surprised that
there is no existing AC_PROG_MAIL macro.
I thought I would add a --disable-unidata-email parameter to disable the
email attempt. (Which I would use myself with nightly testing.)
Is there some good reason not to do this? But I have never seen it
done...
Thanks!
Ed
--
Ed
to tell AC_SEARCH_LIBS to add another directory to search for the
library?
Thanks!
Ed
--
Ed Hartnett -- e...@unidata.ucar.edu
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
| end
In fact, the smallest test program I can get gfortran to compile is:
program main
use netcdf
res = nf90_open('lll', 0, ncid)
end program main
Does anyone use the AC_SEARCH_LIBS macro with fortran?
Thanks!
Ed
--
Ed Hartnett -- e...@unidata.ucar.edu
this, but what is it?
Thanks!
Ed
--
Ed Hartnett -- e...@unidata.ucar.edu
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
. There is a need for a preferred method
for doing this and
documenting how you do it. Cheers - Bruce
I use sed at configure time to muck with my package's main .h file.
(But I don't install a config.h. That sounds a little scary.)
Ed
--
Ed Hartnett -- [EMAIL PROTECTED
suggestions how? I've seen people write scripts to modify/rename
config.h but that seems overkill in this case.
Thanks much.
-DB
How about something like:
#ifndef MY_CONFIG_INCLUDED
#include ../my/config.h
#define MY_CONFIG_INCLUDED
#endif
--
Ed Hartnett -- [EMAIL PROTECTED
.) To get the answer in config.h, use
AC_DEFINE to define something.
Good luck!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
, and as this is something I'm about
to try myself, I might have a different viewpoint in six months...
Thanks,
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
Wildenhues for the
patches, and all the autotools help in general!)
Thanks Ralf!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
Ralf Wildenhues [EMAIL PROTECTED] writes:
* Ed Hartnett wrote on Wed, Aug 09, 2006 at 07:22:40PM CEST:
Ralf Wildenhues [EMAIL PROTECTED] writes:
* John Wohlbier wrote on Tue, Jul 25, 2006 at 04:12:57PM CEST:
http://cvs.savannah.gnu.org/viewcvs/~checkout~/config/config/config.guess
http
of shared libraries, they must run make clean first
too, right?)
Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
Keith MARSHALL [EMAIL PROTECTED] writes:
Ed Hartnett wrote, quoting me:
However, IMO that's a lazy hack -- it suggests that you aren't
really thinking about *why* you specified those attributes in the
first place, and what the implications are, when you port your code
to a different OS
should not be cross-compiling
anyway, I should just be building under mingw. Which I now do.
Thanks for all the help!!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
#endif
#ifndef S_IWOTH
#define S_IWOTH 002
#endif
Thanks,
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
lines of code in my C
files.
Thanks,
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
compiler... yes
What is with all the i686-pc-mingw32- stuff?
Any help appreciated. Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
understand why doing this on the configure command
line is better than setting environment variables.
And what about the case where you set the value before ./configure?
CPPFLAGS=-Dsomething ./configure
Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED
two different build methods for a minor
gain in build speed in one of them. Buy a faster workstation instead,
it's a lot cheaper than extra configuration maintenance.
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
platforms he uses as development
environment, but that's it:
* Compilers might not support -q64.
* -q64 might have different meanings for different compilers.
* -q64 could have different meanings for different architectures.
This will cause errors when building his package.
Yep.
Ed
--
Ed
me
and ask me how to set it, so this saves us all the trouble.)
Good luck!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
. This works if I supply TEMP_LARGE on the command line with the
make command, but it doesn't pick up the TEMP_LARGE defined during
configure.
So how do I take TEMP_LARGE and get it into one of my test programs?
Any suggestions welcome.
Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED
was a significant problem.
Now I am much happier, my build system is far more robust, and adding
a new system is no work or very little work. The autotools way is hard
at first but scales very well. The hand-written makefile is even
harder to write, and scales poorly.
