diff options
| author | Marek Polacek <polacek@redhat.com> | 2026-01-28 17:30:06 -0500 |
|---|---|---|
| committer | Marek Polacek <polacek@redhat.com> | 2026-01-29 08:42:04 -0500 |
| commit | 101f2970adc0a7ac5785e40eb8ad63facb4f582a (patch) | |
| tree | 08cc77488340a8cb55d6a076055865a6d394cadd /libjava/testsuite/libjava.lang/PR25535.java | |
| parent | bc10a49c37501e0603236fd9ca1f3a63bec24cc7 (diff) | |
| download | gcc-101f2970adc0a7ac5785e40eb8ad63facb4f582a.zip gcc-101f2970adc0a7ac5785e40eb8ad63facb4f582a.tar.gz gcc-101f2970adc0a7ac5785e40eb8ad63facb4f582a.tar.bz2 | |
c++/reflection: enhance splice error checking
As discussed in
<https://gcc.gnu.org/pipermail/gcc-patches/2026-January/705756.html>:
> Absolutely, I mean anything other than REFLECT_EXPR or error_mark_node
> should be an error.
I'm not giving another error for error_mark_node since we've already
complained (but make sure we have).
gcc/cp/ChangeLog:
* reflect.cc (class_members_of): Adjust comment.
(splice): Give an error if the expression doesn't evaluate to
a REFLECT_EXPR_P.
Reviewed-by: Jason Merrill <jason@redhat.com>
Diffstat (limited to 'libjava/testsuite/libjava.lang/PR25535.java')
0 files changed, 0 insertions, 0 deletions