[CUDA] added include paths for both sides of CUDA compilation.

In order to compile a CUDA file clang must be able to find
include files for both both host and device.

This patch passes AuxToolchain to AddPreprocessingOptions and
uses it to add include paths for the opposite side of compilation.

We also must be able to find CUDA include files. If the driver
found CUDA installation, it adds appropriate include path
to CUDA headers. This can be disabled with '-nocudainc'.

- Added include paths for the opposite side of compilation.
- Added include paths to detected CUDA installation.
- Added -nocudainc to prevent adding CUDA include path.
- Added test cases to verify new functionality.

Differential Revision: http://reviews.llvm.org/D13170

llvm-svn: 253386

8 files changed, 66 insertions, 6 deletions

diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td
index 4ebde1f7979..0e1d3035e42 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -1596,6 +1596,7 @@ def no_pedantic : Flag<["-", "--"], "no-pedantic">, Group<pedantic_Group>;
 def no__dead__strip__inits__and__terms : Flag<["-"], "no_dead_strip_inits_and_terms">;
 def nobuiltininc : Flag<["-"], "nobuiltininc">, Flags<[CC1Option]>,
   HelpText<"Disable builtin #include directories">;
+def nocudainc : Flag<["-"], "nocudainc">;
 def nodefaultlibs : Flag<["-"], "nodefaultlibs">;
 def nofixprebinding : Flag<["-"], "nofixprebinding">;
 def nolibc : Flag<["-"], "nolibc">;
diff --git a/clang/include/clang/Driver/ToolChain.h b/clang/include/clang/Driver/ToolChain.h
index 7626af8fb42..79fe149fec8 100644
--- a/clang/include/clang/Driver/ToolChain.h
+++ b/clang/include/clang/Driver/ToolChain.h
@@ -397,6 +397,10 @@ public:
   virtual void addProfileRTLibs(const llvm::opt::ArgList &Args,
                                 llvm::opt::ArgStringList &CmdArgs) const;
 
+  /// \brief Add arguments to use system-specific CUDA includes.
+  virtual void AddCudaIncludeArgs(const llvm::opt::ArgList &DriverArgs,
+                                  llvm::opt::ArgStringList &CC1Args) const;
+
   /// \brief Return sanitizers which are available in this toolchain.
   virtual SanitizerMask getSupportedSanitizers() const;
 };
diff --git a/clang/lib/Driver/ToolChain.cpp b/clang/lib/Driver/ToolChain.cpp
index 221bed0bf11..80783bdcc4a 100644
--- a/clang/lib/Driver/ToolChain.cpp
+++ b/clang/lib/Driver/ToolChain.cpp
@@ -662,3 +662,6 @@ SanitizerMask ToolChain::getSupportedSanitizers() const {
     Res |= CFIICall;
   return Res;
 }
+
+void ToolChain::AddCudaIncludeArgs(const ArgList &DriverArgs,
+                                   ArgStringList &CC1Args) const {}
diff --git a/clang/lib/Driver/ToolChains.cpp b/clang/lib/Driver/ToolChains.cpp
index ed223219f65..34504732a12 100644
--- a/clang/lib/Driver/ToolChains.cpp
+++ b/clang/lib/Driver/ToolChains.cpp
@@ -4104,6 +4104,15 @@ void Linux::AddClangCXXStdlibIncludeArgs(const ArgList &DriverArgs,
   }
 }
 
+void Linux::AddCudaIncludeArgs(const ArgList &DriverArgs,
+                               ArgStringList &CC1Args) const {
+  if (DriverArgs.hasArg(options::OPT_nocudainc))
+    return;
+
+  if (CudaInstallation.isValid())
+    addSystemInclude(DriverArgs, CC1Args, CudaInstallation.getIncludePath());
+}
+
 bool Linux::isPIEDefault() const { return getSanitizerArgs().requiresPIE(); }
 
