XCore target: Add TypeString meta data to IR output.

[robertlytton]

Summary

This includes the addition of the virtual function:

virtual void TargetCodeGenInfo::EmitTargetMD(const Decl *D, llvm::GlobalValue *GV, CodeGen::CodeGenModule &M) const { }

Richard Osborne to review from XCore perspective
Also need review from Clang viewpoint re: EmitTargetMD - Rafael, can you?
Also the code correctness - Richard again.

Thank you

Robert

[robertlytton]

There is a problem when running the test-suite,
I'll investigate, and add patch to code/tests as necessary.

[• rafael]

Please use the current naming convesion, so emitTargetMD. Eric has
worked more with medatata (for debug info) and probably has a better
idea than I if this is the correct way to plug it.

Can you describe a bit what this metadata is?

[echristo]

What are you trying to do here?

[robertlytton]

Hi Eric,

The XCore ABI includes a type information section that communicates symbol type information to the linker.
The linker uses this information to verify safety/correctness of things such as array bound and pointers etc.

This type information (TypeString) is emitted into meta data for all global symbols - definitions, declarations, functions & variables.
The TypeString carries type, qualifier, name, size & value details - as verified in xcore-stringtype.c

Currently meta data seems the best way to get this information to the back-end and hence the linker.
It seems the IR could encode this information as attributes for functions definitions & declaration - but global variables don't have attributes.

The addition of emitTargetMD() is because SetTargetAttributes() only visits function definitions.
Is there a better hook or better mechanism?

Thank you.

Robert

[robertlytton]

Add check for null Decl pointer.
(Will be null for function declarations added by clang viz exception handling support)

[robertlytton]

Add handling of anonymous structures, unions and enums.
Also reduce the expansion of recursive record types.

[robertlytton]

fix comment in test file.

[echristo]

This needs the changes that Rafael mentioned as well as needing the format that you're emitting documented.

Few other comments inline for now.

test/CodeGen/xcore-stringtype.c

7

"their"

Also why 33? Also what's supported/unsupported? It's not obvious from the comment in the testcase and would require someone knowing the entire scheme to debug.

36

It would be nice if the code in the testcase followed the llvm coding standard.

[robertlytton]

Renamed EmitTargetMD emitTargetMD.
Added explanatory comment to GetTypeString().
Added explanatory comment, lowered initial function capitals and reduced line lengths in test/CodeGen/xcore-stringtype.c

[echristo]

Some more inline comments...

lib/CodeGen/TargetInfo.cpp

6085

Coding style here (and elsewhere):

http://llvm.org/docs/CodingStandards.html#name-types-functions-variables-and-enumerators-properly

6274

I'd say that since this is an odd vendor thing that we might want to just include the format in the text that describes it?

On the other hand if xmos isn't in business anymore we could just delete the port immediately for lack of documentation? :)

[robertlytton]

Hi Eric,

I feel suitably embarrassed and have reread the coding standard again as penance.
Functions are now initial lower case.
I have also used auto for iterator declarations.

The format details for the TypeString metadata is quite long and I would agree with your sentiments.

Robert

[echristo]

Added a few more comments inline.

Thanks!

lib/CodeGen/TargetInfo.cpp

6084

What's going on with IncompleteUsedCount? I don't quite get it.

6218

Comment for the function.

6222

Sentences begin with a capital letter :)

6228

Comment for the function.

6247

Do you check the return value of this? Also comment.

6257

Comment on what's going on here. Why does "str" check the state?

6332

You can check whether the record is a union without passing it down as a bool.

6362

Why not check Success before the line above it?

6364

Sorting union encodings and not struct ones? Weird.

[robertlytton]

Hi Eric,

I've extended the comments as suggested.
I have also extended the algorithm to be able to cache recursive types for use when they are non-members.
The interface into the cache has also be tidied.

Thank you for your input.

Robert

[robertlytton]

(remove stray '/' that got accidentally added to diff)

[robertlytton]

I added a bug with the change in the algorithm.
Here is an extra test and corrective code.

[echristo]

Needs block comments for all of the static functions. Otherwise I don't see anything terrible.

[robertlytton]

Added comments to static functions.
cheers.
Good to go?

[echristo]

LGTM.

[robertlytton]

Just spotted that a object had a constructor as its type rather than the class.

[• friedgold]

lib/CodeGen/TargetInfo.cpp

6117

TypesString -> TypeString

6118

A found this a bit unclear. How about:

An incomplete TypesString that has been used in a Recursive type encoding

6154

The members of the Entry struct could do with comments. What does Swapped represent?

6163

Could you rename this to have lookup or find in the name - it isn't clear what the method does from the name.

6174

& should be next to the variable name.

6271

& should be next to the variable name.

6294

What's the difference between self recusive typestrings and recursive typestrings?

[robertlytton]

Hi Richard,
I've reworked comments etc based on you input.
There is no difference between 'self recusive typestrings' and 'recursive typestrings'.
I've renamed the offending variables.

