[libgfortran,patch] Widen default formats for formatted output of floating-point numbers

[FX Coudert]

Attached patch follows the point of Dominique who showed that our  
default formats for formatted floating-point output aren't wide  
enough to allow writing and reading back values.  I've widened the  
formats for kinds 4, 8 and 10, after checking on x86_64-linux; the  
testcase is also an extension of what Dominique posted in the PR. I  
haven't fixed the case of real(kind=16) yet, because it's more  
complicated there (see, once again, Dominique's testing in the PR).  
It seemed that there is no reason to wait for the real(kind=16) issue  
to be resolved to check in that patch, so I submit it here for review.

Bootstrapped on x86_64-linux, regtest is still progressing (but a  
previous version did regtest fine), OK to commit if no error show up?

Thanks,
FX


:ADDPATCH libgfortran:
-------------- next part --------------
A non-text attachment was scrubbed...
Name: widen_formats.ChangeLog
Type: application/octet-stream
Size: 312 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20071001/c1a69396/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: widen_formats.diff
Type: application/octet-stream
Size: 5285 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20071001/c1a69396/attachment-0001.obj>