The assert macro usage should not produce 3 clang-tidy warnings!

[ClausKlein]

Is this compiler header problem or a clang-tidy-7 error?

packages/platform/src/RRCP/Connection.cpp|146 col 13| warning: do not implicitly decay an array into a pointer; consider using gsl::array_view or an explicit cast instead [cppcoreguidelines-pro-bounds-array-to-pointer-decay]
||             assert(getServer() != nullptr);
||             ^
/usr/include/assert.h|95 col 51| note: expanded from macro 'assert'
||       : __assert_fail (#expr, __FILE__, __LINE__, __ASSERT_FUNCTION))
||                                                   ^
/usr/include/assert.h|129 col 30| note: expanded from macro '__ASSERT_FUNCTION'
|| #   define __ASSERT_FUNCTION    __extension__ __PRETTY_FUNCTION__
||                                 ^

[jwakely]

Clang-tidy should suppress that warning. It's unhelpful (it seems to be complaining about converting a compiler-defined null-terminated array into a const char* which is guaranteed to be safe) and users can't do anything about it anyway.

[ClausKlein]

But a

static_cast<const char *>(__PRETTY_FUNCTION__)

would also help and may be correct, or not?

[jwakely]

It adds nothing, and it wouldn't help when assert is compiled by C code.

[ClausKlein]

And what is this cppcoreguidelines-pro-bounds-array-to-pointer-decay check for if it should be suppressed ?

[jwakely]

N.B. re your issue title, there is no such thing as the std::assert macro, because macros don't use scopes.

The guidelines are for application developers and library writers, not for the implementation. assert is part of the C++ (or C) implementation and so it is useless for clang-tidy to check it and tell you to fix it.

Just because the check doesn't make sense for assert doesn't mean it's useless elsewhere, and it's silly to infer that that's what I meant.

[phillipjohnston]

I haven't run into this before, which doesn't mean much. I can think of two possible problems:

1. The build is not correctly including system headers using -isystem
2. clang-tidy is being invoked with --system-headers

I also see this clang-tidy change, which seems to be relevant and also abandoned: https://reviews.llvm.org/D31130

[ClausKlein]

I used clang-7 on debian stable, no really actual

[ClausKlein]

No, it happens with clang-9 on OS X too:

part of compile_commands.json:

{
  "directory": "/Users/clausklein/Workspace/cpp/.build-samples-Debug",
  "command": "/opt/local/bin/clang++  -DGSL_THROW_ON_CONTRACT_VIOLATION  -g   -Wall -Wextra -Wpedantic -Werror -Wno-unused-parameter -Wno-unused-variable -Wno-unknown-warning-option -std=c++17 -o CMakeFiles/ClonableBase.dir/ClonableBase.o -c /Users/clausklein/Workspace/cpp/samples/ClonableBase.cpp",
  "file": "/Users/clausklein/Workspace/cpp/samples/ClonableBase.cpp"
},

Claus-MBP:samples clausklein$ clang-tidy --version
LLVM (http://llvm.org/):
  LLVM version 9.0.0svn
  Optimized build with assertions.
  Default target: x86_64-apple-darwin15.6.0
  Host CPU: penryn


/Users/clausklein/Workspace/cpp/samples/ClonableBase.cpp|84 col 5| warning: do not implicitly decay an array into a pointer; consider using gsl::array_view or an explicit cast instead [cppcoreguidelines-pro-bounds-array-to-pointer-decay]
||     assert(origin.number() == 1);
||     ^
/usr/include/assert.h|93 col 47| note: expanded from macro 'assert'
||     (__builtin_expect(!(e), 0) ? __assert_rtn(__func__, __FILE__, __LINE__, #e) : (void)0)
||                                               ^
/Users/clausklein/Workspace/cpp/samples/ClonableBase.cpp|91 col 5| warning: do not implicitly decay an array into a pointer; consider using gsl::array_view or an explicit cast instead [cppcoreguidelines-pro-bounds-array-to-pointer-decay]
||     assert(copy->number() == 1);
||     ^
/usr/include/assert.h|93 col 47| note: expanded from macro 'assert'
||     (__builtin_expect(!(e), 0) ? __assert_rtn(__func__, __FILE__, __LINE__, #e) : (void)0)
||                                               ^

[phillipjohnston]

As Jonathan mentioned, the proper place to follow up is with the clang-tidy project. I recommend chiming in on this MR: https://reviews.llvm.org/D31130

[ClausKlein]

see too https://bugs.llvm.org/show_bug.cgi?id=28480
and https://bugs.llvm.org/show_bug.cgi?id=32239
OK, thanks