[• friedgold]

LGTM

lib/CodeGen/CodeGenModule.cpp

     }
   }
 
+  getTargetCodeGenInfo().emitTargetMD(D, F, *this);
+
   // Make sure the result is of the requested type.
   if (!IsIncompleteFunction) {
     assert(F->getType()->getElementType() == Ty);

       isExternallyVisible(D->getLinkageAndVisibility().getLinkage()))
     GV->setSection(".cp.rodata");
 
+  getTargetCodeGenInfo().emitTargetMD(D, GV, *this);
+
   return GV;
 }
 

lib/CodeGen/TargetInfo.h

     virtual void SetTargetAttributes(const Decl *D, llvm::GlobalValue *GV,
                                      CodeGen::CodeGenModule &M) const { }
 
+    /// EmitTargetMD - Provides a convenient hook to handle extra
+    /// target-specific metadata for the given global.
+    virtual void emitTargetMD(const Decl *D, llvm::GlobalValue *GV,
+                              CodeGen::CodeGenModule &M) const { }
+
     /// Determines the size of struct _Unwind_Exception on this platform,
     /// in 8-bit units.  The Itanium ABI defines this as:
     ///   struct _Unwind_Exception {

lib/CodeGen/TargetInfo.cpp

 #include "llvm/IR/DataLayout.h"
 #include "llvm/IR/Type.h"
 #include "llvm/Support/raw_ostream.h"
+
+#include <algorithm>    // std::sort
+
 using namespace clang;
 using namespace CodeGen;
 

 //===----------------------------------------------------------------------===//
 // XCore ABI Implementation
 //===----------------------------------------------------------------------===//
+
 namespace {
+
+/// A SmallStringEnc instance is used to build up the TypeString by passing
+/// it by reference between functions that append to it.
+typedef llvm::SmallString<128> SmallStringEnc;
+
+/// TypeStringCache caches the meta encodings of Types.
+///
+/// The reason for caching TypeStrings is two fold:
+///   1. To cache a type's encoding for later uses;
+///   2. As a means to break recursive member type inclusion.
+///
+/// A cache Entry can have a Status of:
+///   NonRecursive:   The type encoding is not recursive;
+///   Recursive:      The type encoding is recursive;
+///   Incomplete:     An incomplete TypeString;
+///   IncompleteUsed: An incomplete TypeString that has been used in a
+///                   Recursive type encoding.
+///
+/// A NonRecursive entry will have all of its sub-members expanded as fully
+/// as possible. Whilst it may contain types which are recursive, the type
+/// itself is not recursive and thus its encoding may be safely used whenever
+/// the type is encountered.
+///
+/// A Recursive entry will have all of its sub-members expanded as fully as
+/// possible. The type itself is recursive and it may contain other types which
+/// are recursive. The Recursive encoding must not be used during the expansion
+/// of a recursive type's recursive branch. For simplicity the code uses
+/// IncompleteCount to reject all usage of Recursive encodings for member types.
+///
+/// An Incomplete entry is always a RecordType and only encodes its
+/// identifier e.g. "s(S){}". Incomplete 'StubEnc' entries are ephemeral and
+/// are placed into the cache during type expansion as a means to identify and
+/// handle recursive inclusion of types as sub-members. If there is recursion
+/// the entry becomes IncompleteUsed.
+///
+/// During the expansion of a RecordType's members:
+///
+///   If the cache contains a NonRecursive encoding for the member type, the
+///   cached encoding is used;
+///
+///   If the cache contains a Recursive encoding for the member type, the
+///   cached encoding is 'Swapped' out, as it may be incorrect, and...
+///
+///   If the member is a RecordType, an Incomplete encoding is placed into the
+///   cache to break potential recursive inclusion of itself as a sub-member;
+///
+///   Once a member RecordType has been expanded, its temporary incomplete
+///   entry is removed from the cache. If a Recursive encoding was swapped out
+///   it is swapped back in;
+///
+///   If an incomplete entry is used to expand a sub-member, the incomplete
+///   entry is marked as IncompleteUsed. The cache keeps count of how many
+///   IncompleteUsed entries it currently contains in IncompleteUsedCount;
+///
+///   If a member's encoding is found to be a NonRecursive or Recursive viz:
+///   IncompleteUsedCount==0, the member's encoding is added to the cache.
+///   Else the member is part of a recursive type and thus the recursion has
+///   been exited too soon for the encoding to be correct for the member.
+///
+class TypeStringCache {
+  enum Status {NonRecursive, Recursive, Incomplete, IncompleteUsed};
+  struct Entry {
+    std::string Str;     // The encoded TypeString for the type.
+    enum Status State;   // Information about the encoding in 'Str'.
+    std::string Swapped; // A temporary place holder for a Recursive encoding
+                         // during the expansion of RecordType's members.
+  };
+  std::map<const IdentifierInfo *, struct Entry> Map;
+  unsigned IncompleteCount;     // Number of Incomplete entries in the Map.
+  unsigned IncompleteUsedCount; // Number of IncompleteUsed entries in the Map.
+public:
+  void addIncomplete(const IdentifierInfo *ID, std::string StubEnc);
+  bool removeIncomplete(const IdentifierInfo *ID);
+  void addIfComplete(const IdentifierInfo *ID, StringRef Str,
+                     bool IsRecursive);
+  StringRef lookupStr(const IdentifierInfo *ID);
+};
+
+/// TypeString encodings for union fields must be order.
+/// FieldEncoding is a helper for this ordering process.
+class FieldEncoding {
+  bool HasName;
+  std::string Enc;
+public:
+  FieldEncoding(bool b, SmallStringEnc &e) : HasName(b), Enc(e.c_str()) {};
+  StringRef str() {return Enc.c_str();};
+  bool operator<(const FieldEncoding &rhs) const {
+    if (HasName != rhs.HasName) return HasName;
+    return Enc < rhs.Enc;
+  }
+};
+
 class XCoreABIInfo : public DefaultABIInfo {
 public:
   XCoreABIInfo(CodeGen::CodeGenTypes &CGT) : DefaultABIInfo(CGT) {}

