NLO erros

Hi,

NLO process requires that your compiler support quadruple precision.
On most machine having gcc 4.6 or later should do the trick.
But looks like some distributions shipped such version of gcc but without the libquadmath library which is required in this case.
(On which OS are you running by the way?)

So far, I do not have enough information on this problem to tell you which apt-get/yum package to install in order to install such library.
Install a new version of gcc from source might do the trick but this is probably an overkill.

Cheers,

Olivier

> On 22 Jun 2017, at 18:17, roberto <email address hidden> wrote:
>
> New question #644533 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/644533
>
> Hi,
>
> I am trying to run a process in NLO, but I am getting the error:
>
> WARNING: fct <function compile_dir at 0xb690c3ac> does not return 0. Stopping the code in a clean way. The error was:
> A compilation Error occurs when trying to compile /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/DY_LQR2_NLO_6_8TeV_test103/SubProcesses/P0_uux_epem.
> The compilation fails with the following output message:
> gfortran -O -fno-automatic -ffixed-line-length-132 -c -I. check_poles.f
> check_poles.f:130.15:
>
> 1 pmass_rambo, prambo)
> 1
> Warning: Actual argument contains too few elements for dummy argument 'xm' (5/100) at (1)
> check_poles.f:160.19:
>
> 1 pmass_rambo, prambo)
> 1
> Warning: Actual argument contains too few elements for dummy argument 'xm' (5/100) at (1)
> run.inc:79.21:
> Included at check_poles.f:48:
>
> common/to_rwgt/ do_rwgt_scale, rw_Fscale_down, rw_Fscale_up, rw_Rscale_do
> 1
> Warning: Padding of 4 bytes required before 'rw_fscale_down' in COMMON 'to_rwgt' at (1); reorder elements or use -fno-align-commons
> gfortran -O -fno-automatic -ffixed-line-length-132 -c -I. BinothLHA.f
> gfortran -O -fno-automatic -ffixed-line-length-132 -c -I. born_hel.f
> for dir in `ls -d V*`; do cd $dir; make; cd ../; done
> make[1]: Entering directory `/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/DY_LQR2_NLO_6_8TeV_test103/SubProcesses/P0_uux_epem/V0_uux_epem'
> rm -f ../libMadLoop.a
> gfortran -fno-automatic -ffixed-line-length-132 -O3 -c polynomial.f -o polynomial.o
> gfortran -O -fno-automatic -ffixed-line-length-132 -c loop_matrix.f -I /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
> gfortran -O -fno-automatic -ffixed-line-length-132 -c improve_ps.f -I /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
> gfortran -O -fno-automatic -ffixed-line-length-132 -c born_matrix.f -I /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
> gfortran -O -fno-automatic -ffixed-line-length-132 -c loop_num.f -I /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
> gfortran -O -fno-automatic -ffixed-line-length-132 -c CT_interface.f -I /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
> CT_interface.f:667.29:
>
> ABSP_TMP(I)=0.E0+0_16
> 1
> Error: Integer kind 16 at (1) not available
> CT_interface.f:669.30:
>
> P_TMP(I,J)=0.E0+0_16
> 1
> Error: Integer kind 16 at (1) not available
> make[1]: *** [CT_interface.o] Error 1
> make[1]: Leaving directory `/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/DY_LQR2_NLO_6_8TeV_test103/SubProcesses/P0_uux_epem/V0_uux_epem'
> gfortran -o check_poles parton_lum_1.o parton_lum_2.o parton_lum_3.o parton_lum_chooser.o matrix_1.o matrix_2.o matrix_3.o real_me_chooser.o fks_inc_chooser.o leshouche_inc_chooser.o genps_fks.o setcuts.o setscales.o born.o sborn_sf.o b_sf_001.o fks_Sij.o trapfpe.o fastjetfortran_madfks_core.o fjcore.o fastjet_wrapper.o fks_singular.o veto_xsec.o montecarlocounter.o reweight_xsec.o boostwdir2.o configs_and_props_inc_chooser.o initcluster.o cluster.o reweight.o get_color.o FKSParamReader.o iproc_map.o MC_integer.o reweight_xsec_events_pdf_dummy.o appl_interface_dummy.o cuts.o pythia_unlops.o recluster.o check_poles.o BinothLHA.o born_hel.o -L../../lib/ libMadLoop.a -lcts -liregi -L/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/lib/ -lninja -L/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/lib/ -lavh_olo -L../../lib/ -ldhelas -lgeneric -lmodel -lpdf -lcernlib -lstdc++ -lstdc++
> gfortran: error: libMadLoop.a: No such file or directory
> make: *** [check_poles] Error 1
>
> Please try to fix this compilations issue and retry.
> Help might be found at https://answers.launchpad.net/mg5amcnlo.
> If you think that this is a bug, you can report this at https://bugs.launchpad.net/mg5amcnlo
>
> What can I do?
>
> Cheers,
>
> Roberto
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier I have verified and my gcc version is 4.8.4
My OS is Ubuntu 14.04.03
What others informations do you need?

