https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128227
--- Comment #8 from Josef Melcr <jmelcr at gcc dot gnu.org> --- Created attachment 66036 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=66036&action=edit patch fixing the issue The callgraph seems consistent as far as I can tell, the issue is the check itself. Redirecting an internall callback does produce an edge without a carrying edge which is not pointing towards an assume body, but I think that it's ok in this case. Attached patch should fix it, currently in testing.
