https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52985
--- Comment #8 from Andrew Pinski <pinskia at gcc dot gnu.org> --- (In reply to Sam James from comment #6) > (In reply to Andrew Pinski from comment #5) > > Created attachment 60858 [details] > > Easy running testcase > > With this, we can make it a regression (5.5 works)? I am not sure though because the testcase in comment #2 always failed. Plus I think a slightly modified testcase to have the ?: be `t1 ? a : b` has always failed. Basically `t1 ? a : a` was reduced down to just `a` before GCC 6 and not exposing the other issue.
