Ambiguous call of ldexp in standalone_test_arithmetic_cpp_rational.cpp in generic_interconvert #697
Description
Building version 1.83 with gcc14.2, this test test errors with:
Am I missing some relevant change?
./boost/multiprecision/detail/generic_interconvert.hpp:375:18: error: call of overloaded 'ldexp(__float128&, int)' is ambiguous
375 | result = ldexp(result, static_cast<int>(-shift));
| ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from libs/multiprecision/test/test.hpp:11:
/usr/include/c++/14.2/cmath:311:3: note: candidate: 'constexpr long double std::ldexp(long double, int)'
311 | ldexp(long double __x, int __exp)
| ^~~~~
/usr/include/c++/14.2/cmath:307:3: note: candidate: 'constexpr float std::ldexp(float, int)'
307 | ldexp(float __x, int __exp)
| ^~~~~
In file included from /usr/include/features.h:524,
from /usr/include/c++/14.2/x86_64-unknown-linux-gnu/bits/os_defines.h:39,
from /usr/include/c++/14.2/x86_64-unknown-linux-gnu/bits/c++config.h:683,
from /usr/include/c++/14.2/functional:48,
from libs/multiprecision/test/test_arithmetic.hpp:10:
/usr/include/bits/mathcalls.h:119:1: note: candidate: 'double ldexp(double, int)'
119 | __MATHCALL (ldexp,, (_Mdouble_ __x, int __exponent));
| ^~~~~~~~~~
Details
=> Testing: boost::multiprecision
Performing configuration checks
- default address-model : 64-bit (cached) [1]
- default architecture : x86 (cached) [1]
- cxx11_rvalue_references : yes (cached) [2]
- cxx11_template_aliases : yes (cached) [2]
- cxx11_hdr_array : yes (cached) [2]
- cxx11_allocator : yes (cached) [2]
- cxx11_constexpr : yes (cached) [2]
- cxx11_explicit_conversion_operators : yes (cached) [2]
- cxx11_ref_qualifiers : yes (cached) [2]
- cxx11_hdr_functional : yes (cached) [2]
- cxx11_variadic_templates : yes (cached) [2]
- cxx11_user_defined_literals : yes (cached) [2]
- cxx11_decltype : yes (cached) [2]
- cxx11_static_assert : yes (cached) [2]
- cxx11_defaulted_functions : yes (cached) [2]
- cxx11_noexcept : yes (cached) [2]
- cxx11_hdr_type_traits : yes (cached) [2]
- cxx14_decltype_auto : yes (cached) [2]
- cxx14_generic_lambdas : yes (cached) [2]
- cxx14_return_type_deduction : yes (cached) [2]
- cxx14_variable_templates : yes (cached) [2]
- cxx11_rvalue_references : yes (cached) [3]
- cxx11_template_aliases : yes (cached) [3]
- cxx11_hdr_array : yes (cached) [3]
- cxx11_allocator : yes (cached) [3]
- cxx11_constexpr : yes (cached) [3]
- cxx11_explicit_conversion_operators : yes (cached) [3]
- cxx11_ref_qualifiers : yes (cached) [3]
- cxx11_hdr_functional : yes (cached) [3]
- cxx11_variadic_templates : yes (cached) [3]
- cxx11_user_defined_literals : yes (cached) [3]
- cxx11_decltype : yes (cached) [3]
- cxx11_static_assert : yes (cached) [3]
- cxx11_defaulted_functions : yes (cached) [3]
- cxx11_noexcept : yes (cached) [3]
- cxx11_hdr_type_traits : yes (cached) [3]
- cxx14_decltype_auto : yes (cached) [3]
- cxx14_generic_lambdas : yes (cached) [3]
- cxx14_return_type_deduction : yes (cached) [3]
- cxx14_variable_templates : yes (cached) [3]
- has_float128 builds : yes (cached) [3]
- has_gmp builds : no (cached) [3]
- cxx11_rvalue_references : yes (cached) [4]
- cxx11_template_aliases : yes (cached) [4]
- cxx11_hdr_array : yes (cached) [4]
- cxx11_allocator : yes (cached) [4]
- cxx11_constexpr : yes (cached) [4]
- cxx11_explicit_conversion_operators : yes (cached) [4]
- cxx11_ref_qualifiers : yes (cached) [4]
- cxx11_hdr_functional : yes (cached) [4]
- cxx11_variadic_templates : yes (cached) [4]
- cxx11_user_defined_literals : yes (cached) [4]
- cxx11_decltype : yes (cached) [4]
- cxx11_static_assert : yes (cached) [4]
- cxx11_defaulted_functions : yes (cached) [4]
- cxx11_noexcept : yes (cached) [4]
- cxx11_hdr_type_traits : yes (cached) [4]
- cxx14_decltype_auto : yes (cached) [4]
