[FrontEnd] Allow overriding the default C/C++ -std via CMake vars

[mgorny]

Summary

Provide two new CMake cache variables -- CLANG_DEFAULT_STD_C
and CLANG_DEFAULT_STD_CXX -- that can be used to override the default
C/ObjC and C++/ObjC++ standards appropriately. They can be set to one of
the identifiers from LangStandards.def, or left unset (the default) to
respect the current platform default.

This option is mostly intended for compiler vendors that do not wish to
continue defaulting to C++98. We are preparing a new Gentoo release that
requires GCC 6, and therefore defaults to -std=gnu++14. Since we want to
maintain as much compatibility between GCC and clang by default, we
would like to adjust the default for clang appropriately.

The code relies on explicit identifiers rather than the string aliases
for simplicity. This saves us from the necessity of parsing aliases at
build-time or adding additional processing at runtime. For the latter
case, it also adds trivial value check -- if incorrect value is passed,
the code simply fails to compile through referencing an undefined
constant.

If the variable is used to redefine the default standard, the explicit
value overrides the special case for PS4. It is done this way mostly
following other kinds of variables where 'platform defaults' are
redefined.

I'm going to submit more test fixes that add explicit -std=
whenever necessary once this is approved.

[mgorny]

Ping.

[mgorny]

Ping again.

[mgorny]

Ping for the third time.

[lebedev.ri]

clang 6.0 defaults to C++14, BTW.

[compnerd]

Is there a need for this given the changes for 6.0?

[rnk]

Let's do this. I thought we'd already have a higher C++ standard version by now, but apparently not.

[mgorny]

@rnk, could you confirm the rebased patch? I'm not sure if I should override InputKind::RenderScript as well.

[rnk]

Yep, looks good. I'd leave renderscript alone. I think it's deprecated.

CMakeLists.txt

 set(DEFAULT_SYSROOT "" CACHE PATH
   "Default <path> to all compiler invocations for --sysroot=<path>." )
 
 set(ENABLE_LINKER_BUILD_ID OFF CACHE BOOL "pass --build-id to ld")
 
 set(ENABLE_X86_RELAX_RELOCATIONS OFF CACHE BOOL
     "enable x86 relax relocations by default")
 
+# TODO: verify the values against LangStandards.def?
+set(CLANG_DEFAULT_STD_C "" CACHE STRING
+  "Default standard to use for C/ObjC code (IDENT from LangStandards.def, empty for platform default)")
+set(CLANG_DEFAULT_STD_CXX "" CACHE STRING
+  "Default standard to use for C++/ObjC++ code (IDENT from LangStandards.def, empty for platform default)")
+
 set(CLANG_DEFAULT_LINKER "" CACHE STRING
   "Default linker to use (linker name or absolute path, empty for platform default)")
 
 set(CLANG_DEFAULT_CXX_STDLIB "" CACHE STRING
   "Default C++ stdlib to use (\"libstdc++\" or \"libc++\", empty for platform default")
 if (NOT(CLANG_DEFAULT_CXX_STDLIB STREQUAL "" OR
         CLANG_DEFAULT_CXX_STDLIB STREQUAL "libstdc++" OR
         CLANG_DEFAULT_CXX_STDLIB STREQUAL "libc++"))

include/clang/Config/config.h.cmake

 /* This generated file is for internal use. Do not include it from headers. */
 
 #ifdef CLANG_CONFIG_H
 #error config.h can only be included once
 #else
 #define CLANG_CONFIG_H
 
 /* Bug report URL. */
 #define BUG_REPORT_URL "${BUG_REPORT_URL}"
 
 /* Default linker to use. */
 #define CLANG_DEFAULT_LINKER "${CLANG_DEFAULT_LINKER}"
 
+/* Default C/ObjC standard to use. */
+#cmakedefine CLANG_DEFAULT_STD_C LangStandard::lang_${CLANG_DEFAULT_STD_C}
+
+/* Default C++/ObjC++ standard to use. */
+#cmakedefine CLANG_DEFAULT_STD_CXX LangStandard::lang_${CLANG_DEFAULT_STD_CXX}
+
 /* Default C++ stdlib to use. */
 #define CLANG_DEFAULT_CXX_STDLIB "${CLANG_DEFAULT_CXX_STDLIB}"
 
 /* Default runtime library to use. */
 #define CLANG_DEFAULT_RTLIB "${CLANG_DEFAULT_RTLIB}"
 
 /* Default OpenMP runtime used by -fopenmp. */
 #define CLANG_DEFAULT_OPENMP_RUNTIME "${CLANG_DEFAULT_OPENMP_RUNTIME}"

lib/Frontend/CompilerInvocation.cpp

if (LangStd == LangStandard::lang_unspecified) {
     case InputKind::OpenCL:
       LangStd = LangStandard::lang_opencl10;
       break;
     case InputKind::CUDA:
       LangStd = LangStandard::lang_cuda;
       break;
     case InputKind::Asm:
     case InputKind::C:
+#if defined(CLANG_DEFAULT_STD_C)
+      LangStd = CLANG_DEFAULT_STD_C;
+#else
       // The PS4 uses C99 as the default C standard.
       if (T.isPS4())
         LangStd = LangStandard::lang_gnu99;
       else
         LangStd = LangStandard::lang_gnu11;
+#endif
       break;
     case InputKind::ObjC:
+#if defined(CLANG_DEFAULT_STD_C)
+      LangStd = CLANG_DEFAULT_STD_C;
+#else
       LangStd = LangStandard::lang_gnu11;
+#endif
       break;
     case InputKind::CXX:
     case InputKind::ObjCXX:
+#if defined(CLANG_DEFAULT_STD_CXX)
+      LangStd = CLANG_DEFAULT_STD_CXX;
+#else
       // The PS4 uses C++11 as the default C++ standard.
       if (T.isPS4())
         LangStd = LangStandard::lang_gnucxx11;
       else
         LangStd = LangStandard::lang_gnucxx98;
+#endif
       break;
     case InputKind::RenderScript:
       LangStd = LangStandard::lang_c99;
       break;
     }
   }
 
   const LangStandard &Std = LangStandard::getLangStandardForKind(LangStd);