 };
 
 class XCoreTargetCodeGenInfo : public TargetCodeGenInfo {
+  mutable TypeStringCache TSC;
 public:
   XCoreTargetCodeGenInfo(CodeGenTypes &CGT)
     :TargetCodeGenInfo(new XCoreABIInfo(CGT)) {}
+  virtual void emitTargetMD(const Decl *D, llvm::GlobalValue *GV,
+                            CodeGen::CodeGenModule &M) const;
 };
+
 } // End anonymous namespace.
 
 llvm::Value *XCoreABIInfo::EmitVAArg(llvm::Value *VAListAddr, QualType Ty,

   return Val;
 }
 
+/// During the expansion of a RecordType, an incomplete TypeString is placed
+/// into the cache as a means to identify and break recursion.
+/// If there is a Recursive encoding in the cache, it is swapped out and will
+/// be reinserted by removeIncomplete().
+/// All other types of encoding should have been used rather than arriving here.
+void TypeStringCache::addIncomplete(const IdentifierInfo *ID,
+                                    std::string StubEnc) {
+  if (!ID)
+    return;
+  Entry &E = Map[ID];
+  assert( (E.Str.empty() || E.State == Recursive) &&
+         "Incorrectly use of addIncomplete");
+  assert(!StubEnc.empty() && "Passing an empty string to addIncomplete()");
+  E.Swapped.swap(E.Str); // swap out the Recursive
+  E.Str.swap(StubEnc);
+  E.State = Incomplete;
+  ++IncompleteCount;
+}
+
+/// Once the RecordType has been expanded, the temporary incomplete TypeString
+/// must be removed from the cache.
+/// If a Recursive was swapped out by addIncomplete(), it will be replaced.
+/// Returns true if the RecordType was defined recursively.
+bool TypeStringCache::removeIncomplete(const IdentifierInfo *ID) {
+  if (!ID)
+    return false;
+  auto I = Map.find(ID);
+  assert(I != Map.end() && "Entry not present");
+  Entry &E = I->second;
+  assert( (E.State == Incomplete ||
+           E.State == IncompleteUsed) &&
+         "Entry must be an incomplete type");
+  bool IsRecursive = false;
+  if (E.State == IncompleteUsed) {
+    // We made use of our Incomplete encoding, thus we are recursive.
+    IsRecursive = true;
+    --IncompleteUsedCount;
+  }
+  if (E.Swapped.empty())
+    Map.erase(I);
+  else {
+    // Swap the Recursive back.
+    E.Swapped.swap(E.Str);
+    E.Swapped.clear();
+    E.State = Recursive;
+  }
+  --IncompleteCount;
+  return IsRecursive;
+}
+
+/// Add the encoded TypeString to the cache only if it is NonRecursive or
+/// Recursive (viz: all sub-members were expanded as fully as possible).
+void TypeStringCache::addIfComplete(const IdentifierInfo *ID, StringRef Str,
+                                    bool IsRecursive) {
+  if (!ID || IncompleteUsedCount)
+    return; // No key or it is is an incomplete sub-type so don't add.
+  Entry &E = Map[ID];
+  if (IsRecursive && !E.Str.empty()) {
+    assert(E.State==Recursive && E.Str.size() == Str.size() &&
+           "This is not the same Recursive entry");
+    // The parent container was not recursive after all, so we could have used
+    // this Recursive sub-member entry after all, but we assumed the worse when
+    // we started viz: IncompleteCount!=0.
+    return;
+  }
+  assert(E.Str.empty() && "Entry already present");
+  E.Str = Str.str();
+  E.State = IsRecursive? Recursive : NonRecursive;
+}
+
+/// Return a cached TypeString encoding for the ID. If there isn't one, or we
+/// are recursively expanding a type (IncompleteCount != 0) and the cached
+/// encoding is Recursive, return an empty StringRef.
+StringRef TypeStringCache::lookupStr(const IdentifierInfo *ID) {
+  if (!ID)
+    return StringRef();   // We have no key.
+  auto I = Map.find(ID);
+  if (I == Map.end())
+    return StringRef();   // We have no encoding.
+  Entry &E = I->second;
+  if (E.State == Recursive && IncompleteCount)
+    return StringRef();   // We don't use Recursive encodings for member types.
+
+  if (E.State == Incomplete) {
+    // The incomplete type is being used to break out of recursion.
+    E.State = IncompleteUsed;
+    ++IncompleteUsedCount;