- cxx14_generic_lambdas : yes (cached) [4]
- cxx14_return_type_deduction : yes (cached) [4]
- cxx14_variable_templates : yes (cached) [4]
- has_gmp builds : no (cached) [4]
- has_mpfr builds : no (cached) [3]
- has_mpfr builds : no (cached) [4]
- has_tommath builds : no (cached) [3]
- has_tommath builds : no (cached) [4]
- has_mpc builds : no (cached) [3]
- has_mpc builds : no (cached) [4]
- has_mpfi builds : no (cached) [3]
- has_mpfi builds : no (cached) [4]
- has_intel_quad builds : no (cached) [3]
- has_intel_quad builds : no (cached) [4]
- cxx17_if_constexpr : yes (cached) [3]
- cxx17_if_constexpr : yes (cached) [4]
- has_float128 builds : yes (cached) [4]
- Sanitizer CI run : no (cached) [3]
- Sanitizer CI run : no (cached) [4]
- has std::atomic_ref : no (cached) [3]
- has -Wl,--no-undefined : yes (cached) [3]
- has statx : yes (cached) [3]
- has init_priority attribute : yes (cached) [3]
- has stat::st_blksize : yes (cached) [3]
- has stat::st_mtim : yes (cached) [3]
- has stat::st_mtimensec : no (cached) [3]
- has stat::st_mtimespec : no (cached) [3]
- has stat::st_birthtim : no (cached) [3]
- has stat::st_birthtimensec : no (cached) [3]
- has stat::st_birthtimespec : no (cached) [3]
- has fdopendir(O_NOFOLLOW) : yes (cached) [3]
- has dirent::d_type : yes (cached) [3]
- has POSIX *at APIs : yes (cached) [3]
- compiler supports SSE2 : yes (cached) [3]
- compiler supports SSE4.1 : yes (cached) [3]
- cxx11_hdr_random : yes (cached) [3]
- cxx14_constexpr : yes (cached) [3]
- has_is_constant_evaluated builds : yes (cached) [3]
- has_constexpr_limits builds : yes (cached) [3]
- cpp_lib_bit_cast : no (cached) [3]
- cpp_lib_bit_cast : no (cached) [4]
- has_eigen builds : no (cached) [3]
- has_eigen builds : no (cached) [4]
- has_gmp builds : no (cached) [2]
- has_mpfr builds : no (cached) [2]
- has_mpfr builds : no (cached) [5]
- has_mpfr builds : no (cached) [6]
- has_gmp builds : no (cached) [5]
- has_gmp builds : no (cached) [6]
- cxx11_rvalue_references : yes (cached) [5]
- cxx11_template_aliases : yes (cached) [5]
- cxx11_hdr_array : yes (cached) [5]
- cxx11_allocator : yes (cached) [5]
- cxx11_constexpr : yes (cached) [5]
- cxx11_explicit_conversion_operators : yes (cached) [5]
- cxx11_ref_qualifiers : yes (cached) [5]
- cxx11_hdr_functional : yes (cached) [5]
- cxx11_variadic_templates : yes (cached) [5]
- cxx11_user_defined_literals : yes (cached) [5]
- cxx11_decltype : yes (cached) [5]
- cxx11_static_assert : yes (cached) [5]
- cxx11_defaulted_functions : yes (cached) [5]
- cxx11_noexcept : yes (cached) [5]
- cxx11_hdr_type_traits : yes (cached) [5]
- cxx14_decltype_auto : yes (cached) [5]
- cxx14_generic_lambdas : yes (cached) [5]
- cxx14_return_type_deduction : yes (cached) [5]
- cxx14_variable_templates : yes (cached) [5]
- has_big_obj builds : no (cached) [3]
- cxx11_lambdas : yes (cached) [3]
- cxx11_nullptr : yes (cached) [3]
- cxx11_numeric_limits : yes (cached) [3]
- BOOST_COMP_GNUC >= 4.3.0 : yes (cached) [7]
- cxx11_hdr_atomic : yes (cached) [3]
- cxx11_hdr_thread : yes (cached) [3]
- cxx11_hdr_atomic : yes (cached) [4]
- cxx11_hdr_thread : yes (cached) [4]
- cxx11_auto_declarations : yes (cached) [3]
- lockfree boost::atomic_flag : yes (cached) [3]
- has_f2c builds : no (cached) [3]
- has_f2c builds : no (cached) [4]
[1] gcc-14
[2] gcc-14/release/architecture-x86/python-3.13/threading-multi/visibility-hidden
[3] gcc-14/release/python-3.13/threading-multi/visibility-hidden
[4] gcc-14/release/build-no/python-3.13/threading-multi/visibility-hidden
[5] gcc-14/release/optimization-space/python-3.13/threading-multi/visibility-hidden
[6] gcc-14/release/build-no/optimization-space/python-3.13/threading-multi/visibility-hidden
[7] gcc-14/release/link-static/python-3.13/threading-multi/visibility-hidden
...patience...
