Merge pull request #1279 from swiftwasm/master

[pull] swiftwasm from master

Author: pull[bot]

include/swift/SIL/SILInstruction.h

@@ -375,6 +375,12 @@ class SILInstruction
     /// The instruction may read memory.
     MayRead,
     /// The instruction may write to memory.
+    /// This includes destroying or taking from memory (e.g. destroy_addr,
+    /// copy_addr [take], load [take]).
+    /// Although, physically, destroying or taking does not modify the memory,
+    /// it is important to model it is a write. Optimizations must not assume
+    /// that the value stored in memory is still available for loading after
+    /// the memory is destroyed or taken.
     MayWrite,
     /// The instruction may read or write memory.
     MayReadWrite,

include/swift/SILOptimizer/PassManager/Passes.def

@@ -283,6 +283,8 @@ PASS(PerformanceSILLinker, "performance-linker",
      "Deserialize all referenced SIL functions")
 PASS(RawSILInstLowering, "raw-sil-inst-lowering",
      "Lower all raw SIL instructions to canonical equivalents.")
+PASS(TempLValueOpt, "temp-lvalue-opt",
+     "Remove short-lived immutable temporary l-values")
 PASS(TempRValueOpt, "temp-rvalue-opt",
      "Remove short-lived immutable temporary copies")
 PASS(SideEffectsDumper, "side-effects-dump",

lib/SILOptimizer/Analysis/AliasAnalysis.cpp

@@ -569,6 +569,15 @@ AliasResult AliasAnalysis::alias(SILValue V1, SILValue V2,
   return Result;
 }
 
+/// Get the underlying object, looking through init_enum_data_addr and
+/// init_existential_addr.
+static SILValue stripInitEnumAndExistentialAddr(SILValue v) {
+  while (isa<InitEnumDataAddrInst>(v) || isa<InitExistentialAddrInst>(v)) {
+    v = getUnderlyingObject(cast<SingleValueInstruction>(v)->getOperand(0));
+  }
+  return v;
+}
+
 /// The main AA entry point. Performs various analyses on V1, V2 in an attempt
 /// to disambiguate the two values.
 AliasResult AliasAnalysis::aliasInner(SILValue V1, SILValue V2,
@@ -614,9 +623,12 @@ AliasResult AliasAnalysis::aliasInner(SILValue V1, SILValue V2,
   LLVM_DEBUG(llvm::dbgs() << "        Underlying V1:" << *O1);
   LLVM_DEBUG(llvm::dbgs() << "        Underlying V2:" << *O2);
 
-  // If O1 and O2 do not equal, see if we can prove that they cannot be the
-  // same object. If we can, return No Alias.
-  if (O1 != O2 && aliasUnequalObjects(O1, O2))
+  // If the underlying objects are not equal, see if we can prove that they
+  // cannot be the same object. If we can, return No Alias.
+  // For this we even look through init_enum_data_addr and init_existential_addr.
+  SILValue StrippedO1 = stripInitEnumAndExistentialAddr(O1);
+  SILValue StrippedO2 = stripInitEnumAndExistentialAddr(O2);
+  if (StrippedO1 != StrippedO2 && aliasUnequalObjects(StrippedO1, StrippedO2))
     return AliasResult::NoAlias;
 
   // Ok, either O1, O2 are the same or we could not prove anything based off of

lib/SILOptimizer/Analysis/MemoryBehavior.cpp

@@ -174,6 +174,7 @@ class MemoryBehaviorVisitor
 
   MemBehavior visitLoadInst(LoadInst *LI);
   MemBehavior visitStoreInst(StoreInst *SI);
+  MemBehavior visitCopyAddrInst(CopyAddrInst *CAI);
   MemBehavior visitApplyInst(ApplyInst *AI);
   MemBehavior visitTryApplyInst(TryApplyInst *AI);
   MemBehavior visitBuiltinInst(BuiltinInst *BI);
@@ -245,6 +246,10 @@ MemBehavior MemoryBehaviorVisitor::visitLoadInst(LoadInst *LI) {
   if (!mayAlias(LI->getOperand()))
     return MemBehavior::None;
 
+  // A take is modelled as a write. See MemoryBehavior::MayWrite.
+  if (LI->getOwnershipQualifier() == LoadOwnershipQualifier::Take)
+      return MemBehavior::MayReadWrite;
+
   LLVM_DEBUG(llvm::dbgs() << "  Could not prove that load inst does not alias "
                              "pointer. Returning may read.\n");
   return MemBehavior::MayRead;
@@ -268,6 +273,34 @@ MemBehavior MemoryBehaviorVisitor::visitStoreInst(StoreInst *SI) {
   return MemBehavior::MayWrite;
 }
 
+MemBehavior MemoryBehaviorVisitor::visitCopyAddrInst(CopyAddrInst *CAI) {
+  // If it's an assign to the destination, a destructor might be called on the
+  // old value. This can have any side effects.
+  // We could also check if it's a trivial type (which cannot have any side
+  // effect on destruction), but such copy_addr instructions are optimized to
+  // load/stores anyway, so it's probably not worth it.
+  if (!CAI->isInitializationOfDest())
+    return MemBehavior::MayHaveSideEffects;
+
+  bool mayWrite = mayAlias(CAI->getDest());
+  bool mayRead = mayAlias(CAI->getSrc());
+  
+  if (mayRead) {
+    if (mayWrite)
+      return MemBehavior::MayReadWrite;
+
+    // A take is modelled as a write. See MemoryBehavior::MayWrite.
+    if (CAI->isTakeOfSrc())
+      return MemBehavior::MayReadWrite;
+
+    return MemBehavior::MayRead;
+  }
+  if (mayWrite)
+    return MemBehavior::MayWrite;
+
+  return MemBehavior::None;
+}
+
 MemBehavior MemoryBehaviorVisitor::visitBuiltinInst(BuiltinInst *BI) {
   // If our callee is not a builtin, be conservative and return may have side
   // effects.

lib/SILOptimizer/PassManager/PassPipeline.cpp

@@ -283,6 +283,9 @@ void addFunctionPasses(SILPassPipelinePlan &P,
   // splits up copy_addr.
   P.addCopyForwarding();
 
+  // Optimize copies from a temporary (an "l-value") to a destination.
+  P.addTempLValueOpt();
+
   // Split up opaque operations (copy_addr, retain_value, etc.).
   P.addLowerAggregateInstrs();
 

lib/SILOptimizer/Transforms/CMakeLists.txt

@@ -38,5 +38,6 @@ target_sources(swiftSILOptimizer PRIVATE
   Sink.cpp
   SpeculativeDevirtualizer.cpp
   StackPromotion.cpp
+  TempLValueOpt.cpp
   TempRValueElimination.cpp
   UnsafeGuaranteedPeephole.cpp)