[Bug tree-optimization/128267] [15/16/17 Regression] wrong code at -O2 with -march=x86-64-v4: a signed char shift induction is 64 again instead of 0 after 256 iterations

[rguenth at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128267

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Interestingly Andreas testcase miscompiles with -march=x86-64-v4 but doesn't
SIGILL on Zen2, it does not miscompile (or vectorize) with -march=znver2
though.