...patience...
...patience...
...patience...
...found 10034 targets...
...updating 4 targets...
gcc.compile.c++ bin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden/standalone_test_arithmetic_cpp_rational.o
In file included from ./boost/multiprecision/detail/generic_interconvert.hpp:12,
from ./boost/multiprecision/number.hpp:12,
from libs/multiprecision/test/test.hpp:16,
from libs/multiprecision/test/test_arithmetic.hpp:16,
from libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.cpp:9:
./boost/multiprecision/detail/default_ops.hpp: In instantiation of 'constexpr typename boost::multiprecision::scalar_result_from_possible_complex<typename boost::multiprecision::detail::expression<tag, Arg1, Arg2, Arg3, Arg4>::result_type>::type boost::multiprecision::norm(const detail::expression<tag, Arg1, Arg2, Arg3, Arg4>&) [with tag = detail::multiply_immediates; A1 = number<backends::rational_adaptor<backends::cpp_int_backend<> > >; A2 = int; A3 = void; A4 = void; typename scalar_result_from_possible_complex<typename detail::expression<tag, Arg1, Arg2, Arg3, Arg4>::result_type>::type = number<backends::rational_adaptor<backends::cpp_int_backend<> > >; typename detail::expression<tag, Arg1, Arg2, Arg3, Arg4>::result_type = number<backends::rational_adaptor<backends::cpp_int_backend<> > >]':
libs/multiprecision/test/test_arithmetic.hpp:2870:4: required from 'typename std::enable_if<(boost::multiprecision::number_category<Num>::value != boost::multiprecision::number_kind_complex)>::type test_members(Real) [with Real = boost::multiprecision::number<boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> > >; typename std::enable_if<(boost::multiprecision::number_category<Num>::value != boost::multiprecision::number_kind_complex)>::type = void]'
2870 | BOOST_CHECK_EQUAL(norm(a * 1), a * a);
libs/multiprecision/test/test_arithmetic.hpp:3391:16: required from 'void test() [with Real = boost::multiprecision::number<boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> > >]'
3391 | test_members(a);
| ~~~~~~~~~~~~^~~
libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.cpp:35:45: required from here
35 | test<boost::multiprecision::cpp_rational>();
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
./boost/multiprecision/detail/default_ops.hpp:2366:54: warning: moving a temporary object prevents copy elision [-Wpessimizing-move]
2366 | return std::move(norm(static_cast<number_type>(v)));
| ^
./boost/multiprecision/detail/default_ops.hpp:2366:54: note: remove 'std::move' call
./boost/multiprecision/detail/generic_interconvert.hpp: In instantiation of 'typename std::enable_if<(boost::multiprecision::is_number<T>::value || std::is_floating_point<_Tp>::value)>::type boost::multiprecision::detail::generic_convert_rational_to_float_imp(To&, Integer&, Integer&, const std::integral_constant<bool, false>&) [with To = __float128; Integer = boost::multiprecision::number<boost::multiprecision::backends::cpp_int_backend<> >; typename std::enable_if<(boost::multiprecision::is_number<T>::value || std::is_floating_point<_Tp>::value)>::type = void]':
./boost/multiprecision/detail/generic_interconvert.hpp:404:41: required from 'void boost::multiprecision::detail::generic_convert_rational_to_float(To&, const From&) [with To = __float128; From = boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> >]'
404 | generic_convert_rational_to_float_imp(result, n, d, dispatch_tag());
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
./boost/multiprecision/rational_adaptor.hpp:393:70: required from 'typename std::enable_if<(boost::multiprecision::number_category<Num>::value == boost::multiprecision::number_kind_floating_point)>::type boost::multiprecision::backends::eval_convert_to(R*, const rational_adaptor<IntBackend>&) [with R = __float128; IntBackend = cpp_int_backend<>; typename std::enable_if<(boost::multiprecision::number_category<Num>::value == boost::multiprecision::number_kind_floating_point)>::type = void]'
393 | ::boost::multiprecision::detail::generic_convert_rational_to_float(*result, backend);
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~