Cheers,

Roberto

--------------------------------------------
Em sex, 23/6/17, Olivier Mattelaer <email address hidden> escreveu:

 Assunto: Re: [Question #644533]: NLO erros
 Para: <email address hidden>
 Data: Sexta-feira, 23 de Junho de 2017, 10:17

 Your question #644533 on MadGraph5_aMC@NLO
 changed:
 https://answers.launchpad.net/mg5amcnlo/+question/644533

     Status: Open =>
 Answered

 Olivier Mattelaer proposed the
 following answer:
 Hi,

 NLO process requires that your compiler
 support quadruple precision.
 On most machine having gcc 4.6 or later
 should do the trick.
 But looks like some distributions
 shipped such version of gcc but without the libquadmath
 library which is required in this case.
 (On which OS are you running by the
 way?)

 So far, I do not have enough
 information on this problem to tell you which apt-get/yum
 package to install in order to install such library.
 Install a new version of gcc from
 source might do the trick but this is probably an overkill.

 Cheers,

 Olivier

 > On 22 Jun 2017, at 18:17, roberto
 <email address hidden>
 wrote:
 >
 > New question #644533 on MadGraph5_aMC@NLO:
 > https://answers.launchpad.net/mg5amcnlo/+question/644533
 >
 > Hi,
 >
 > I am trying to run a process in
 NLO, but I am getting the error:
 >
 > WARNING: fct <function
 compile_dir at 0xb690c3ac> does not return 0. Stopping
 the code in a clean way. The error was:
 > A compilation Error occurs when
 trying to compile
 /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/DY_LQR2_NLO_6_8TeV_test103/SubProcesses/P0_uux_epem.
 > The compilation fails with the
 following output message:
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c -I. check_poles.f
 >
 check_poles.f:130.15:
 >
 >
 1        pmass_rambo, prambo)

 >
          1
 >    Warning: Actual
 argument contains too few elements for dummy argument 'xm'
 (5/100) at (1)
 >
 check_poles.f:160.19:
 >
 >
 1            pmass_rambo,
 prambo)

 >
              1
 >    Warning: Actual
 argument contains too few elements for dummy argument 'xm'
 (5/100) at (1)
 >    run.inc:79.21:
 >
 Included at check_poles.f:48:
 >
 >
 common/to_rwgt/ do_rwgt_scale, rw_Fscale_down, rw_Fscale_up,
 rw_Rscale_do
 >
                1
 >    Warning: Padding of 4
 bytes required before 'rw_fscale_down' in COMMON 'to_rwgt'
 at (1); reorder elements or use -fno-align-commons
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c -I. BinothLHA.f
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c -I. born_hel.f
 >    for dir in `ls -d
 V*`; do cd $dir; make; cd ../; done
 >    make[1]: Entering
 directory
 `/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/DY_LQR2_NLO_6_8TeV_test103/SubProcesses/P0_uux_epem/V0_uux_epem'
 >    rm -f ../libMadLoop.a

 >    gfortran
 -fno-automatic -ffixed-line-length-132 -O3  -c
 polynomial.f -o polynomial.o
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c loop_matrix.f -I
 /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c improve_ps.f -I
 /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c born_matrix.f -I
 /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c loop_num.f -I
 /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
 >    gfortran -O
 -fno-automatic -ffixed-line-length-132 -c CT_interface.f -I
 /home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/include
 >
 CT_interface.f:667.29:
 >
 >
   ABSP_TMP(I)=0.E0+0_16

