diff options
| author | Jonathan Wakely <jwakely@redhat.com> | 2019-02-05 15:45:24 +0000 |
|---|---|---|
| committer | Jonathan Wakely <redi@gcc.gnu.org> | 2019-02-05 15:45:24 +0000 |
| commit | ed99e818e02cd7b0c10c9e8cba7fec8ee17b0ca7 (patch) | |
| tree | 05d62d0809c00286cbb2770e05c70fdde6e5dfc1 /gcc | |
| parent | 7356fbde50dd4bfd55eaa0bed31b31da55d1461c (diff) | |
| download | gcc-ed99e818e02cd7b0c10c9e8cba7fec8ee17b0ca7.zip gcc-ed99e818e02cd7b0c10c9e8cba7fec8ee17b0ca7.tar.gz gcc-ed99e818e02cd7b0c10c9e8cba7fec8ee17b0ca7.tar.bz2 | |
PR libstdc++/89194 untangle is_convertible and is_nothrow_convertible
The additional logic added to __is_convertible_helper in order to
support is_nothrow_convertible makes some uses of is_convertible
ill-formed. This appears to be due to PR c++/87603, but can be avoided
just by defining a separate helper for is_nothrow_convertible. The same
problems are likely to still exist for is_nothrow_convertible, but that
is new and so won't cause regressions for existing users of
is_convertible.
PR libstdc++/89194
* include/std/type_traits (__is_convertible_helper)
(__is_convertible_helper<_From, _To, false>): Revert changes to
support is_nothrow_convertible.
(__is_nt_convertible_helper): New helper.
(is_nothrow_convertible): Use __is_nt_convertible_helper.
From-SVN: r268543
Diffstat (limited to 'gcc')
0 files changed, 0 insertions, 0 deletions