-malign-double failure on i686-pc-linux-gnuaout
Steve Edwards
sgedward@korime.demon.co.uk
Thu May 31 22:24:00 GMT 2007
On Thu, 31 May 2007 14:31:39 -0700
"Andrew Pinski" <pinskia@gmail.com> wrote:
> On 5/31/07, Steve Edwards <sgedward@korime.demon.co.uk> wrote:
> > I am not sure if this is known or not, but I ran into a problem building an old fortran library. Taking as an example:
> >
> > > cat check.f
> > program check
> > write(6,1000)
> > 1000 format('check')
> > end
>
> Did you read the documentation for -malign-double?
Yes, but I wasn't the person who originally wrote the build system for the librray in question. I looked it up once I found it was casuing the error and I cannot really understand why it was using it. The case I posted was just a trivial reduction that showed the effect I was getting.
> It specificly says -malign-double will change the ABI.
I certainly don't normally use that flag purely because of that warning.
> Also why are you using i686-pc-linux-gnuaout? a.out support for linux
> has not been supported for a while now, I don't even think glibc uses
> it anymore.
This surprised me too. I ran config.guess from my automake RPM package and that was the result on my Fedora Core #6 installation. As far as I'm aware it is a standard installation - I certainly made no special modifications to the basic system (other than build the svn gfortran anyway!).
Steve
--
----------------------------------------------------------
----------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20070531/5b07c630/attachment.sig>
More information about the Fortran
mailing list