clang-format: fail. Please format your changes with clang-format by running git-clang-format HEAD^ or applying this patch.

Build artifacts: diff.json, clang-format.patch, CMakeCache.txt, console-log.txt, test-results.xml

[gchatelet]

• Adding back unrelated file to please the Merge bot

[merge_guards_bot]

Unit tests: pass. 62017 tests passed, 0 failed and 783 were skipped.

clang-tidy: unknown.

clang-format: fail. Please format your changes with clang-format by running git-clang-format HEAD^ or applying this patch.

Build artifacts: diff.json, clang-format.patch, CMakeCache.txt, console-log.txt, test-results.xml

[gchatelet]

• Fixing clang-format

llvm/include/llvm/IR/IRBuilder.h

public:
   /// specified pointers.
   ///
   /// DstAlign/SrcAlign are the alignments of the Dst/Src pointers, respectively.
   ///
   /// If the pointers aren't i8*, they will be converted.  If a TBAA tag is
   /// specified, it will be added to the instruction. Likewise with alias.scope
   /// and noalias tags.
   CallInst *CreateElementUnorderedAtomicMemCpy(
-      Value *Dst, unsigned DstAlign, Value *Src, unsigned SrcAlign,
-      uint64_t Size, uint32_t ElementSize, MDNode *TBAATag = nullptr,
+      Value *Dst, Align DstAlign, Value *Src, Align SrcAlign, Value *Size,
+      uint32_t ElementSize, MDNode *TBAATag = nullptr,
       MDNode *TBAAStructTag = nullptr, MDNode *ScopeTag = nullptr,
-      MDNode *NoAliasTag = nullptr) {
+      MDNode *NoAliasTag = nullptr);
+
+  /// FIXME: Remove this function once transition to Align is over.
+  /// Use the version that takes Align instead of this one.
+  LLVM_ATTRIBUTE_DEPRECATED(CallInst *CreateElementUnorderedAtomicMemCpy(
+                                Value *Dst, unsigned DstAlign, Value *Src,
+                                unsigned SrcAlign, uint64_t Size,
+                                uint32_t ElementSize, MDNode *TBAATag = nullptr,
+                                MDNode *TBAAStructTag = nullptr,
+                                MDNode *ScopeTag = nullptr,
+                                MDNode *NoAliasTag = nullptr),
+                            "Use the version that takes Align instead") {
     return CreateElementUnorderedAtomicMemCpy(
-        Dst, DstAlign, Src, SrcAlign, getInt64(Size), ElementSize, TBAATag,
-        TBAAStructTag, ScopeTag, NoAliasTag);
+        Dst, Align(DstAlign), Src, Align(SrcAlign), getInt64(Size), ElementSize,
+        TBAATag, TBAAStructTag, ScopeTag, NoAliasTag);
   }
 
-  CallInst *CreateElementUnorderedAtomicMemCpy(
-      Value *Dst, unsigned DstAlign, Value *Src, unsigned SrcAlign, Value *Size,
+  /// FIXME: Remove this function once transition to Align is over.
+  /// Use the version that takes Align instead of this one.
+  LLVM_ATTRIBUTE_DEPRECATED(CallInst *CreateElementUnorderedAtomicMemCpy(
+                                Value *Dst, unsigned DstAlign, Value *Src,
+                                unsigned SrcAlign, Value *Size,
                                 uint32_t ElementSize, MDNode *TBAATag = nullptr,
-      MDNode *TBAAStructTag = nullptr, MDNode *ScopeTag = nullptr,
-      MDNode *NoAliasTag = nullptr);
+                                MDNode *TBAAStructTag = nullptr,
+                                MDNode *ScopeTag = nullptr,
+                                MDNode *NoAliasTag = nullptr),
+                            "Use the version that takes Align instead") {
+    return CreateElementUnorderedAtomicMemCpy(
+        Dst, Align(DstAlign), Src, Align(SrcAlign), Size, ElementSize, TBAATag,
+        TBAAStructTag, ScopeTag, NoAliasTag);
+  }
 
   /// Create and insert a memmove between the specified
   /// pointers.
   ///
   /// If the pointers aren't i8*, they will be converted.  If a TBAA tag is
   /// specified, it will be added to the instruction. Likewise with alias.scope
   /// and noalias tags.
   /// FIXME: Remove this function once transition to Align is over.