 SanitizerMask Linux::getSupportedSanitizers() const {
diff --git a/clang/lib/Driver/ToolChains.h b/clang/lib/Driver/ToolChains.h
index 17e790ffd25..b1b670feb85 100644
--- a/clang/lib/Driver/ToolChains.h
+++ b/clang/lib/Driver/ToolChains.h
@@ -784,6 +784,8 @@ public:
   void AddClangCXXStdlibIncludeArgs(
       const llvm::opt::ArgList &DriverArgs,
       llvm::opt::ArgStringList &CC1Args) const override;
+  void AddCudaIncludeArgs(const llvm::opt::ArgList &DriverArgs,
+                          llvm::opt::ArgStringList &CC1Args) const override;
   bool isPIEDefault() const override;
   SanitizerMask getSupportedSanitizers() const override;
   void addProfileRTLibs(const llvm::opt::ArgList &Args,
diff --git a/clang/lib/Driver/Tools.cpp b/clang/lib/Driver/Tools.cpp
index 3c593fc66b8..5df8efafd5a 100644
--- a/clang/lib/Driver/Tools.cpp
+++ b/clang/lib/Driver/Tools.cpp
@@ -278,7 +278,8 @@ void Clang::AddPreprocessingOptions(Compilation &C, const JobAction &JA,
                                     const Driver &D, const ArgList &Args,
                                     ArgStringList &CmdArgs,
                                     const InputInfo &Output,
-                                    const InputInfoList &Inputs) const {
+                                    const InputInfoList &Inputs,
+                                    const ToolChain *AuxToolChain) const {
   Arg *A;
 
   CheckPreprocessingOptions(D, Args);
@@ -470,12 +471,26 @@ void Clang::AddPreprocessingOptions(Compilation &C, const JobAction &JA,
   // OBJCPLUS_INCLUDE_PATH - system includes enabled when compiling ObjC++.
   addDirectoryList(Args, CmdArgs, "-objcxx-isystem", "OBJCPLUS_INCLUDE_PATH");
 
+  // Optional AuxToolChain indicates that we need to include headers
+  // for more than one target. If that's the case, add include paths
+  // from AuxToolChain right after include paths of the same kind for
+  // the current target.
+
   // Add C++ include arguments, if needed.
-  if (types::isCXX(Inputs[0].getType()))
+  if (types::isCXX(Inputs[0].getType())) {
     getToolChain().AddClangCXXStdlibIncludeArgs(Args, CmdArgs);
+    if (AuxToolChain)
+      AuxToolChain->AddClangCXXStdlibIncludeArgs(Args, CmdArgs);
+  }
 
   // Add system include arguments.
   getToolChain().AddClangSystemIncludeArgs(Args, CmdArgs);
+  if (AuxToolChain)
+      AuxToolChain->AddClangCXXStdlibIncludeArgs(Args, CmdArgs);
+
+  // Add CUDA include arguments, if needed.
+  if (types::isCuda(Inputs[0].getType()))
+    getToolChain().AddCudaIncludeArgs(Args, CmdArgs);
 }
 
 // FIXME: Move to target hook.
@@ -3262,12 +3277,12 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
   CmdArgs.push_back("-triple");
   CmdArgs.push_back(Args.MakeArgString(TripleStr));
 
+  const ToolChain *AuxToolChain = nullptr;
   if (IsCuda) {
     // FIXME: We need a (better) way to pass information about
     // particular compilation pass we're constructing here. For now we
     // can check which toolchain we're using and pick the other one to
     // extract the triple.
-    const ToolChain *AuxToolChain;
     if (&getToolChain() == C.getCudaDeviceToolChain())
       AuxToolChain = C.getCudaHostToolChain();
     else if (&getToolChain() == C.getCudaHostToolChain())
@@ -4085,7 +4100,8 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
   //
   // FIXME: Support -fpreprocessed
   if (types::getPreprocessedType(InputType) != types::TY_INVALID)
-    AddPreprocessingOptions(C, JA, D, Args, CmdArgs, Output, Inputs);
+    AddPreprocessingOptions(C, JA, D, Args, CmdArgs, Output, Inputs,
+                            AuxToolChain);
 
