[Bug tree-optimization/94216] [10 Regression] ICE in maybe_canonicalize_mem_ref_addr, at gimple-fold.c:4899 since r10-7237-g4e3d3e40726e1b68bf52fa205c68495124ea60b8

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94216

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rgue...@gcc.gnu.org>:

https://gcc.gnu.org/g:73bc09fa8c6b973a928a599498caa66a25c8bc8d

commit r10-7272-g73bc09fa8c6b973a928a599498caa66a25c8bc8d
Author: Richard Biener <rguent...@suse.de>
Date:   Thu Mar 19 10:15:52 2020 +0100

    middle-end/94216 fix another build_fold_addr_expr use

    2020-03-19  Richard Biener  <rguent...@suse.de>

            PR middle-end/94216
            * fold-const.c (fold_binary_loc): Avoid using
            build_fold_addr_expr when we really want an ADDR_EXPR.

            * g++.dg/torture/pr94216.C: New testcase.