+  }
+  return E.Str.c_str();
+}
+
+/// The XCore ABI includes a type information section that communicates symbol
+/// type information to the linker. The linker uses this information to verify
+/// safety/correctness of things such as array bound and pointers et al.
+/// The ABI only requires C (and XC) language modules to emit TypeStrings.
+/// This type information (TypeString) is emitted into meta data for all global
+/// symbols: definitions, declarations, functions & variables.
+///
+/// The TypeString carries type, qualifier, name, size & value details.
+/// Please see 'Tools Development Guide' section 2.16.2 for format details:
+/// <https://www.xmos.com/download/public/Tools-Development-Guide%28X9114A%29.pdf>
+/// The output is tested by test/CodeGen/xcore-stringtype.c.
+///
+static bool getTypeString(SmallStringEnc &Enc, const Decl *D,
+                          CodeGen::CodeGenModule &CGM, TypeStringCache &TSC);
+
+/// XCore uses emitTargetMD to emit TypeString metadata for global symbols.
+void XCoreTargetCodeGenInfo::emitTargetMD(const Decl *D, llvm::GlobalValue *GV,
+                                          CodeGen::CodeGenModule &CGM) const {
+  SmallStringEnc Enc;
+  if (getTypeString(Enc, D, CGM, TSC)) {
+    llvm::LLVMContext &Ctx = CGM.getModule().getContext();
+    llvm::SmallVector<llvm::Value *, 2> MDVals;
+    MDVals.push_back(GV);
+    MDVals.push_back(llvm::MDString::get(Ctx, Enc.str()));
+    llvm::NamedMDNode *MD =
+      CGM.getModule().getOrInsertNamedMetadata("xcore.typestrings");
+    MD->addOperand(llvm::MDNode::get(Ctx, MDVals));
+  }
+}
+
+static bool appendType(SmallStringEnc &Enc, QualType QType,
+                       const CodeGen::CodeGenModule &CGM,
+                       TypeStringCache &TSC);
+
+/// Helper function for appendRecordType().
+/// Builds a SmallVector containing the encoded field types in declaration order.
+static bool extractFieldType(SmallVectorImpl<FieldEncoding> &FE,
+                             const RecordDecl *RD,
+                             const CodeGen::CodeGenModule &CGM,
+                             TypeStringCache &TSC) {
+  for (RecordDecl::field_iterator I = RD->field_begin(), E = RD->field_end();
+       I != E; ++I) {
+    SmallStringEnc Enc;
+    Enc += "m(";
+    Enc += I->getName();
+    Enc += "){";
+    if (I->isBitField()) {
+      Enc += "b(";
+      llvm::raw_svector_ostream OS(Enc);
+      OS.resync();
+      OS << I->getBitWidthValue(CGM.getContext());
+      OS.flush();
+      Enc += ':';
+    }
+    if (!appendType(Enc, I->getType(), CGM, TSC))
+      return false;
+    if (I->isBitField())
+      Enc += ')';
+    Enc += '}';
+    FE.push_back(FieldEncoding(!I->getName().empty(), Enc));
+  }
+  return true;
+}
+
+/// Appends structure and union types to Enc and adds encoding to cache.
+/// Recursively calls appendType (via extractFieldType) for each field.
+/// Union types have their fields ordered according to the ABI.
+static bool appendRecordType(SmallStringEnc &Enc, const RecordType *RT,
+                             const CodeGen::CodeGenModule &CGM,
+                             TypeStringCache &TSC, const IdentifierInfo *ID) {
+  // Append the cached TypeString if we have one.
+  StringRef TypeString = TSC.lookupStr(ID);
+  if (!TypeString.empty()) {
+    Enc += TypeString;
+    return true;
+  }
+
+  // Start to emit an incomplete TypeString.
+  size_t Start = Enc.size();
+  Enc += (RT->isUnionType()? 'u' : 's');
+  Enc += '(';
+  if (ID)
+    Enc += ID->getName();
+  Enc += "){";
+
+  // We collect all encoded fields and order as necessary.
+  bool IsRecursive = false;
+  SmallVector<FieldEncoding, 16> FE;
+  const RecordDecl *RD = RT->getDecl()->getDefinition();
+  if (RD && !RD->field_empty()) {
+    // An incomplete TypeString stub is placed in the cache for this RecordType
+    // so that recursive calls to this RecordType will use it whilst building a
+    // complete TypeString for this RecordType.
