[cmsdy] in dsample.f of pp_dy3j.mad P0_gux_taptamggux (also copied to…

… gg_tt.mad), comment out dead if/then branches (for warnings that are commented out)

This is another minor component of madgraph5#969. It gives almost insignificant performance improvements, but it simplifies the code.

CUDACPP_RUNTIME_DISABLEFPE=1 ./build.cuda_d_inl0_hrd0/madevent_cuda < /tmp/avalassi/input_dy3j_x1_cudacpp
 [COUNTERS] PROGRAM TOTAL                         :    4.1574s
 [COUNTERS] Fortran Other                  (  0 ) :    0.1706s
 [COUNTERS] Fortran Initialise(I/O)        (  1 ) :    0.0670s
 [COUNTERS] Fortran Random2Momenta         (  3 ) :    2.8950s for  1170103 events => throughput is 2.47E-06 events/s
 [COUNTERS] Fortran PDFs                   (  4 ) :    0.1021s for    49152 events => throughput is 2.08E-06 events/s
 [COUNTERS] Fortran UpdateScaleCouplings   (  5 ) :    0.1360s for    16384 events => throughput is 8.30E-06 events/s
 [COUNTERS] Fortran Reweight               (  6 ) :    0.0518s for    16384 events => throughput is 3.16E-06 events/s
 [COUNTERS] Fortran Unweight(LHE-I/O)      (  7 ) :    0.0679s for    16384 events => throughput is 4.15E-06 events/s
 [COUNTERS] Fortran SamplePutPoint         (  8 ) :    0.1401s for  1170103 events => throughput is 1.20E-07 events/s
 [COUNTERS] CudaCpp Initialise             ( 11 ) :    0.4658s
 [COUNTERS] CudaCpp Finalise               ( 12 ) :    0.0263s
 [COUNTERS] CudaCpp MEs                    ( 19 ) :    0.0347s for    16384 events => throughput is 2.12E-06 events/s
 [COUNTERS] OVERALL NON-MEs                ( 21 ) :    4.1227s
 [COUNTERS] OVERALL MEs                    ( 22 ) :    0.0347s for    16384 events => throughput is 2.12E-06 events/s

epochX/cudacpp/gg_tt.mad/Source/dsample.f

@@ -1349,10 +1349,10 @@ subroutine sample_get_x(wgt, x, j, ipole, xmin, xmax)
             call ntuple(ddum(j),max(xbin_min,dble(int(tx(2,j)))),
      $           min(xbin_max,dble(int(tx(2,j))+1)),j,ipole)
 
-            if(max(xbin_min,dble(int(tx(2,j)))).gt.
-     $           min(xbin_max,dble(int(tx(2,j))+1))) then
+c           if(max(xbin_min,dble(int(tx(2,j)))).gt.
+c    $           min(xbin_max,dble(int(tx(2,j))+1))) then
 c               write(*,*) 'not good'
-            endif
+c           endif
 
 c            write(*,'(2i6,4e15.5)') nzoom,j,ddum(j),tx(2,j),
 c     $           max(xbin_min,dble(int(tx(2,j)))),
@@ -1426,10 +1426,10 @@ subroutine sample_get_x(wgt, x, j, ipole, xmin, xmax)
          endif
          icount=icount+1
       endif
-      if (x .lt. xmin .or. x .gt. xmax) then
+c     if (x .lt. xmin .or. x .gt. xmax) then
 c         write(*,'(a,4i4,2f24.16,1e10.2)') 'Bad x',ij,int(xbin_min),ip,
 c     &        int(xbin_max),xmin,x,xmax-xmin
-      endif
+c     endif
 
       wgt = wgt * xo * dble(xbin_max-xbin_min)
 c      print*,'Returning x',ij,ipole,j,x

epochX/cudacpp/pp_dy3j.mad/Source/dsample.f

@@ -1349,10 +1349,10 @@ subroutine sample_get_x(wgt, x, j, ipole, xmin, xmax)
             call ntuple(ddum(j),max(xbin_min,dble(int(tx(2,j)))),
      $           min(xbin_max,dble(int(tx(2,j))+1)),j,ipole)
 
-            if(max(xbin_min,dble(int(tx(2,j)))).gt.
-     $           min(xbin_max,dble(int(tx(2,j))+1))) then
+c           if(max(xbin_min,dble(int(tx(2,j)))).gt.
+c    $           min(xbin_max,dble(int(tx(2,j))+1))) then
 c               write(*,*) 'not good'
-            endif
+c           endif
 
 c            write(*,'(2i6,4e15.5)') nzoom,j,ddum(j),tx(2,j),
 c     $           max(xbin_min,dble(int(tx(2,j)))),
@@ -1426,10 +1426,10 @@ subroutine sample_get_x(wgt, x, j, ipole, xmin, xmax)
          endif
          icount=icount+1
       endif
-      if (x .lt. xmin .or. x .gt. xmax) then
+c     if (x .lt. xmin .or. x .gt. xmax) then
 c         write(*,'(a,4i4,2f24.16,1e10.2)') 'Bad x',ij,int(xbin_min),ip,
 c     &        int(xbin_max),xmin,x,xmax-xmin
-      endif
+c     endif
 
       wgt = wgt * xo * dble(xbin_max-xbin_min)
 c      print*,'Returning x',ij,ipole,j,x