Created attachment 176384 [details] Fix for lang/gcc5
Thanks for looking into this, Kenneth! I am not familiar with this port, nor even Eigen, so I hope some other committer will grab this, but one comment/suggestion on this patch: ++// fix for lang/gcc5 ++#if __GNUC__ == 5 && __GNUC_MINOR__ == 4 ++#undef minor ++#endif Why check for __GNUC_MINOR__, especially with ==? What if the lang/gcc5 port is updated to version 5.5? Wouldn't it be best to just check for __GNUC__, ignoring __GNUC_MINOR__, and perhaps even __GNUC__ >= 5?
Ha! Neither was I, but it's a prereq for building Kalzium, which is a neat little program part of kdeedu. I don't disagree with your suggest to test for __GNUC__ >= 5, problem is though I have only tested with lang/gcc5 and do not know if lang/gcc6 has the same issue or not. Thanks for your support, Gerald!
(In reply to Gerald Pfeifer from comment #1) Hi, Gerald. I have tested lang/gcc6 and determined it does NOT have the same issue as lang/gcc5, so my test for __GNUC__ == 5 still stands as valid. I can consider removing __GNUC_MINOR__ == 4 however.
Created attachment 176419 [details] Fix for lang/gcc5 Removed test for __GNUC_MINOR__
A commit references this bug: Author: miwi Date: Sat May 6 15:17:38 UTC 2017 New revision: 440273 URL: https://svnweb.freebsd.org/changeset/ports/440273 Log: - Fix build with gcc5 PR: 213985 Changes: head/science/avogadro/files/patch-libavogadro_src_extensions_crystallography_crystallographyextension.cpp
Isn't this going to break again when lang/gcc5 goes from version 5.4 to 5.5? It appears you applied the original patch instead of the one after my feedback? Cf. comment #1 and comment #4.
Assignee timeout. Give back to the pool.
GCC 7 is the default now, GCC 5 is deprecated. Can this be closed or do we need to amend the previous commit?