+    std::string StubEnc(Enc.substr(Start).str());
+    StubEnc += '}';  // StubEnc now holds a valid incomplete TypeString.
+    TSC.addIncomplete(ID, std::move(StubEnc));
+    if (!extractFieldType(FE, RD, CGM, TSC)) {
+      (void) TSC.removeIncomplete(ID);
+      return false;
+    }
+    IsRecursive = TSC.removeIncomplete(ID);
+    // The ABI requires unions to be sorted but not structures.
+    // See FieldEncoding::operator< for sort algorithm.
+    if (RT->isUnionType())
+      std::sort(FE.begin(), FE.end());
+  }
+
+  // We can now complete the TypeString.
+  if (unsigned E = FE.size())
+    for (unsigned I = 0; I != E; ++I) {
+      if (I)
+        Enc += ',';
+      Enc += FE[I].str();
+    }
+  Enc += '}';
+  TSC.addIfComplete(ID, Enc.substr(Start), IsRecursive);
+  return true;
+}
+
+/// Appends enum types to Enc and adds the encoding to the cache.
+static bool appendEnumType(SmallStringEnc &Enc, const EnumType *ET,
+                           TypeStringCache &TSC,
+                           const IdentifierInfo *ID) {
+  // Append the cached TypeString if we have one.
+  StringRef TypeString = TSC.lookupStr(ID);
+  if (!TypeString.empty()) {
+    Enc += TypeString;
+    return true;
+  }
+
+  size_t Start = Enc.size();
+  Enc += "e(";
+  if (ID)
+    Enc += ID->getName();
+  Enc += "){";
+  if (const EnumDecl *ED = ET->getDecl()->getDefinition()) {
+    auto I = ED->enumerator_begin();
+    auto E = ED->enumerator_end();
+    while (I != E) {
+      Enc += "m(";
+      Enc += I->getName();
+      Enc += "){";
+      I->getInitVal().toString(Enc);
+      Enc += '}';
+      ++I;
+      if (I != E)
+        Enc += ',';
+    }
+  }
+  Enc += '}';
+  TSC.addIfComplete(ID, Enc.substr(Start), false);
+  return true;
+}
+
+/// Appends type's qualifier to Enc.
+/// This is done prior to appending the type's encoding.
+static void appendQualifier(SmallStringEnc &Enc, QualType QT) {
+  // Qualifiers are emitted in alphabetical order.
+  static const char *Table[] = {"","c:","r:","cr:","v:","cv:","rv:","crv:"};
+  int Lookup = 0;
+  if (QT.isConstQualified())
+    Lookup += 1<<0;
+  if (QT.isRestrictQualified())
+    Lookup += 1<<1;
+  if (QT.isVolatileQualified())
+    Lookup += 1<<2;
+  Enc += Table[Lookup];
+}
+
+/// Appends built-in types to Enc.
+static bool appendBuiltinType(SmallStringEnc &Enc, const BuiltinType *BT) {
+  const char *EncType;
+  switch (BT->getKind()) {
+    case BuiltinType::Void:
+      EncType = "0";
+      break;
+    case BuiltinType::Bool:
+      EncType = "b";
+      break;
+    case BuiltinType::Char_U:
+      EncType = "uc";
+      break;
+    case BuiltinType::UChar:
+      EncType = "uc";
+      break;
+    case BuiltinType::SChar:
+      EncType = "sc";
+      break;
+    case BuiltinType::UShort:
+      EncType = "us";
+      break;
+    case BuiltinType::Short:
+      EncType = "ss";
+      break;
+    case BuiltinType::UInt:
+      EncType = "ui";
+      break;
+    case BuiltinType::Int:
+      EncType = "si";
+      break;
+    case BuiltinType::ULong:
+      EncType = "ul";
+      break;
+    case BuiltinType::Long:
+      EncType = "sl";
+      break;
+    case BuiltinType::ULongLong:
+      EncType = "ull";
+      break;
+    case BuiltinType::LongLong:
+      EncType = "sll";
+      break;
+    case BuiltinType::Float:
+      EncType = "ft";
+      break;
+    case BuiltinType::Double:
+      EncType = "d";
+      break;
+    case BuiltinType::LongDouble:
+      EncType = "ld";
+      break;
+    default:
+      return false;
+  }
+  Enc += EncType;
+  return true;
+}
+
+/// Appends a pointer encoding to Enc before calling appendType for the pointee.
+static bool appendPointerType(SmallStringEnc &Enc, const PointerType *PT,
+                              const CodeGen::CodeGenModule &CGM,
+                              TypeStringCache &TSC) {
+  Enc += "p(";
+  if (!appendType(Enc, PT->getPointeeType(), CGM, TSC))
+    return false;
