[CUDA] Copy host builtin types to NVPTXTargetInfo.

Summary:
Host and device types must match, otherwise when we pass values back and
forth between the host and device, we will get the wrong result.

This patch makes NVPTXTargetInfo inherit most of its type information
from the host's target info.

Reviewers: rsmith

Subscribers: cfe-commits, jhen, tra

Differential Revision: http://reviews.llvm.org/D19346

llvm-svn: 268131

3 files changed, 71 insertions, 2 deletions

diff --git a/clang/lib/Basic/Targets.cpp b/clang/lib/Basic/Targets.cpp
index 0a87cdf061e..1df35ab8465 100644
--- a/clang/lib/Basic/Targets.cpp
+++ b/clang/lib/Basic/Targets.cpp
@@ -65,6 +65,9 @@ static void defineCPUMacros(MacroBuilder &Builder, StringRef CPUName,
     Builder.defineMacro("__tune_" + CPUName + "__");
 }
 
+static TargetInfo *AllocateTarget(const llvm::Triple &Triple,
+                                  const TargetOptions &Opts);
+
 //===----------------------------------------------------------------------===//
 // Defines specific to certain operating systems.
 //===----------------------------------------------------------------------===//
@@ -1627,7 +1630,7 @@ class NVPTXTargetInfo : public TargetInfo {
   } GPU;
 
 public:
-  NVPTXTargetInfo(const llvm::Triple &Triple, const TargetOptions &)
+  NVPTXTargetInfo(const llvm::Triple &Triple, const TargetOptions &Opts)
       : TargetInfo(Triple) {
     BigEndian = false;
     TLSSupported = false;
@@ -1639,6 +1642,65 @@ public:
     NoAsmVariants = true;
     // Set the default GPU to sm20
     GPU = GK_SM20;
+
+    // If possible, get a TargetInfo for our host triple, so we can match its
+    // types.
+    llvm::Triple HostTriple(Opts.HostTriple);
+    if (HostTriple.isNVPTX())
+      return;
+    std::unique_ptr<TargetInfo> HostTarget(
+        AllocateTarget(llvm::Triple(Opts.HostTriple), Opts));
+    if (!HostTarget) {
+      return;
+    }
+
+    PointerWidth = HostTarget->getPointerWidth(/* AddrSpace = */ 0);
+    PointerAlign = HostTarget->getPointerAlign(/* AddrSpace = */ 0);
+    BoolWidth = HostTarget->getBoolWidth();
+    BoolAlign = HostTarget->getBoolAlign();
+    IntWidth = HostTarget->getIntWidth();
+    IntAlign = HostTarget->getIntAlign();
+    HalfWidth = HostTarget->getHalfWidth();
+    HalfAlign = HostTarget->getHalfAlign();
+    FloatWidth = HostTarget->getFloatWidth();
+    FloatAlign = HostTarget->getFloatAlign();
+    DoubleWidth = HostTarget->getDoubleWidth();
+    DoubleAlign = HostTarget->getDoubleAlign();
+    LongWidth = HostTarget->getLongWidth();
+    LongAlign = HostTarget->getLongAlign();
+    LongLongWidth = HostTarget->getLongLongWidth();
+    LongLongAlign = HostTarget->getLongLongAlign();
+    MinGlobalAlign = HostTarget->getMinGlobalAlign();
+    DefaultAlignForAttributeAligned =
+        HostTarget->getDefaultAlignForAttributeAligned();
+    SizeType = HostTarget->getSizeType();
+    IntMaxType = HostTarget->getIntMaxType();
+    PtrDiffType = HostTarget->getPtrDiffType(/* AddrSpace = */ 0);
+    IntPtrType = HostTarget->getIntPtrType();
+    WCharType = HostTarget->getWCharType();
+    WIntType = HostTarget->getWIntType();
+    Char16Type = HostTarget->getChar16Type();
+    Char32Type = HostTarget->getChar32Type();
+    Int64Type = HostTarget->getInt64Type();
+    SigAtomicType = HostTarget->getSigAtomicType();
+    ProcessIDType = HostTarget->getProcessIDType();
+
+    UseBitFieldTypeAlignment = HostTarget->useBitFieldTypeAlignment();
+    UseZeroLengthBitfieldAlignment =
+        HostTarget->useZeroLengthBitfieldAlignment();
+    UseExplicitBitFieldAlignment = HostTarget->useExplicitBitFieldAlignment();
+    ZeroLengthBitfieldBoundary = HostTarget->getZeroLengthBitfieldBoundary();
+
+    // Properties intentionally not copied from host:
+    // - LargeArrayMinWidth, LargeArrayAlign: Not visible across the
+    //   host/device boundary.
+    // - SuitableAlign: Not visible across the host/device boundary, and may
+    //   correctly be different on host/device, e.g. if host has wider vector
+    //   types than device.
+    // - LongDoubleWidth, LongDoubleAlign: nvptx's long double type is the same
+    //   as its double type, but that's not necessarily true on the host.
+    //   TODO: nvcc emits a warning when using long double on device; we should
+    //   do the same.
   }
   void getTargetDefines(const LangOptions &Opts,
                         MacroBuilder &Builder) const override {
diff --git a/clang/lib/Frontend/CompilerInstance.cpp b/clang/lib/Frontend/CompilerInstance.cpp
index cac73befff0..d6f4749db24 100644
--- a/clang/lib/Frontend/CompilerInstance.cpp
+++ b/clang/lib/Frontend/CompilerInstance.cpp
@@ -836,8 +836,9 @@ bool CompilerInstance::ExecuteAction(FrontendAction &Act) {
 
   // Create TargetInfo for the other side of CUDA compilation.
   if (getLangOpts().CUDA && !getFrontendOpts().AuxTriple.empty()) {
-    std::shared_ptr<TargetOptions> TO(new TargetOptions);
+    auto TO = std::make_shared<TargetOptions>();
     TO->Triple = getFrontendOpts().AuxTriple;
+    TO->HostTriple = getTarget().getTriple().str();
     setAuxTarget(TargetInfo::CreateTargetInfo(getDiagnostics(), TO));
   }
 
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp b/clang/lib/Frontend/CompilerInvocation.cpp
index 4bef159f3bf..02a570076b9 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2155,6 +2155,12 @@ bool CompilerInvocation::CreateFromArgs(CompilerInvocation &Res,
       LangOpts.ObjCExceptions = 1;
   }
 
+  // During CUDA device-side compilation, the aux triple is the triple used for
+  // host compilation.
+  if (LangOpts.CUDA && LangOpts.CUDAIsDevice) {
+    Res.getTargetOpts().HostTriple = Res.getFrontendOpts().AuxTriple;
+  }
+
   // FIXME: Override value name discarding when asan or msan is used because the
   // backend passes depend on the name of the alloca in order to print out
   // names.