This program constexpr bool f() { union U{ std::string s; constexpr ~U(){ s.~basic_string(); } } u{}; return true; } static_assert( f() ); is accepted in MSVC and Clang with libc++. But in GCC it produces the error: error: non-constant condition for static assertion in 'constexpr' expansion of 'f()' in 'constexpr' expansion of '(& u)->f()::U::~U()' in 'constexpr' expansion of '((f()::U*)this)->f()::U::s.std::__cxx11::basic_string<char>::~basic_string()' /opt/compiler-explorer/gcc-13.1.0/include/c++/13.1.0/bits/basic_string.h:803:19: in 'constexpr' expansion of '((std::__cxx11::basic_string<char>*)this)->std::__cxx11::basic_string<char>::_M_dispose()' error: accessing 'std::__cxx11::basic_string<char>::<unnamed union>::_M_allocated_capacity' member instead of initialized 'std::__cxx11::basic_string<char>::<unnamed union>::_M_local_buf' member in constant expression Online demo: https://gcc.godbolt.org/z/bbf4Yo3v9
Here is a slightly reduced testcase (for a slightly different issue still dealing with unions): ``` struct str1 { // bool a; char *var; union { char t[15]; int allocated; }; constexpr str1() : var(new char[2]) { t[0] = 0; } constexpr ~str1() {if (var != t) delete[] var; } }; typedef str1 str; constexpr bool f1() { str t{}; return true; } static_assert( f1() ); constexpr bool f() { union U{ str s; constexpr ~U(){ s.~str(); } } u{}; return true; } static_assert( f() ); ```
It works if we activate the union's member: constexpr bool f() { union U{ std::string s = {}; constexpr ~U(){ s.~basic_string(); } } u{}; return true; } static_assert( f() ); So perhaps this is a dup of PR110245?
*** Bug 110900 has been marked as a duplicate of this bug. ***
Here's a repro without the std::string inside a union. It is the SSO union inside the string that causes the error. https://gcc.godbolt.org/z/T8oM8vYnq ``` #include <string> template <class T, class I, class S, class F> constexpr T fold(T init, I i, S s, F f) { while (true) { if (i == s) return init; else init = f(std::move(init), *i++); } } constexpr char v[] = {'a', 'b', 'c'}; static_assert(fold(std::string(), std::begin(v), std::end(v), [](std::string acc, char v) { acc.push_back(v); return acc; }) == "abc"); int main() {} ``` <source>:18:23: error: non-constant condition for static assertion 14 | static_assert(fold(std::string(), std::begin(v), std::end(v), | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 15 | [](std::string acc, char v) { | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 16 | acc.push_back(v); | ~~~~~~~~~~~~~~~~~ 17 | return acc; | ~~~~~~~~~~~ 18 | }) == "abc"); | ~~~^~~~~~~~ <source>:18:32: in 'constexpr' expansion of 'fold(T, I, S, F) [with T = std::__cxx11::basic_string<char>; I = const char*; S = const char*; F = <lambda(std::string, char)>](std::begin<const char, 3>(v), std::end<const char, 3>(v), (<lambda closure object><lambda(std::string, char)>(), <lambda(std::string, char)>()))' <source>:18:32: in 'constexpr' expansion of 'std::__cxx11::basic_string<char>((* & std::move<__cxx11::basic_string<char>&>(init)))' <source>:18:23: error: accessing 'std::__cxx11::basic_string<char>::<unnamed union>::_M_allocated_capacity' member instead of initialized 'std::__cxx11::basic_string<char>::<unnamed union>::_M_local_buf' member in constant expression ASM generation compiler returned: 1
As a means to work around #4 in static asserts, making your string long enough that it avoids the SSO will allow the compiler to accept it.
In case it is of help, here's an even smaller repro that clang reports the error in libstdc++. For whatever reason gcc does not notice the libstdc++ bug here. https://gcc.godbolt.org/z/Phndafeoe ``` #include <string> struct S { constexpr ~S() {} std::string s; }; constexpr std::string foo(const S& s) { return s.s; } static_assert(foo(S("hi")) == "hi"); // Fails. static_assert(foo(S("a longer string works")) == "a longer string works"); int main() {} ``` The error: ``` <source>:11:15: error: static assertion expression is not an integral constant expression static_assert(foo(S("hi")) == "hi"); ^~~~~~~~~~~~~~~~~~~~ /opt/compiler-explorer/gcc-12.2.0/lib/gcc/x86_64-linux-gnu/12.2.0/../../../../include/c++/12.2.0/bits/basic_string.h:356:10: note: assignment to member '_M_local_buf' of union with no active member is not allowed in a constant expression __c = _CharT(); ^ /opt/compiler-explorer/gcc-12.2.0/lib/gcc/x86_64-linux-gnu/12.2.0/../../../../include/c++/12.2.0/bits/basic_string.tcc:229:4: note: in call to '&S(("hi")).s->_M_use_local_data()' _M_use_local_data(); ^ /opt/compiler-explorer/gcc-12.2.0/lib/gcc/x86_64-linux-gnu/12.2.0/../../../../include/c++/12.2.0/bits/basic_string.h:642:2: note: in call to '&S(("hi")).s->_M_construct(&"hi"[0], &"hi"[2], {{}})' _M_construct(__s, __end, forward_iterator_tag()); ^ <source>:11:21: note: in call to 'basic_string(&"hi"[0], std::allocator<char>())' static_assert(foo(S("hi")) == "hi"); ^ 1 error generated. ```
That bug in comment 6 has already been fixed in gcc-12.3, by the last patch for PR 103295. The one in comment 4 is different, and will be fixed by Nathaniel's latest patch: https://gcc.gnu.org/pipermail/gcc-patches/2023-September/631642.html
It seems the comment #4 testcase is still rejected on trunk after/despite r14-4334-g28adad7a32ed92?
Odd, I thought I'd checked it when testing r14-4334-g28adad7a32ed92. Seems like the same issue as PR 112642 though (which has a minimized version without std::string).