[Fortran, committed to trunk] Fixed powerpc "unsigned" printing of

John David Anglin dave@hiauly1.hia.nrc.ca
Thu Apr 11 13:15:00 GMT 2002


> > On the latter [powerpc-unknown-linux-gnu], even the following is 
> > incorrect:
> > 
> >       INTEGER*1 I
> >       I = -9
> >       PRINT*,I
> >       END
> > 
> > which prints 247 ;-)
> 
> I fixed this with the attached patch [applied to trunk].  Make bootstrap
> (C and Fortran), make check (Fortran only) and make install on
> powerpc-unknown-linux-gnu.

Did that fix f90-intrinsic-numeric.f on the powerpc?

I did a liitle more digging on the pa.  It beginning to look like some
kind of aliasing problem.  It important to note that the test passes
at -O0 but not at -O1.  At -O0, there are insns that write the i*2
values of j and ja to virtual stack slots.  However, at -O1, it appears
the write for ja is deleted and as a result ka is read from an unitialized
slot.  I got as far as ffestd_stmt_pass_ in trying to locate the problem.

Dave
-- 
J. David Anglin                                  dave.anglin@nrc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6605)



More information about the Gcc-patches mailing list