Re: AW: MPI recognition

2010-12-22 Thread Olaf Lenz
On 12/22/2010 06:40 PM, Carsten Heinrici wrote: > If a project requires a specialized tool chain, I would consider it wrong to > just silently fall back to something different as required. Well, the point is that our project doesn't require MPI, but it profits from it. If MPI is not available, y

AW: MPI recognition

2010-12-22 Thread Carsten Heinrici
Your request sounds strange to me. If a project requires a specialized tool chain, I would consider it wrong to just silently fall back to something different as required. It is similar to the always re appearing discussion about 64bit libs. It is in the responsibility of the integrator to c

Re: MPI recognition

2010-12-22 Thread Olaf Lenz
Hi Ralf, Erik! On 12/22/2010 07:44 AM, Ralf Wildenhues wrote: > One problem with the above is that the settings 'gcc' and 'cc' will fail > to be correct MPI compilers, as they likely won't link in the MPI > library. Nowadays this doesn't happen so often any more but on some > systems it is good t

Re: MPI recognition

2010-12-21 Thread Ralf Wildenhues
Hi Olaf, and sorry for the delay in replying, Eric, * Eric Blake wrote on Tue, Dec 21, 2010 at 04:57:04PM CET: > On 12/21/2010 07:16 AM, Olaf Lenz wrote: > > > After some trial and error I have found a solution for my problem that > > is very simple. Now I wonder whether this is actually the corr

Re: MPI recognition

2010-12-21 Thread Eric Blake
On 12/21/2010 07:16 AM, Olaf Lenz wrote: > I have attached my simplified attempt do this. When I run autoconf, I > get the warning > > configure.ac:4: warning: AC_REQUIRE: `AC_PROG_CC' was expanded before > it was required > > and configure doesn't seem to be looking for mpicc, but always finds

Re: MPI recognition

2010-12-21 Thread Olaf Lenz
Hello again! On 12/21/2010 03:16 PM, Olaf Lenz wrote: > The problem seems to be that AC_PROG_CC doesn't work when using it in > a conditional, is that right? Is there any simple way to do what I > want? After some trial and error I have found a solution for my problem that is very simple. Now I w

MPI recognition

2010-12-21 Thread Olaf Lenz
Hello everybody! For our scientific software ESPResSo[1], I am trying to get configure to be able to handle MPI compilers correctly. MPI is a parallelization library and it often brings its own compiler, i.e. actually it only brings a wrapper to your compiler, which is usually called mpicc. In our