new gfortran 4.1 branch failures

[Steve Kargl]

On Wed, Apr 19, 2006 at 10:58:10PM -0400, Jack Howarth wrote:
> Steve,
>     I find the following for all of the transpose_reshape_r10.f90
> tests...
> 
> /usr/bin/ld: Undefined symbols:
> __gfortran_reshape_16
> __gfortran_transpose_i16

(snip)

> 
> I assume those symbols should be defined in libgfortran, no? I only have...
> 
> nm /sw/lib/gcc4/lib/libgfortran.1.dylib | grep gfortran_transpose
> 0005225c T __gfortran_transpose
> 000318d0 T __gfortran_transpose_c16
> 000314c8 T __gfortran_transpose_c4
> 000316cc T __gfortran_transpose_c8
> 00052250 T __gfortran_transpose_char
> 000310c8 T __gfortran_transpose_i4
> 000312c4 T __gfortran_transpose_i8
> 

Yes, these undefined symbols should be in libgfortran.  I'm
at home, so checking/comparing ChangeLogs between 4.1 and 
trunk is somewhat inconvenient.  I'll assume Thomas will
see this thread, but you might want to ping him. 

-- 
Steve