]] Hendrik Sattler
(Please don't Cc me, I read the list, as noted in my M-F-T header)
| Am Donnerstag 29 Juli 2010, 19:23:40 schrieb Tollef Fog Heen:
| > The reason for stripping -I is so you can
| > have in-tree include files that get included correctly even if their
| > names overlap the files
Hendrik Sattler writes ("Re: Problem with gfortran and pkg-config"):
> Am Donnerstag 29 Juli 2010, 19:23:40 schrieb Tollef Fog Heen:
> > The reason for stripping -I is so you can
> > have in-tree include files that get included correctly even if their
> > names ov
Am Donnerstag 29 Juli 2010, 19:23:40 schrieb Tollef Fog Heen:
> The reason for stripping -I is so you can
> have in-tree include files that get included correctly even if their
> names overlap the files in /usr/include.
This is not the case with current gcc. From "man gcc":
"Directories named by -
]] Alastair McKinstry
| Which is wrong, gfortran not reading /usr/include by default, or
| pkg-config for removing /usr/include from CFLAGS ?
Either stop shipping non-C header files in /usr/include (aka, move the
.mod file) or fix gfortran to look in /usr/include by default. If
there's a defaul
On Thu, Jul 29, 2010 at 12:52:32 +0100, Alastair McKinstry wrote:
> Hi,
>
> I've a problem using pkg-config for Fortran programs.
> As shipped in sid (and squeeze), libnetcdf-dev includes a netcdf.pc
> and /usr/include/netcdf.mod
>
> Now, to compile a fortran90 program I need to include
> gfort
Alastair McKinstry wrote:
Hi,
> Which is wrong, gfortran not reading /usr/include by default, or
> pkg-config for
> removing /usr/include from CFLAGS ?
I'll add to the above:
- netcdf-dev for installing netcdf.mod in the wrong place?
Is there a standard place for .mod files?
JB.
--
Julien
6 matches
Mail list logo
