[Bug tree-optimization/93054] ICE in gimple_set_lhs, at gimple.c:1820

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93054

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2020-01-09
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
So we're running into

      notice_special_calls (call_stmt);
      if (!gimple_call_noreturn_p (call_stmt) || !should_remove_lhs_p (*to_p))
        gimple_call_set_lhs (call_stmt, *to_p);
      else if (TREE_CODE (*to_p) == SSA_NAME)
        /* The above is somewhat premature, avoid ICEing later for a
           SSA name w/o a definition.  We may have uses in the GIMPLE IL.
           ???  This doesn't make it a default-def.  */
        SSA_NAME_DEF_STMT (*to_p) = gimple_build_nop ();

and later don't deal with this.