./boost/multiprecision/number.hpp:1046:22: required from 'constexpr void boost::multiprecision::number<Backend, ExpressionTemplates>::convert_to_imp(T*) const [with T = __float128; Backend = boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> >; boost::multiprecision::expression_template_option ExpressionTemplates = boost::multiprecision::et_on]'
1046 | eval_convert_to(result, m_backend);
| ~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~
./boost/multiprecision/number.hpp:1063:21: required from 'constexpr T boost::multiprecision::number<Backend, ExpressionTemplates>::convert_to() const [with T = __float128; Backend = boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> >; boost::multiprecision::expression_template_option ExpressionTemplates = boost::multiprecision::et_on]'
1063 | convert_to_imp(&result);
| ~~~~~~~~~~~~~~^~~~~~~~~
libs/multiprecision/test/test_arithmetic.hpp:2253:4: required from 'void test_mixed(const std::integral_constant<bool, true>&) [with Real = boost::multiprecision::number<boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> > >; Num = __float128]'
2253 | BOOST_CHECK_EQUAL(Real(n1).template convert_to<Num>(), n1);
libs/multiprecision/test/test_arithmetic.hpp:3216:35: required from 'void test() [with Real = boost::multiprecision::number<boost::multiprecision::backends::rational_adaptor<boost::multiprecision::backends::cpp_int_backend<> > >]'
3216 | test_mixed<Real, __float128>(tag);
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~
libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.cpp:35:45: required from here
35 | test<boost::multiprecision::cpp_rational>();
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
./boost/multiprecision/detail/generic_interconvert.hpp:375:18: error: call of overloaded 'ldexp(__float128&, int)' is ambiguous
375 | result = ldexp(result, static_cast<int>(-shift));
| ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from libs/multiprecision/test/test.hpp:11:
/usr/include/c++/14.2/cmath:311:3: note: candidate: 'constexpr long double std::ldexp(long double, int)'
311 | ldexp(long double __x, int __exp)
| ^~~~~
/usr/include/c++/14.2/cmath:307:3: note: candidate: 'constexpr float std::ldexp(float, int)'
307 | ldexp(float __x, int __exp)
| ^~~~~
In file included from /usr/include/features.h:524,
from /usr/include/c++/14.2/x86_64-unknown-linux-gnu/bits/os_defines.h:39,
from /usr/include/c++/14.2/x86_64-unknown-linux-gnu/bits/c++config.h:683,
from /usr/include/c++/14.2/functional:48,
from libs/multiprecision/test/test_arithmetic.hpp:10:
/usr/include/bits/mathcalls.h:119:1: note: candidate: 'double ldexp(double, int)'
119 | __MATHCALL (ldexp,, (_Mdouble_ __x, int __exponent));
| ^~~~~~~~~~
"g++" -fvisibility-inlines-hidden -fstack-clash-protection -D_FORTIFY_SOURCE=2 -mtune=generic -O2 -pipe -ffile-prefix-map=/builddir/boost-1.83.0=. -fPIC -m64 -pthread -O3 -finline-functions -Wno-inline -Wall -fvisibility=hidden -Wall -Wextra -fstack-clash-protection -D_FORTIFY_SOURCE=2 -mtune=generic -O2 -pipe -ffile-prefix-map=/builddir/boost-1.83.0=. -DBOOST_ALL_NO_LIB=1 -DNDEBUG -I"." -I"libs/multiprecision/include" -c -o "bin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden/standalone_test_arithmetic_cpp_rational.o" "libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.cpp"
...failed gcc.compile.c++ bin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden/standalone_test_arithmetic_cpp_rational.o...
...skipped <pbin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden>standalone_test_arithmetic_cpp_rational for lack of <pbin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden>standalone_test_arithmetic_cpp_rational.o...
...skipped <pbin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden>standalone_test_arithmetic_cpp_rational.run for lack of <pbin.v2/libs/multiprecision/test/standalone_test_arithmetic_cpp_rational.test/gcc-14/release/threading-multi/visibility-hidden>standalone_test_arithmetic_cpp_rational...
...failed updating 1 target...
...skipped 3 targets...
=> ERROR: do_check: boost::multiprecision: failure