ARM cost model: Penalize insertelement into D subregisters

Swift has a renaming dependency if we load into D subregisters. We don't have a
way of distinguishing between insertelement operations of values from loads and
other values. Therefore, we are pessimistic for now (The performance problem
showed up in example 14 of gcc-loops).

radar://13096933

llvm-svn: 174300

llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp

@@ -117,6 +117,7 @@ public:
   unsigned getCastInstrCost(unsigned Opcode, Type *Dst,
                                       Type *Src) const;
 
+  unsigned getVectorInstrCost(unsigned Opcode, Type *Val, unsigned Index) const;
   /// @}
 };
 
@@ -197,3 +198,15 @@ unsigned ARMTTI::getCastInstrCost(unsigned Opcode, Type *Dst,
 
   return TargetTransformInfo::getCastInstrCost(Opcode, Dst, Src);
 }
+
+unsigned ARMTTI::getVectorInstrCost(unsigned Opcode, Type *ValTy,
+                                    unsigned Index) const {
+  // Penalize inserting into an D-subregister.
+  if (ST->isSwift() &&
+      Opcode == Instruction::InsertElement &&
+      ValTy->isVectorTy() &&
+      ValTy->getScalarSizeInBits() <= 32)
+    return 2;
+
+  return TargetTransformInfo::getVectorInstrCost(Opcode, ValTy, Index);
+}

llvm/test/Analysis/CostModel/ARM/insertelement.ll

@@ -0,0 +1,46 @@
+; RUN: opt -cost-model -analyze -mtriple=thumbv7-apple-ios6.0.0 -mcpu=swift < %s | FileCheck %s
+
+target datalayout = "e-p:32:32:32-i1:8:32-i8:8:32-i16:16:32-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:32:64-v128:32:128-a0:0:32-n32-S32"
+target triple = "thumbv7-apple-ios6.0.0"
+
+; Multiple insert elements from loads into d subregisters are expensive on swift
+; due to renaming constraints.
+%T_i8v = type <8 x i8>
+%T_i8 = type i8
+; CHECK: insertelement_i8
+define void @insertelement_i8(%T_i8* %saddr,
+                           %T_i8v* %vaddr) {
+  %v0 = load %T_i8v* %vaddr
+  %v1 = load %T_i8* %saddr
+;CHECK: estimated cost of 2 for {{.*}} insertelement <8 x i8>
+  %v2 = insertelement %T_i8v %v0, %T_i8 %v1, i32 1
+  store %T_i8v %v2, %T_i8v* %vaddr
+  ret void
+}
+
+
+%T_i16v = type <4 x i16>
+%T_i16 = type i16
+; CHECK: insertelement_i16
+define void @insertelement_i16(%T_i16* %saddr,
+                           %T_i16v* %vaddr) {
+  %v0 = load %T_i16v* %vaddr
+  %v1 = load %T_i16* %saddr
+;CHECK: estimated cost of 2 for {{.*}} insertelement <4 x i16>
+  %v2 = insertelement %T_i16v %v0, %T_i16 %v1, i32 1
+  store %T_i16v %v2, %T_i16v* %vaddr
+  ret void
+}
+
+%T_i32v = type <2 x i32>
+%T_i32 = type i32
+; CHECK: insertelement_i32
+define void @insertelement_i32(%T_i32* %saddr,
+                           %T_i32v* %vaddr) {
+  %v0 = load %T_i32v* %vaddr
+  %v1 = load %T_i32* %saddr
+;CHECK: estimated cost of 2 for {{.*}} insertelement <2 x i32>
+  %v2 = insertelement %T_i32v %v0, %T_i32 %v1, i32 1
+  store %T_i32v %v2, %T_i32v* %vaddr
+  ret void
+}

llvm/test/Analysis/CostModel/ARM/lit.local.cfg

@@ -0,0 +1,6 @@
+config.suffixes = ['.ll', '.c', '.cpp']
+
+targets = set(config.root.targets_to_build.split())
+if not 'ARM' in targets:
+    config.unsupported = True
+