https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89677

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
We're analyzing a CHREC of type float, sth the affine machinery isn't really
prepared to handle.  The rev in question makes us more defensive but in the
end we're lucky.

Testing a real patch for the issue.

Reply via email to