https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85921
--- Comment #6 from John Simon <gcc at mailinator dot com> --- > I doubt this warrants a change on GCC side, given that a workaround is simple. If I understood it correctly the local "solution" would be to fix a system header (which is often not possible because of missing write access). The "simple workaround" would be to change a GCC distributed file before the build. Requesting this from a user who doesn't know much about what he is requested to do - if he founds the reason for the issue at all - seems to be the wrong thing. If it is unlikely that undefining noinline would do any harm I'd propose to include the patch (I can provide a signed one, if necessary) [I just don't know f the change is "correct" - but the quick hack seemed to work, still bootstrapping...]. Simon Sobisch
