https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71261
--- Comment #4 from Anton Mitrokhin <anton.mitrokhin at phystech dot edu> --- PR70902 does not hang on knl or broadwell, while this test case hangs on knl, broadwell and it seems to hang on skylake-avx512 as well. So, I am not sure if this case triggers the same bug as the PR70902, or it is a different one (which looks similar). It is also very difficult to reduce this test case, as you can't tell if the compiler hangs or it is an another 40-second compilation. Overall, 40 seconds for compilation over 10 lines of code seems like a bug as it is, but I will try to generate a better (hanging :) ) test case by tomorrow!