+  Enc += ')';
+  return true;
+}
+
+/// Appends array encoding to Enc before calling appendType for the element.
+static bool appendArrayType(SmallStringEnc &Enc, const ArrayType *AT,
+                            const CodeGen::CodeGenModule &CGM,
+                            TypeStringCache &TSC, StringRef NoSizeEnc) {
+  if (AT->getSizeModifier() != ArrayType::Normal)
+    return false;
+  Enc += "a(";
+  if (const ConstantArrayType *CAT = dyn_cast<ConstantArrayType>(AT))
+    CAT->getSize().toStringUnsigned(Enc);
+  else
+    Enc += NoSizeEnc; // Global arrays use "*", otherwise it is "".
+  Enc += ':';
+  if (!appendType(Enc, AT->getElementType(), CGM, TSC))
+    return false;
+  Enc += ')';
+  return true;
+}
+
+/// Appends a function encoding to Enc, calling appendType for the return type
+/// and the arguments.
+static bool appendFunctionType(SmallStringEnc &Enc, const FunctionType *FT,
+                             const CodeGen::CodeGenModule &CGM,
+                             TypeStringCache &TSC) {
+  Enc += "f{";
+  if (!appendType(Enc, FT->getReturnType(), CGM, TSC))
+    return false;
+  Enc += "}(";
+  if (const FunctionProtoType *FPT = FT->getAs<FunctionProtoType>()) {
+    // N.B. we are only interested in the adjusted param types.
+    auto I = FPT->param_type_begin();
+    auto E = FPT->param_type_end();
+    if (I != E) {
+      do {
+        if (!appendType(Enc, *I, CGM, TSC))
+          return false;
+        ++I;
+        if (I != E)
+          Enc += ',';
+      } while (I != E);
+      if (FPT->isVariadic())
+        Enc += ",va";
+    } else {
+      if (FPT->isVariadic())
+        Enc += "va";
+      else
+        Enc += '0';
+    }
+  }
+  Enc += ')';
+  return true;
+}
+
+/// Handles the type's qualifier before dispatching a call to handle specific
+/// type encodings.
+static bool appendType(SmallStringEnc &Enc, QualType QType,
+                       const CodeGen::CodeGenModule &CGM,
+                       TypeStringCache &TSC) {
+
+  QualType QT = QType.getCanonicalType();
+
+  appendQualifier(Enc, QT);
+
+  if (const BuiltinType *BT = QT->getAs<BuiltinType>())
+    return appendBuiltinType(Enc, BT);
+
+  if (const ArrayType *AT = QT->getAsArrayTypeUnsafe())
+    return appendArrayType(Enc, AT, CGM, TSC, "");
+
+  if (const PointerType *PT = QT->getAs<PointerType>())
+    return appendPointerType(Enc, PT, CGM, TSC);
+
+  if (const EnumType *ET = QT->getAs<EnumType>())
+    return appendEnumType(Enc, ET, TSC, QT.getBaseTypeIdentifier());
+
+  if (const RecordType *RT = QT->getAsStructureType())
+    return appendRecordType(Enc, RT, CGM, TSC, QT.getBaseTypeIdentifier());
+
+  if (const RecordType *RT = QT->getAsUnionType())
+    return appendRecordType(Enc, RT, CGM, TSC, QT.getBaseTypeIdentifier());
+
+  if (const FunctionType *FT = QT->getAs<FunctionType>())
+    return appendFunctionType(Enc, FT, CGM, TSC);
+
+  return false;
+}
+
+static bool getTypeString(SmallStringEnc &Enc, const Decl *D,
+                          CodeGen::CodeGenModule &CGM, TypeStringCache &TSC) {
+  if (!D)
+    return false;
+
+  if (const FunctionDecl *FD = dyn_cast<FunctionDecl>(D)) {
+    if (FD->getLanguageLinkage() != CLanguageLinkage)
+      return false;
+    return appendType(Enc, FD->getType(), CGM, TSC);
+  }
+
+  if (const VarDecl *VD = dyn_cast<VarDecl>(D)) {
+    if (VD->getLanguageLinkage() != CLanguageLinkage)
+      return false;
+    QualType QT = VD->getType().getCanonicalType();
+    if (const ArrayType *AT = QT->getAsArrayTypeUnsafe()) {
+      // Global ArrayTypes are given a size of '*' if the size is unknown.
+      appendQualifier(Enc, QT);
+      return appendArrayType(Enc, AT, CGM, TSC, "*");
+    }
+    return appendType(Enc, QT, CGM, TSC);
+  }
+  return false;
+}
+
+
 //===----------------------------------------------------------------------===//
 // Driver code
 //===----------------------------------------------------------------------===//

