[globalisel][tablegen] Demote OptForSize/OptForMinSize/ForCodeSize to per-function...

Summary:
This causes them to be re-computed more often than necessary but resolves
objections that were raised post-commit on r301750.

Reviewers: qcolombet, ab, t.p.northover, rovka, kristof.beyls

Reviewed By: qcolombet

Subscribers: igorb, llvm-commits

Differential Revision: https://reviews.llvm.org/D32861

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@303418 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/AArch64/AArch64InstrInfo.td b/lib/Target/AArch64/AArch64InstrInfo.td
index 5ddf66654a675f47c76baeb25053319ef1b25059..da68f3165c5eee70a684a9deb0a45cd2a09ec60a 100644 (file)
--- a/lib/Target/AArch64/AArch64InstrInfo.td
+++ b/lib/Target/AArch64/AArch64InstrInfo.td
@@ -315,8 +315,14 @@ def AArch64umaxv    : SDNode<"AArch64ISD::UMAXV", SDT_AArch64UnaryVec>;
 // AArch64 Instruction Predicate Definitions.
 def IsDarwin  : Predicate<"Subtarget->isTargetDarwin()">;
 def IsNotDarwin: Predicate<"!Subtarget->isTargetDarwin()">;
-def ForCodeSize   : Predicate<"Subtarget->getForCodeSize()">;
-def NotForCodeSize   : Predicate<"!Subtarget->getForCodeSize()">;
+
+// We could compute these on a per-module basis but doing so requires accessing
+// the Function object through the <Target>Subtarget and objections were raised
+// to that (see post-commit review comments for r301750).
+let RecomputePerFunction = 1 in {
+  def ForCodeSize   : Predicate<"MF->getFunction()->optForSize()">;
+  def NotForCodeSize   : Predicate<"!MF->getFunction()->optForSize()">;
+}
 
 include "AArch64InstrFormats.td"
 

diff --git a/lib/Target/AArch64/AArch64Subtarget.cpp b/lib/Target/AArch64/AArch64Subtarget.cpp
index 1c81d34014fda6305f895286440d34c597c405bd..b369ee7e4ba214720a03be03291db0f85cb8e2d0 100644 (file)
--- a/lib/Target/AArch64/AArch64Subtarget.cpp
+++ b/lib/Target/AArch64/AArch64Subtarget.cpp
@@ -164,12 +164,11 @@ struct AArch64GISelActualAccessor : public GISelAccessor {
 
 AArch64Subtarget::AArch64Subtarget(const Triple &TT, const std::string &CPU,
                                    const std::string &FS,
-                                   const TargetMachine &TM, bool LittleEndian,
-                                   bool ForCodeSize)
+                                   const TargetMachine &TM, bool LittleEndian)
     : AArch64GenSubtargetInfo(TT, CPU, FS), ReserveX18(TT.isOSDarwin()),
       IsLittle(LittleEndian), TargetTriple(TT), FrameLowering(),
       InstrInfo(initializeSubtargetDependencies(FS, CPU)), TSInfo(),
-      TLInfo(TM, *this), GISel(), ForCodeSize(ForCodeSize) {
+      TLInfo(TM, *this), GISel() {
 #ifndef LLVM_BUILD_GLOBAL_ISEL
   GISelAccessor *AArch64GISel = new GISelAccessor();
 #else

diff --git a/lib/Target/AArch64/AArch64Subtarget.h b/lib/Target/AArch64/AArch64Subtarget.h
index df54bf3f48e181abe0c68fd1396498ecd7e440ea..7933e58c49eed755619e37800a62596e19ebc4e3 100644 (file)
--- a/lib/Target/AArch64/AArch64Subtarget.h
+++ b/lib/Target/AArch64/AArch64Subtarget.h
@@ -128,8 +128,6 @@ protected:
   /// an optional library.
   std::unique_ptr<GISelAccessor> GISel;
 
-  bool ForCodeSize;
-
 private:
   /// initializeSubtargetDependencies - Initializes using CPUString and the
   /// passed in feature string so that we can use initializer lists for
@@ -145,7 +143,7 @@ public:
   /// of the specified triple.
   AArch64Subtarget(const Triple &TT, const std::string &CPU,
                    const std::string &FS, const TargetMachine &TM,
-                   bool LittleEndian, bool ForCodeSize);
+                   bool LittleEndian);
 
   /// This object will take onwership of \p GISelAccessor.
   void setGISelAccessor(GISelAccessor &GISel) {
@@ -274,8 +272,6 @@ public:
     }
   }
 