   // Don't warn about "clang -c -DPIC -fPIC test.i" because libtool.m4 assumes
   // that "The compiler can only warn and ignore the option if not recognized".
diff --git a/clang/lib/Driver/Tools.h b/clang/lib/Driver/Tools.h
index 7a7c642e356..95c2063695c 100644
--- a/clang/lib/Driver/Tools.h
+++ b/clang/lib/Driver/Tools.h
@@ -57,7 +57,8 @@ private:
                                const Driver &D, const llvm::opt::ArgList &Args,
                                llvm::opt::ArgStringList &CmdArgs,
                                const InputInfo &Output,
-                               const InputInfoList &Inputs) const;
+                               const InputInfoList &Inputs,
+                               const ToolChain *AuxToolChain) const;
 
   void AddAArch64TargetArgs(const llvm::opt::ArgList &Args,
                             llvm::opt::ArgStringList &CmdArgs) const;
diff --git a/clang/test/Driver/cuda-detect.cu b/clang/test/Driver/cuda-detect.cu
index 31e8494027a..182e379df5b 100644
--- a/clang/test/Driver/cuda-detect.cu
+++ b/clang/test/Driver/cuda-detect.cu
@@ -1,10 +1,34 @@
 // REQUIRES: clang-driver
 // REQUIRES: x86-registered-target
+// REQUIRES: nvptx-registered-target
 //
+// # Check that we properly detect CUDA installation.
 // RUN: %clang -v --target=i386-unknown-linux \
-// RUN:   --sysroot=/tmp/no-cuda-there 2>&1 | FileCheck %s -check-prefix NOCUDA
+// RUN:   --sysroot=%S/no-cuda-there 2>&1 | FileCheck %s -check-prefix NOCUDA
+// RUN: %clang -v --target=i386-unknown-linux \
+// RUN:   --sysroot=%S/Inputs/CUDA 2>&1 | FileCheck %s
 // RUN: %clang -v --target=i386-unknown-linux \
 // RUN:   --cuda-path=%S/Inputs/CUDA/usr/local/cuda 2>&1 | FileCheck %s
 
+// Verify that CUDA include path gets added
+// RUN: %clang -### -v --target=i386-unknown-linux --cuda-gpu-arch=sm_35 \
+// RUN:   --cuda-path=%S/Inputs/CUDA/usr/local/cuda %s 2>&1 \
+// RUN:   | FileCheck %s -check-prefix COMMON -check-prefix CUDAINC
+// Verify that -nocudainc disables CUDA include paths.
+// RUN: %clang -### -v --target=i386-unknown-linux --cuda-gpu-arch=sm_35 \
+// RUN:   -nocudainc --cuda-path=%S/Inputs/CUDA/usr/local/cuda %s 2>&1 \
+// RUN:   | FileCheck %s -check-prefix COMMON -check-prefix NOCUDAINC
+// We should not add any CUDA include paths if there's no valid CUDA installation
+// RUN: %clang -### -v --target=i386-unknown-linux --cuda-gpu-arch=sm_35 \
+// RUN:   --cuda-path=%S/no-cuda-there %s 2>&1 \
+// RUN:   | FileCheck %s -check-prefix COMMON -check-prefix NOCUDAINC
+
 // CHECK: Found CUDA installation: {{.*}}/Inputs/CUDA/usr/local/cuda
 // NOCUDA-NOT: Found CUDA installation:
+
+// COMMON: "-triple" "nvptx-nvidia-cuda"
+// COMMON-SAME: "-fcuda-is-device"
+// CUDAINC-SAME: "-internal-isystem" "{{.*}}/Inputs/CUDA/usr/local/cuda/include"
+// NOCUDAINC-NOT: "-internal-isystem" "{{.*}}/cuda/include"
+// COMMON-SAME: "-x" "cuda"
+