Ed
--
Ed Hartnett -- [EMAIL PROTECTED
and the user emails me.
Yes, in theory the user should read the documentation and pick the
right parameters themselves.
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
macro).
However, how to handle cases where one of these is already installed
on the end-user system?
Anyone autotool-using library builder out there who has tried this?
Any comments on this sort of problem, and solutions for it, would be
very helpful.
Thanks!
Ed
--
Ed Hartnett -- [EMAIL
Howdy all!
I have a configure script which needs to know whether it is dealing
with the f95 distributed with gcc-4.x, or the g95, which is apparently
an independent project, which handles slightly differently.
In both cases ac_cv_fc_compiler_gnu is set to yes.
At the moment I am distinguishing
, including telling me somehow whether it can handle F90
code.
Is there a better way? Or is it required that I have AC_PROG_F77 and
AC_PROG_FC?
Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org
.)
I just use autoreconf -i, which seems to do everything needed without me
having to remember which order to call various tools, or to maintain a
bootstrap script...
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
with some frequency, so many
developers will have to keep up with autotools versions anyway.
Since that's already the case for us, we are going to stop putting
configure and Makefile(s) in the repository.
Good luck!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED
makefiles and
automakefiles, I am glad to leave makefiles behind forever. They are
not that fun!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
it appear in the CFLAGS, FFLAGS and
CPPFLAGS as -L and -I options.
Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
that
the user has specified. How do I get this to work without setting
CFLAGS in configure?
Thanks for any suggestions!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
pgf90 epcf90 g77])
So here I call first AC_PROG_F77, then AC_PROG_FC.
If I just call AC_PROG_FC, the later, when I call AC_PROG_LIBTOOL it
will call AC_PROG_F77, ignoring the fact that FC is set to the fortran
compiler.
Is there a better way to handle this?
Thanks!
Ed
--
Ed Hartnett
],
[AS_HELP_STRING([--disable-f77],
[don't try to build Fortran 77 API])])
test x$enable_f77 = xno nc_build_f77=no
test x$nc_null_fc = xyes nc_build_f77=no
AC_MSG_RESULT([$nc_build_f77])
--
Ed Hartnett -- [EMAIL PROTECTED
sometimes accepted
headers that clashed at compile-time. If you need to check whether a
header is preprocessable, you can use `AC_PREPROC_IFELSE' (*note
Running the Preprocessor::).
Thanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf
to config\
ure.])
fi
fi
I have to append -I${with_hdf5}/include to the CPPFLAGS. Is this the
usual way to do such a thing?
Or is there a better way?
THanks!
Ed
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http
Bob Friesenhahn [EMAIL PROTECTED] writes:
On Fri, 3 Jun 2005, Ed Hartnett wrote:
I have to append -I${with_hdf5}/include to the CPPFLAGS. Is this the
usual way to do such a thing?
Or is there a better way?
Another way is to define a new variable HDF5INCLUDES and export for
substitution
configure wants to set a flag, but the user wants to try
a contradictory flag. So I allow a way to stop the configure from
messing with the flags, so that the user can take complete control if
need be.
Ed
--
Ed Hartnett -- [EMAIL PROTECTED
AM_CONDITIONAL(BUILD_CXX, [test x$nc_build_cxx = xyes])
--
Ed Hartnett -- [EMAIL PROTECTED]
___
Autoconf mailing list
Autoconf@gnu.org
http://lists.gnu.org/mailman/listinfo/autoconf
Howdy all!
We're using autoconf, which I love, to help install netcdf, a freeware
scientific data library for Earth Science.
Where does autoconf look for compilers? For example, I am on a mac,
and autoconf finds a C++ compiler called CC, but it is not in my path.
So how did autoconf find it?
[EMAIL PROTECTED] (Bob Proulx) writes:
Jeff Sheinberg wrote:
Ed Hartnett writes:
I wonder if there is any autoconf macro for running getconf to find
the compile flags needed for 64-bit compilation on platforms that
support it. For example, both the AIX and solaris platforms have
40 matches
Mail list logo