On Sun, Jun 03, 2001 at 11:03:23PM -0400, Harlan Stenn wrote:
> > Besides the reason that it's not nessasary are there any reasons why
> > the automake inference rules for C and C++ don't use -o ?
> >
> > I seem to remember that some compilers don't accept -o is this correct?
>
> Yes. Some older compilers will not accept both -c and -o .
>
> > If so what is the typical way of dealing with the lack of -o when
> > you want it?
>
> either mv the produced .o name to what you want, or find a way to compile
> the source from a faked .c file.
Is there any kind of concensus on what would be the 'correct' way to do
this in automake? Is there a consensus that automake should continue to
support these old compilers? I am attempting to modify automake to
support paths and want the output files to end up in the same directory
structure as the original files.
I am not an expert in autoconf or automake, so I ask those who are, would
a macro to detect whether the compiler supports -o with -c and some
additional code to handle the case when it dosn't be a better way to
handle this condition?
>
> You could also mv an existing .o out of the way, compile, and rename back,
> but that gets messy and doesn't do well with parallel makes.
>
> H
PGP signature
