https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114065
--- Comment #81 from Eric Botcazou <ebotcazou at gcc dot gnu.org> --- > Eric, I have removed the hardcoded 8 from comment 73, but I fail to > understand whether you also dislike the complex conditional. Anyhow, I am > fine with any of the following suggestions. Thanks, let's go with the last one, since my (limited) understanding is that the formula indeed works in all cases; I'm going to apply the changes.