-  bool getForCodeSize() const { return ForCodeSize; }
-
   /// ParseSubtargetFeatures - Parses features string setting specified
   /// subtarget options.  Definition of function is auto generated by tblgen.
   void ParseSubtargetFeatures(StringRef CPU, StringRef FS);

diff --git a/lib/Target/AArch64/AArch64TargetMachine.cpp b/lib/Target/AArch64/AArch64TargetMachine.cpp
index 71d9fb9e8f99aa9138d57a29cff665940ce5e044..132f192f2a9a77359c356f0adea7bc4bcd541684 100644 (file)
--- a/lib/Target/AArch64/AArch64TargetMachine.cpp
+++ b/lib/Target/AArch64/AArch64TargetMachine.cpp
@@ -214,7 +214,6 @@ const AArch64Subtarget *
 AArch64TargetMachine::getSubtargetImpl(const Function &F) const {
   Attribute CPUAttr = F.getFnAttribute("target-cpu");
   Attribute FSAttr = F.getFnAttribute("target-features");
-  bool ForCodeSize = F.optForSize();
 
   std::string CPU = !CPUAttr.hasAttribute(Attribute::None)
                         ? CPUAttr.getValueAsString().str()
@@ -222,17 +221,15 @@ AArch64TargetMachine::getSubtargetImpl(const Function &F) const {
   std::string FS = !FSAttr.hasAttribute(Attribute::None)
                        ? FSAttr.getValueAsString().str()
                        : TargetFS;
-  std::string ForCodeSizeStr =
-      std::string(ForCodeSize ? "+" : "-") + "forcodesize";
 
-  auto &I = SubtargetMap[CPU + FS + ForCodeSizeStr];
+  auto &I = SubtargetMap[CPU + FS];
   if (!I) {
     // This needs to be done before we create a new subtarget since any
     // creation will depend on the TM and the code generation flags on the
     // function that reside in TargetOptions.
     resetTargetOptions(F);
     I = llvm::make_unique<AArch64Subtarget>(TargetTriple, CPU, FS, *this,
-                                            isLittle, ForCodeSize);
+                                            isLittle);
   }
   return I.get();
 }

diff --git a/lib/Target/X86/X86InstrInfo.td b/lib/Target/X86/X86InstrInfo.td
index 4d7d8ece92d9ac8bfb974233fb1541009a1855ee..01df07e1715f76db89a19a6e02785d4d8d399a5e 100644 (file)
--- a/lib/Target/X86/X86InstrInfo.td
+++ b/lib/Target/X86/X86InstrInfo.td
@@ -896,9 +896,16 @@ def KernelCode   : Predicate<"TM.getCodeModel() == CodeModel::Kernel">;
 def NearData     : Predicate<"TM.getCodeModel() == CodeModel::Small ||"
                              "TM.getCodeModel() == CodeModel::Kernel">;
 def IsNotPIC     : Predicate<"!TM.isPositionIndependent()">;
-def OptForSize   : Predicate<"Subtarget->getOptForSize()">;
-def OptForMinSize : Predicate<"Subtarget->getOptForMinSize()">;
-def OptForSpeed  : Predicate<"!Subtarget->getOptForSize()">;
+
+// We could compute these on a per-module basis but doing so requires accessing
+// the Function object through the <Target>Subtarget and objections were raised
+// to that (see post-commit review comments for r301750).
+let RecomputePerFunction = 1 in {
+  def OptForSize   : Predicate<"MF->getFunction()->optForSize()">;
+  def OptForMinSize : Predicate<"MF->getFunction()->optForMinSize()">;
+  def OptForSpeed  : Predicate<"!MF->getFunction()->optForSize()">;
+}
+
 def FastBTMem    : Predicate<"!Subtarget->isBTMemSlow()">;
 def CallImmAddr  : Predicate<"Subtarget->isLegalToCallImmediateAddr()">;
 def FavorMemIndirectCall  : Predicate<"!Subtarget->callRegIndirect()">;

diff --git a/lib/Target/X86/X86Subtarget.cpp b/lib/Target/X86/X86Subtarget.cpp
index e8f1fa09741efb2565520f6ce5109951c82a05f8..2b1f43bffd71789c684cc086d047f0c9e28edb76 100644 (file)
--- a/lib/Target/X86/X86Subtarget.cpp
+++ b/lib/Target/X86/X86Subtarget.cpp
@@ -337,8 +337,7 @@ X86Subtarget &X86Subtarget::initializeSubtargetDependencies(StringRef CPU,
 