 >

        1
 >    Error: Integer kind
 16 at (1) not available
 >
 CT_interface.f:669.30:
 >
 >
     P_TMP(I,J)=0.E0+0_16

 >

         1
 >    Error: Integer kind
 16 at (1) not available
 >    make[1]: ***
 [CT_interface.o] Error 1
 >    make[1]: Leaving
 directory
 `/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/DY_LQR2_NLO_6_8TeV_test103/SubProcesses/P0_uux_epem/V0_uux_epem'
 >    gfortran -o
 check_poles parton_lum_1.o parton_lum_2.o parton_lum_3.o
 parton_lum_chooser.o matrix_1.o matrix_2.o matrix_3.o
 real_me_chooser.o fks_inc_chooser.o leshouche_inc_chooser.o
 genps_fks.o setcuts.o setscales.o born.o sborn_sf.o
 b_sf_001.o fks_Sij.o trapfpe.o fastjetfortran_madfks_core.o
 fjcore.o fastjet_wrapper.o fks_singular.o veto_xsec.o
 montecarlocounter.o reweight_xsec.o boostwdir2.o
 configs_and_props_inc_chooser.o initcluster.o cluster.o
 reweight.o get_color.o FKSParamReader.o iproc_map.o
 MC_integer.o reweight_xsec_events_pdf_dummy.o
 appl_interface_dummy.o cuts.o pythia_unlops.o recluster.o
 check_poles.o BinothLHA.o born_hel.o -L../../lib/
 libMadLoop.a -lcts -liregi
 -L/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/lib/
 -lninja
 -L/home/roger/work_ROBERTO/MG5_aMC_v2_5_5/HEPTools/lib/
 -lavh_olo  -L../../lib/ -ldhelas -lgeneric -lmodel
 -lpdf -lcernlib  -lstdc++ -lstdc++
 >    gfortran: error:
 libMadLoop.a: No such file or directory
 >    make: ***
 [check_poles] Error 1
 >
 > Please try to fix this
 compilations issue and retry.
 > Help might be found at
 https://answers.launchpad.net/mg5amcnlo.
 > If you think that this is a bug,
 you can report this at https://bugs.launchpad.net/mg5amcnlo
 >
 > What can I do?
 >
 > Cheers,
 >
 > Roberto
 >
 > --
 > You received this question
 notification because you are an answer
 > contact for MadGraph5_aMC@NLO.

 --
 If this answers your question, please
 go to the following page to let us
 know that it is solved:
 https://answers.launchpad.net/mg5amcnlo/+question/644533/+confirm?answer_id=0

 If you still need help, you can reply
 to this email or go to the
 following page to enter your feedback:
 https://answers.launchpad.net/mg5amcnlo/+question/644533

 You received this question notification
 because you asked the question.

Your issue here is not related to MG5aMC at all.
The compilation error you face is straight-forward to decipher:

ABSP_TMP(I)=0.E0+0_16
                                 1
    Error: Integer kind 16 at (1) not available

meaning that you must have a compiler suite that supports quadruple precision. Normally all gfortran compilers of version 4.6 or above do support it, but for some reason yours satisfy this criterion but still doesn't ship with libquadmath.

In any case, simply create a little fortran code like this one (in a file called 'testCompiler.f' for instance):

      program test

          real*16 MyDoublePrecisionVariable

          MyDoublePrecisionVariable = 2.0E+00_16
          MyDoublePrecisionVariable = SQRT(MyDoublePrecisionVariable)
          MyDoublePrecisionVariable = MyDoublePrecisionVariable**2
          write(*,*) MyDoublePrecisionVariable

      end program test

And then try compiling and running it with:

gfortran -o test.exe testCompiler.f90
./test.exe

On your system, this compilation will probably fail, so use all sorts of internet resources and/or upgrade your compiler suite so as to fix this issue.
Only come back to MG5_aMC once the above compiles, since as Olivier correctly stressed, this is a requirement for running NLO processes.

Hi,

then according to this page:
https://packages.ubuntu.com/fr/trusty/gcc-4.8
libquadmath is only suggested and not automatic on that system

The following page should contain what you need:
https://packages.ubuntu.com/fr/trusty/libquadmath0-dbg
(but It might be possible to install those via apt-get (probably apt-get install libquadmath0-dbg)

Cheers,

Olivier