https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86306
--- Comment #6 from Jonathan Wakely <redi at gcc dot gnu.org> --- (In reply to zhonghao from comment #5) > Another bug report (https://bugs.llvm.org/show_bug.cgi?id=21835) also used > the code. In particular, David Majnemer said that the following code, > > typedef _Atomic(int) T; > T a = T(); > > is a reasonable test case. It's reasonable FOR CLANG. It's not reasonable for g++. > Please note that the use the the command, clang -fno-crash-diagnostics > -std=c++11 -xc++ -c -emit-llvm, to trigger the bug. The command line > indicates the programers of clang handle the code as a C++ code, instead of > C code. Yes, because it's a non-standard extension that Clang supports. GCC does not support it. GCC does not have to support it. It's non-standard. Bugs in non-standard Clang extensions are not bugs in GCC. If you don't understand that, stop reporting bugs. > I just want to improve the quality of both compilers. I do not understand > why my reported bugs make you mad. Because you are just blindly copying things from one bugzilla to the other, without making any effort to check if what you are reporting is sensible. If you want to improve GCC then stop wasting my time so I can actually work on GCC. If you spend more time making good bug reports and fewer bad reports, then GCC devs will get more work done.