test/CodeGen/xcore-stringtype.c

+// REQUIRES: xcore-registered-target
+// RUN: %clang_cc1 -triple xcore-unknown-unknown -fno-signed-char -fno-common -emit-llvm -o - %s | FileCheck %s
+
+// CHECK: target triple = "xcore-unknown-unknown"
+
+// In the tests below, some types are not supported by the ABI (_Complex,
+// variable length arrays) and will thus emit no meta data.
+// The 33 tests that do emit typstrings are gathered into '!xcore.typestrings'
+// Please see 'Tools Developement Guide' section 2.16.2 for format details:
+// <https://www.xmos.com/download/public/Tools-Development-Guide%28X9114A%29.pdf>
+
+// CHECK: !xcore.typestrings = !{!0, !1, !2, !3, !4, !5, !6, !7, !8, !9, !10,
+// CHECK: !11, !12, !13, !14, !15, !16, !17, !18, !19, !20, !21, !22, !23,
+// CHECK: !24, !25, !26, !27, !28, !29, !30, !31, !32, !33, !34}
+
+
+// test BuiltinType
+// CHECK: !0 = metadata !{void (i1, i8, i8, i8, i16, i16, i16, i32, i32, i32,
+// CHECK:      i32, i32, i32, i64, i64, i64, float, double, double)*
+// CHECK:      @builtinType, metadata !"f{0}(b,uc,uc,sc,ss,us,ss,si,ui,si,sl,
+// CHECK:      ul,sl,sll,ull,sll,ft,d,ld)"}
+void builtinType(_Bool B, char C, unsigned char UC, signed char SC, short S,
+                 unsigned short US, signed short SS, int I, unsigned int UI,
+                 signed int SI, long L, unsigned long UL, signed long SL,
+                 long long LL, unsigned long long ULL, signed long long SLL,
+                 float F, double D, long double LD) {}
+double _Complex Complex; // not supported
+
+
+// test FunctionType & Qualifiers
+// CHECK: !1 = metadata !{void ()* @gI, metadata !"f{0}()"}
+// CHECK: !2 = metadata !{void (...)* @eI, metadata !"f{0}()"}
+// CHECK: !3 = metadata !{void ()* @gV, metadata !"f{0}(0)"}
+// CHECK: !4 = metadata !{void ()* @eV, metadata !"f{0}(0)"}
+// CHECK: !5 = metadata !{void (i32, ...)* @gVA, metadata !"f{0}(si,va)"}
+// CHECK: !6 = metadata !{void (i32, ...)* @eVA, metadata !"f{0}(si,va)"}
+// CHECK: !7 = metadata !{i32* (i32*)* @gQ, metadata !"f{crv:p(cv:si)}(p(cv:si))"}
+// CHECK: !8 = metadata !{i32* (i32*)* @eQ, metadata !"f{crv:p(cv:si)}(p(cv:si))"}
+extern void eI();
+void gI() {eI();};
+extern void eV(void);
+void gV(void) {eV();}
+extern void eVA(int, ...);
+void gVA(int i, ...) {eVA(i);}
+extern const volatile int* volatile restrict const
+    eQ(const volatile int * volatile restrict const);
+const volatile int* volatile restrict const
+    gQ(const volatile int * volatile restrict const i) {return eQ(i);}
+
+
+// test PointerType
+// CHECK: !9 = metadata !{i32* (i32*, i32* (i32*)*)*
+// CHECK:      @pointerType, metadata !"f{p(si)}(p(si),p(f{p(si)}(p(si))))"}
+// CHECK: !10 = metadata !{i32** @EP, metadata !"p(si)"}
+// CHECK: !11 = metadata !{i32** @GP, metadata !"p(si)"}
+extern int* EP;
+int* GP;
+int* pointerType(int *I, int * (*FP)(int *)) {
+  return I? EP : GP;
+}
+
+
+// test ArrayType
+// CHECK: !12 = metadata !{[2 x i32]* (i32*, i32*, [2 x i32]*, [2 x i32]*, i32*)*
+// CHECK:       @arrayType, metadata !"f{p(a(2:si))}(p(si),p(si),p(a(2:si)),
+// CHECK:       p(a(2:si)),p(si))"}
+// CHECK: !13 = metadata !{[0 x i32]* @EA1, metadata !"a(*:si)"}
+// CHECK: !14 = metadata !{[2 x i32]* @EA2, metadata !"a(2:si)"}
+// CHECK: !15 = metadata !{[0 x [2 x i32]]* @EA3, metadata !"a(*:a(2:si))"}
+// CHECK: !16 = metadata !{[3 x [2 x i32]]* @EA4, metadata !"a(3:a(2:si))"}
+// CHECK: !17 = metadata !{[2 x i32]* @GA1, metadata !"a(2:si)"}
+// CHECK: !18 = metadata !{void ([2 x i32]*)* @arrayTypeVariable1,
+// CHECK:       metadata !"f{0}(p(a(2:si)))"}
+// CHECK: !19 = metadata !{void (void ([2 x i32]*)*)* @arrayTypeVariable2,
+// CHECK:       metadata !"f{0}(p(f{0}(p(a(2:si)))))"}
+// CHECK: !20 = metadata !{[3 x [2 x i32]]* @GA2, metadata !"a(3:a(2:si))"}
+extern int EA1[];
+extern int EA2[2];
+extern int EA3[][2];
+extern int EA4[3][2];
+int GA1[2];
+int GA2[3][2];
+extern void arrayTypeVariable1(int[*][2]);