llvm/lib/IR/IRBuilder.cpp

CallInst *IRBuilderBase::CreateMemCpy(Value *Dst, MaybeAlign DstAlign,
 
   if (NoAliasTag)
     CI->setMetadata(LLVMContext::MD_noalias, NoAliasTag);
 
   return CI;
 }
 
 CallInst *IRBuilderBase::CreateElementUnorderedAtomicMemCpy(
-    Value *Dst, unsigned DstAlign, Value *Src, unsigned SrcAlign, Value *Size,
+    Value *Dst, Align DstAlign, Value *Src, Align SrcAlign, Value *Size,
     uint32_t ElementSize, MDNode *TBAATag, MDNode *TBAAStructTag,
     MDNode *ScopeTag, MDNode *NoAliasTag) {
   assert(DstAlign >= ElementSize &&
          "Pointer alignment must be at least element size");
   assert(SrcAlign >= ElementSize &&
          "Pointer alignment must be at least element size");
   Dst = getCastedInt8PtrValue(Dst);
   Src = getCastedInt8PtrValue(Src);

llvm/lib/Transforms/Scalar/LoopIdiomRecognize.cpp

bool LoopIdiomRecognize::processLoopStoreOfLoopLoad(StoreInst *SI,
   //  If the load or store are atomic, then they must necessarily be unordered
   //  by previous checks.
   if (!SI->isAtomic() && !LI->isAtomic())
     NewCall = Builder.CreateMemCpy(StoreBasePtr, SI->getAlign(), LoadBasePtr,
                                    LI->getAlign(), NumBytes);
   else {
     // We cannot allow unaligned ops for unordered load/store, so reject
     // anything where the alignment isn't at least the element size.
-    unsigned Align = std::min(SI->getAlignment(), LI->getAlignment());
-    if (Align < StoreSize)
+    const MaybeAlign StoreAlign = SI->getAlign();
+    const MaybeAlign LoadAlign = LI->getAlign();
+    if (StoreAlign == None || LoadAlign == None)
+      return false;
+    if (*StoreAlign < StoreSize || *LoadAlign < StoreSize)
       return false;
 
     // If the element.atomic memcpy is not lowered into explicit
     // loads/stores later, then it will be lowered into an element-size
     // specific lib call. If the lib call doesn't exist for our store size, then
     // we shouldn't generate the memcpy.
     if (StoreSize > TTI->getAtomicMemIntrinsicMaxElementSize())
       return false;
 
     // Create the call.
     // Note that unordered atomic loads/stores are *required* by the spec to
     // have an alignment but non-atomic loads/stores may not.
     NewCall = Builder.CreateElementUnorderedAtomicMemCpy(
-        StoreBasePtr, SI->getAlignment(), LoadBasePtr, LI->getAlignment(),
-        NumBytes, StoreSize);
+        StoreBasePtr, *StoreAlign, LoadBasePtr, *LoadAlign, NumBytes,
+        StoreSize);
   }
   NewCall->setDebugLoc(SI->getDebugLoc());
 
   LLVM_DEBUG(dbgs() << "  Formed memcpy: " << *NewCall << "\n"
                     << "    from load ptr=" << *LoadEv << " at: " << *LI << "\n"
                     << "    from store ptr=" << *StoreEv << " at: " << *SI
                     << "\n");
 

mlir/lib/Dialect/GPU/Transforms/AllReduceLowering.cpp

void rewrite() {
     create<gpu::BarrierOp>();
     Value result = create<LoadOp>(valueType, buffer, zero);
 
     rewriter.replaceOp(reduceOp, result);
   }
 
 private:
   // Shortcut to create an op from rewriter using loc as the first argument.
-  template <typename T, typename... Args> T create(Args... args) {
+  template <typename T, typename... Args>
+  T create(Args... args) {
     return rewriter.create<T>(loc, std::forward<Args>(args)...);
   }
 
   // Creates dimension op of type T, with the result casted to int32.
-  template <typename T> Value getDimOp(StringRef dimension) {
+  template <typename T>
+  Value getDimOp(StringRef dimension) {
     Value dim = create<T>(indexType, rewriter.getStringAttr(dimension));
     return create<IndexCastOp>(int32Type, dim);
   }
 
   /// Adds type to funcOp's workgroup attributions.
   Value createWorkgroupBuffer() {
     int workgroupMemoryAddressSpace = 3;
     auto bufferType =
AccumulatorFactory getFactory(StringRef opName) {
     if (opName == "add")
       return isFloatingPoint ? getFactory<AddFOp>() : getFactory<AddIOp>();
     if (opName == "mul")
       return isFloatingPoint ? getFactory<MulFOp>() : getFactory<MulIOp>();
     return AccumulatorFactory();
   }
 
   /// Returns an accumulator factory that creates an op of type T.
-  template <typename T> AccumulatorFactory getFactory() {
+  template <typename T>
+  AccumulatorFactory getFactory() {
     return [&](Value lhs, Value rhs) {
       return create<T>(lhs.getType(), lhs, rhs);
     };
   }
 
   /// Creates an if-block skeleton and calls the two factories to generate the
   /// ops in the `then` and `else` block..
   ///