c++: Pointer-to-member fn conversion with noexcept [PR99374]
authorMarek Polacek <polacek@redhat.com>
Thu, 4 Mar 2021 19:25:01 +0000 (14:25 -0500)
committerMarek Polacek <polacek@redhat.com>
Fri, 5 Mar 2021 23:00:02 +0000 (18:00 -0500)
commit02a3554879001e8f1405d17e096ed68fc3f76975
tree15d3d2b3ec01576d67f0731aa4cad0024211b245
parentc2e64c33d9d903f0a52565ad98300feea0ffc580
c++: Pointer-to-member fn conversion with noexcept [PR99374]

The issue in this PR is that we wrongly reject converting pointers to
member function of incomplete types, one of which has noexcept.  Recall
that pointers (including pointers to member functions) to non-throwing
functions can be implicitly converted to potentially-throwing functions
(but not vice versa).

We reject the conversion when called from can_convert_arg_bad because
standard_conversion can't create such a conversion.  It comes down to
the DERIVED_FROM_P check in the TYPE_PTRMEMFUNC_P block.  It considers
every class derived from itself, but not when the class is incomplete.
But surely we want to reach fnptr_conv_p when tbase is fbase (one of
them could be an alias to the other so use same_type_p instead of ==).

Another approach would be to not perform DERIVED_FROM_P at all when
either tbase or fbase are incomplete (so perhaps something like at the
end of ptr_reasonably_similar).

gcc/cp/ChangeLog:

PR c++/99374
* call.c (standard_conversion): When converting pointers to
member, don't return NULL when the bases are equivalent but
incomplete.

gcc/testsuite/ChangeLog:

PR c++/99374
* g++.dg/cpp1z/noexcept-type23.C: New test.
gcc/cp/call.c
gcc/testsuite/g++.dg/cpp1z/noexcept-type23.C [new file with mode: 0644]