+extern void arrayTypeVariable2( void(*fp)(int[*][2]) );
+extern void arrayTypeVariable3(int[3][*]);                // not supported
+extern void arrayTypeVariable4( void(*fp)(int[3][*]) );   // not supported
+typedef int RetType[2];
+RetType* arrayType(int A1[], int A2[2], int A3[][2], int A4[3][2],
+                   int A5[const volatile restrict static 2]) {
+  if (A1) return &EA1;
+  if (A2) return &EA2;
+  if (A3) return EA3;
+  if (A4) return EA4;
+  if (A5) return &GA1;
+  arrayTypeVariable1(EA4);
+  arrayTypeVariable2(arrayTypeVariable1);
+  arrayTypeVariable3(EA4);
+  arrayTypeVariable4(arrayTypeVariable3);
+  return GA2;
+}
+
+
+// test StructureType
+// CHECK: !21 = metadata !{void (%struct.S1*)* @structureType1, metadata
+// CHECK:       !"f{0}(s(S1){m(ps2){p(s(S2){m(ps3){p(s(S3){m(s1){s(S1){}}})}})}})"}
+// CHECK: !22 = metadata !{void (%struct.S2*)* @structureType2, metadata
+// CHECK:       !"f{0}(s(S2){m(ps3){p(s(S3){m(s1){s(S1){m(ps2){p(s(S2){})}}}})}})"}
+// CHECK: !23 = metadata !{void (%struct.S3*)* @structureType3, metadata
+// CHECK:       !"f{0}(s(S3){m(s1){s(S1){m(ps2){p(s(S2){m(ps3){p(s(S3){})}})}}}})"}
+// CHECK: !24 = metadata !{void (%struct.S4*)* @structureType4, metadata
+// CHECK:       !"f{0}(s(S4){m(s1){s(S1){m(ps2){p(s(S2){m(ps3){p(s(S3){m(s1){s(S1){}}})}})}}}})"}
+// CHECK: !25 = metadata !{%struct.anon* @StructAnon, metadata !"s(){m(A){si}}"}
+// CHECK: !26 = metadata !{i32 (%struct.SB*)* @structureTypeB, metadata
+// CHECK:       !"f{si}(s(SB){m(){b(4:si)},m(){b(2:si)},m(N4){b(4:si)},
+// CHECK:       m(N2){b(2:si)},m(){b(4:ui)},m(){b(4:si)},m(){b(4:c:si)},
+// CHECK:       m(){b(4:c:si)},m(){b(4:cv:si)}})"}
+struct S2;
+struct S1{struct S2 *ps2;};
+struct S3;
+struct S2{struct S3 *ps3;};
+struct S3{struct S1 s1;};
+struct S4{struct S1 s1;};
+void structureType1(struct S1 s1){}
+void structureType2(struct S2 s2){}
+void structureType3(struct S3 s3){}
+void structureType4(struct S4 s4){}
+struct {int A;} StructAnon = {1};
+struct SB{int:4; int:2; int N4:4; int N2:2; unsigned int:4; signed int:4;
+          const int:4; int const :4; volatile const int:4;};
+int structureTypeB(struct SB sb){return StructAnon.A;}
+
+
+// test UnionType
+// CHECK: !27 = metadata !{void (%union.U1*)* @unionType1, metadata
+// CHECK:       !"f{0}(u(U1){m(pu2){p(u(U2){m(pu3){p(u(U3){m(u1){u(U1){}}})}})}})"}
+// CHECK: !28 = metadata !{void (%union.U2*)* @unionType2, metadata
+// CHECK:       !"f{0}(u(U2){m(pu3){p(u(U3){m(u1){u(U1){m(pu2){p(u(U2){})}}}})}})"}
+// CHECK: !29 = metadata !{void (%union.U3*)* @unionType3, metadata
+// CHECK:       !"f{0}(u(U3){m(u1){u(U1){m(pu2){p(u(U2){m(pu3){p(u(U3){})}})}}}})"}
+// CHECK: !30 = metadata !{void (%union.U4*)* @unionType4, metadata
+// CHECK:       !"f{0}(u(U4){m(u1){u(U1){m(pu2){p(u(U2){m(pu3){p(u(U3){m(u1){u(U1){}}})}})}}}})"}
+// CHECK: !31 = metadata !{%union.anon* @UnionAnon, metadata !"u(){m(A){si}}"}
+// CHECK: !32 = metadata !{i32 (%union.UB*)* @unionTypeB, metadata
+// CHECK:       !"f{si}(u(UB){m(N2){b(2:si)},m(N4){b(4:si)},m(){b(2:si)},
+// CHECK:       m(){b(4:c:si)},m(){b(4:c:si)},m(){b(4:cv:si)},m(){b(4:si)},
+// CHECK:       m(){b(4:si)},m(){b(4:ui)}})"}
+union U2;
+union U1{union U2 *pu2;};
+union U3;
+union U2{union U3 *pu3;};
+union U3{union U1 u1;};
+union U4{union U1 u1;};
+void unionType1(union U1 u1) {}
+void unionType2(union U2 u2) {}
+void unionType3(union U3 u3) {}
+void unionType4(union U4 u4) {}
+union UB{int:4; int:2; int N4:4; int N2:2; unsigned int:4; signed int:4;
+         const int:4; int const :4; volatile const int:4;};
+union {int A;} UnionAnon = {1};
+int unionTypeB(union UB ub) {return UnionAnon.A;}
+
+
+// test EnumType
+// CHECK: !33 = metadata !{i32* @EnumAnon, metadata !"e(){m(EA){3}}"}
+// CHECK: !34 = metadata !{i32 (i32)* @enumType, metadata
+// CHECK:       !"f{si}(e(E){m(A){0},m(B){1},m(C){5},m(D){6}})"}
+enum E {A, B, C=5, D};
+enum {EA=3} EnumAnon = EA;
+int enumType(enum E e) {return EnumAnon;}