[patch, libgfortran] Final patch for PR29277, LF conversion to CR-LF

[Bud Davis]

jerry,

OK.

--bud

--- Jerry DeLisle <jvdelisle@verizon.net> wrote:

> Danny Smith wrote:
> >> 2006-10-17  Jerry DeLisle 
> <jvdelisle@gcc.gnu.org>
> >>
> >> 	PR libgfortran/29277
> >> 	* io/write.c (write_a): Add conversion of LF to
> LF-CR 
> >> for systems with
> >> 	#define HAVE_CRLF.
> >>
> > 
> 
> I see I need to change the ChangeLog to from LF-CR
> to CR-LF  :)
> 
> > Looks good to me.
> > 
> > Before this patch hex dump of output file is:
> > lookatme.txt:
> > 00000000  7265 636f 7264 310a 0a0a 0a0a 7265 636f
> record1.....reco
> > 00000010  7264 310a 0a0a 0a72 6563 6f0a          
> rd1....reco.
> > 
> > 
> > After patch
> > lookatme.txt:
> > 00000000  7265 636f 7264 310d 0a0d 0a0d 0a0d 0a0d
> record1.........
> > 00000010  0a72 6563 6f72 6431 0d0a 0d0a 0d0a 0d0a
> .record1........
> > 00000020  7265 636f 0d0a                         
> reco..
> > 
> > 
> 
> This looks great!  Thanks Danny for testing.
> 
> > No new libgortran failures on mingw32.
> > 
> > Thanks
> > Danny
> > 
> > Umm, would you consider a command line option (in
> next stage 1) that
> > would override the default format mode to allow,
> eg, using 0_BINARY 
> > newlines  on targets such as mingw32.
> > 
> >
> I thought about a command line option may be useful.
>  We will do that for 4.3 
> with concurrence from the team.
> 
> All that remains is approval from one of our
> Gfortran maintainers.  I have an OK 
> from Mark Mitchell.  But after this we go strict on
> release rules.
> 
> Regards,
> 
> Jerry
> 
>