[Bug sanitizer/70147] testcase from hana testsuite gets miscompiled with -fsanitize=undefined

["jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70147

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jason at gcc dot gnu.org

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Indeed, in std::basic_istream::basic_istream
  MEM[(struct  &)this] = {CLOBBER};
  this->_vptr.basic_istream = 0B;
  D.162921 = this->_vptr.basic_istream;
  D.162922 = D.162921 + 18446744073709551592;
  D.162923 = MEM[(long int *)D.162922];
  D.162924 = (sizetype) D.162923;
  D.162925 = this + D.162924;
  D.162925->D.41813._vptr.ios_base = 0B;
looks wrong, because it dereferences NULL - 24.
If I look what the actual normal initialization does later on:
        D.162930 = *__vtt_parm;
        this->_vptr.basic_istream = D.162930;
        D.162921 = this->_vptr.basic_istream;
        D.162922 = D.162921 + 18446744073709551592;
        D.162923 = MEM[(long int *)D.162922];
        D.162924 = (sizetype) D.162923;
        D.162925 = this + D.162924;
        D.162931 = MEM[(const void * *)__vtt_parm + 8B];
        D.162925->D.41813._vptr.ios_base = D.162931;
then we probably need to pretend that we've stored *__vtt_parm into
this->_vptr.basic_istream (e.g. replace that read with *__vtt_parm).
Now to figure out what is __vtt_parm...