The following code shows that `decltype(auto)` is deduced to the wrong type when used in template code for an expression like `(var.static_data_member>)`. ```cpp #include <cstddef> #include <type_traits> #include <utility> template <typename T> struct type_displayer; #define TYPE_DISPLAY(x) type_displayer<decltype(x)> test_obj; struct SkipField {}; struct TestSkip { template <typename T, std::size_t> struct FIELD; int value; static SkipField unused; template <typename T> struct FIELD<T, 0> { T &&obj; FIELD(T &&ref) : obj(std::forward<T>(ref)) { } decltype(auto) value() { return (obj.value); } static constexpr const char *name() { return "value"; } }; template <typename T> struct FIELD<T, 0 + 1> { T &&obj; FIELD(T &&ref) : obj(std::forward<T>(ref)) { } decltype(auto) value() { return (obj.unused); } decltype((obj.unused)) valueAlt() { return (obj.unused); } static constexpr const char *name() { return "unused"; } }; }; int main() { TestSkip s; // decltype((TestSkip::FIELD<TestSkip&, 0>(s).value())) is int& // decltype((TestSkip::FIELD<TestSkip&, 1>(s).value())) is SkipField, but it should be SkipField& // decltype((TestSkip::FIELD<TestSkip&, 1>(s).valueAlt())) is SkipField& } ``` This is a simplified usage scenario, where I want to use a macro to generate code to emulate static reflection. Here is the problematic part: ```cpp decltype(auto) value() { return (obj.unused); } ``` This function returns `SkipField`, instead of the expected `SkipField&`. I have verified that clang does what I want. Currently I have to avoid `decltype(auto)` as a workaround: ```cpp decltype((obj.unused)) valueAlt() { return (obj.unused); } ``` (I used the macro TYPE_DISPLAY to check the type of an expression.)
Reduced testcase that is accepted but should not be: struct TestSkip { static int unused; }; template <class T> struct FIELD { TestSkip *obj; decltype(auto) value() { return (obj->unused); } }; int main() { TestSkip s; int sf; int i; decltype((FIELD<int>().value())) b; } Confirmed.
The master branch has been updated by Patrick Palka <ppalka@gcc.gnu.org>: https://gcc.gnu.org/g:e4f1ea5d8b96fc9fbe4fc0e0e0a4938ceeef092f commit r14-560-ge4f1ea5d8b96fc9fbe4fc0e0e0a4938ceeef092f Author: Patrick Palka <ppalka@redhat.com> Date: Sun May 7 11:57:22 2023 -0400 c++: parenthesized -> resolving to static data member [PR98283] Here we're neglecting to propagate parenthesized-ness when the member access (this->m) resolves to a static data member (and thus finish_class_member_access_expr yields a VAR_DECL instead of a COMPONENT_REF). PR c++/98283 gcc/cp/ChangeLog: * pt.cc (tsubst_copy_and_build) <case COMPONENT_REF>: Propagate REF_PARENTHESIZED_P more generally via force_paren_expr. * semantics.cc (force_paren_expr): Document default argument. gcc/testsuite/ChangeLog: * g++.dg/cpp1y/paren6.C: New test.
Fixed for GCC 14, thanks for the bug report