[llvm-bugs] [Bug 26594] New: Compilation error with typdef, template expansion and conditional

via llvm-bugs llvm-bugs at lists.llvm.org
Fri Feb 12 10:24:51 PST 2016


https://llvm.org/bugs/show_bug.cgi?id=26594

            Bug ID: 26594
           Summary: Compilation error with typdef, template expansion and
                    conditional
           Product: clang
           Version: 3.7
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: C++14
          Assignee: unassignedclangbugs at nondot.org
          Reporter: joe_woodbury at yahoo.com
                CC: llvm-bugs at lists.llvm.org
    Classification: Unclassified

Created attachment 15892
  --> https://llvm.org/bugs/attachment.cgi?id=15892&action=edit
The source code cited in the description.

I'm not sure how to summarize, this the following builds on the latest MSC 2015
update 1 and g++ 5.3, but not clang 3.7 or 3.7.1. The following code has been
severely reduced from the original. There may be a better way to do this and
I'm very open to the possibility that clang is right and Microsoft and g++ are
wrong.

If TEST_TYPEDEF is defined, the generates the following error on clang 3.7.x.
The static_assert is correct; the question is why is it generating that
operator:


[jwoodbury at joew test]$ clang -c -std=c++14 test.cpp
test.cpp:9:5: error: static_assert failed "uint8_t, uint16_t, uint32_t
required."
    static_assert(std::is_same<T, uint8_t>::value || std::is_same<T,
uint16_t>::value ||
    ^            
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test.cpp:71:20: note: in instantiation of template class
'bit_pack<bit_pack<unsigned int>::prohibit_operator_uint16_conversion>'
      requested here
const lmap bit16 = bit_pack_factory<16>::nth_bit();

/////////////////////////////////////

#include <cstddef>
#include <cstdint>
#include <limits>
#include <type_traits>

template <typename T>
struct bit_pack
{
    static_assert(std::is_same<T, uint8_t>::value || std::is_same<T,
uint16_t>::value ||
        std::is_same<T, uint32_t>::value, "uint8_t, uint16_t, uint32_t
required.");

    class prohibit_operator_uint16_conversion {};
    class prohibit_operator_uint32_conversion {};

    typedef T Type;

    bit_pack() : the_backing_store(0)
    {
    }

    bit_pack(T const& value) : the_backing_store(value)
    {
    }

    operator bit_pack<typename std::conditional<
        sizeof(T) <= sizeof(uint16_t), uint16_t,
        prohibit_operator_uint16_conversion>::type>() const
    {
        return bit_pack<uint16_t>(the_backing_store);
    }

    operator bit_pack<typename std::conditional<
        sizeof(T) <= sizeof(uint32_t), uint32_t,
        prohibit_operator_uint32_conversion>::type>() const
    {
        return bit_pack<uint32_t>(the_backing_store);
    }

    T the_backing_store;
};

#define TEST_TYPEDEF
#ifdef TEST_TYPEDEF

typedef bit_pack<uint8_t>  bmap;
typedef bit_pack<uint16_t> wmap;
typedef bit_pack<uint32_t> lmap;

#define BIT_PACK_NTH_BIT_MAP_TYPE \
    typename std::conditional<bit_index < 8, bmap, typename
std::conditional<bit_index < 16, wmap, lmap>::type>::type

#else

#define BIT_PACK_NTH_BIT_MAP_TYPE \
    typename std::conditional<bit_index < 8, uint8_t, typename
std::conditional<bit_index < 16, uint16_t, uint32_t>::type>::type

#endif

template <size_t bit_index>
struct bit_pack_factory
{
    static BIT_PACK_NTH_BIT_MAP_TYPE nth_bit()
    {
        // With clang and TEST_TYPEDEF, this does not give a warning
        BIT_PACK_NTH_BIT_MAP_TYPE const result = 1u << bit_index;
        return result;
    }
};

#ifdef TEST_TYPEDEF
const lmap bit16 = bit_pack_factory<16>::nth_bit();
#else
bit_pack<uint32_t> bit_pack_int = bit_pack_factory<16>::nth_bit();
#endif

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20160212/3c65f936/attachment-0001.html>


More information about the llvm-bugs mailing list