[Nix-dev] Re: gcc broken for FORTRAN build

[Pjotr Prins]

On Fri, Aug 08, 2008 at 10:51:12AM +0200, Peter Simons wrote:
> Hi Pjotr,
> 
>  > The time.patch works fine! This was a response to Eelco.
>  > Without the patch coreutils does not compile.
> 
> Eelco was wondering what functionality is lost if the patch is applied.
> If we apply that patch, clearly it will fix the compile errors in
> coreutils some people have experienced. That's good. But what is the
> downside? What do we lose if we apply that patch?

I haven't looked closely. The issue is here:

http://bugs.strategoxt.org/browse/NIXPKGS-94

The patch describes:

+ /* POSIX 200x added two interfaces to set file timestamps with
+- nanosecond resolution. */
++ nanosecond resolution. We provide a fallback for ENOSYS (for
++ example, compiling against Linux 2.6.25 kernel headers and glibc
++ 2.7, but running on Linux 2.6.18 kernel). */ 

Pj.