 X86Subtarget::X86Subtarget(const Triple &TT, StringRef CPU, StringRef FS,
                            const X86TargetMachine &TM,
-                           unsigned StackAlignOverride, bool OptForSize,
-                           bool OptForMinSize)
+                           unsigned StackAlignOverride)
     : X86GenSubtargetInfo(TT, CPU, FS), X86ProcFamily(Others),
       PICStyle(PICStyles::None), TM(TM), TargetTriple(TT),
       StackAlignOverride(StackAlignOverride),
@@ -348,8 +347,7 @@ X86Subtarget::X86Subtarget(const Triple &TT, StringRef CPU, StringRef FS,
       In16BitMode(TargetTriple.getArch() == Triple::x86 &&
                   TargetTriple.getEnvironment() == Triple::CODE16),
       InstrInfo(initializeSubtargetDependencies(CPU, FS)), TLInfo(TM, *this),
-      FrameLowering(*this, getStackAlignment()), OptForSize(OptForSize),
-      OptForMinSize(OptForMinSize) {
+      FrameLowering(*this, getStackAlignment()) {
   // Determine the PICStyle based on the target selected.
   if (!isPositionIndependent())
     setPICStyle(PICStyles::None);

diff --git a/lib/Target/X86/X86Subtarget.h b/lib/Target/X86/X86Subtarget.h
index 02be95e2e5562897404c03cf75eda62eef932d9b..a9f3a2aee1bed494884677b13813f4a2344f50fd 100644 (file)
--- a/lib/Target/X86/X86Subtarget.h
+++ b/lib/Target/X86/X86Subtarget.h
@@ -336,16 +336,12 @@ private:
   X86TargetLowering TLInfo;
   X86FrameLowering FrameLowering;
 
-  bool OptForSize;
-  bool OptForMinSize;
-
 public:
   /// This constructor initializes the data members to match that
   /// of the specified triple.
   ///
   X86Subtarget(const Triple &TT, StringRef CPU, StringRef FS,
-               const X86TargetMachine &TM, unsigned StackAlignOverride,
-               bool OptForSize, bool OptForMinSize);
+               const X86TargetMachine &TM, unsigned StackAlignOverride);
 
   /// This object will take onwership of \p GISelAccessor.
   void setGISelAccessor(GISelAccessor &GISel) { this->GISel.reset(&GISel); }
@@ -513,9 +509,6 @@ public:
   bool isSLM() const { return X86ProcFamily == IntelSLM; }
   bool useSoftFloat() const { return UseSoftFloat; }
 
-  bool getOptForSize() const { return OptForSize; }
-  bool getOptForMinSize() const { return OptForMinSize; }
-
   /// Use mfence if we have SSE2 or we're on x86-64 (even if we asked for
   /// no-sse2). There isn't any reason to disable it if the target processor
   /// supports it.

diff --git a/lib/Target/X86/X86TargetMachine.cpp b/lib/Target/X86/X86TargetMachine.cpp
index e717f84104c5cca19fa9ffc094309fe8e4817d12..53a8e83b36fca3d2e2e376e6756fba4875a67dd0 100644 (file)
--- a/lib/Target/X86/X86TargetMachine.cpp
+++ b/lib/Target/X86/X86TargetMachine.cpp
@@ -270,12 +270,6 @@ X86TargetMachine::getSubtargetImpl(const Function &F) const {
 
   FS = Key.substr(CPU.size());
 
-  bool OptForSize = F.optForSize();
-  bool OptForMinSize = F.optForMinSize();
-
-  Key += std::string(OptForSize ? "+" : "-") + "optforsize";
-  Key += std::string(OptForMinSize ? "+" : "-") + "optforminsize";
-
   auto &I = SubtargetMap[Key];
   if (!I) {
     // This needs to be done before we create a new subtarget since any
@@ -283,8 +277,7 @@ X86TargetMachine::getSubtargetImpl(const Function &F) const {
     // function that reside in TargetOptions.
     resetTargetOptions(F);
     I = llvm::make_unique<X86Subtarget>(TargetTriple, CPU, FS, *this,
-                                        Options.StackAlignmentOverride,
-                                        OptForSize, OptForMinSize);
+                                        Options.StackAlignmentOverride);
 #ifndef LLVM_BUILD_GLOBAL_ISEL
     GISelAccessor *GISel = new GISelAccessor();
 #else

diff --git a/unittests/Target/AArch64/InstSizes.cpp b/unittests/Target/AArch64/InstSizes.cpp
index 5adbd7d2de4391d67f1d11d77de1e502c1fa836a..22b47c6852ab62c17a38b036492ac7ecd99e453c 100644 (file)
--- a/unittests/Target/AArch64/InstSizes.cpp
+++ b/unittests/Target/AArch64/InstSizes.cpp
@@ -30,8 +30,7 @@ std::unique_ptr<TargetMachine> createTargetMachine() {
 
 std::unique_ptr<AArch64InstrInfo> createInstrInfo(TargetMachine *TM) {
   AArch64Subtarget ST(TM->getTargetTriple(), TM->getTargetCPU(),
-                      TM->getTargetFeatureString(), *TM, /* isLittle */ false,
-                      /* ForCodeSize */ false);
+                      TM->getTargetFeatureString(), *TM, /* isLittle */ false);
   return llvm::make_unique<AArch64InstrInfo>(ST);
 }