Various issues with HRDCOD=0 and HRDCOD=1 builds of no_b_mass and loop_no_b_mass SM processes

[valassi]

Add support for HRDCOD=0 builds of no_b_mass SM processes

Presently these processes fail the build like non-SM processes. The problem, as usual, is the handling of parameter dependencies.

For instance in ttW (which I just added as suggested by Francesco Giuli in ATLAS):

[avalassi@itscrd80 gcc11.2/cvmfs] /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/pp_ttW.sa/SubProcesses/P0_Sigma_loop_sm_no_b_mass_dux_ttxwm> make 
...
Building in BUILDDIR=. for tag=512y_d_inl0_hrd0_hasCurand (USEBUILDDIR is not set)
make -C ../../src  -f cudacpp_src.mk
make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/pp_ttW.sa/src'
AVX=512y
ccache /cvmfs/sft.cern.ch/lcg/releases/gcc/11.2.0-ad950/x86_64-centos8/bin/g++  -O3  -std=c++17 -I.  -fPIC -Wall -Wshadow -Wextra -ffast-math  -fopenmp -march=skylake-avx512 -mprefer-vector-width=256  -DMGONGPU_FPTYPE_DOUBLE -DMGONGPU_FPTYPE2_DOUBLE -c Parameters_loop_sm_no_b_mass.cc -o Parameters_loop_sm_no_b_mass.o
In file included from Parameters_loop_sm_no_b_mass.cc:15:
Parameters_loop_sm_no_b_mass.h:26:2: error: #error This non-SM physics process only supports MGONGPU_HARDCODE_PARAM builds (#439): please run "make HRDCOD=1"
   26 | #error This non-SM physics process only supports MGONGPU_HARDCODE_PARAM builds (#439): please run "make HRDCOD=1"
      |  ^~~~~

[valassi]

The process is

    pp_ttW)
      cmd="import model loop_sm-no_b_mass
      define p = p b b~
      define j = p
      define w = w+ w- # W case only
      generate p p > t t~ w @0
      add process p p > t t~ w j @1"
      ;;

[valassi]

Note, this affects ALL no_b_mass SM processes - including other ATLAS processes like #673

[valassi]

I have renamed this to mention that BOTH no_b_mass and loop_no_b_mass processes are affected for HRDCOD=0.

Conversely, for HRDCOD=1, only loop_no_b_mass processes cannot be built (#696), while no_b_mass can be built (MR #706), although they currently fail at runtime (#733).

[valassi]

I have renamed this to mention that BOTH no_b_mass and loop_no_b_mass processes are affected for HRDCOD=0.

Conversely, for HRDCOD=1, only loop_no_b_mass processes cannot be built (#696), while no_b_mass can be built (MR #706), although they currently fail at runtime (#733).

The first statement is now wrong. I am not sure if something changed, or most likely I was confused when I wrote it.

I confirm that nobm_pp_ttW.mad builds ok also for HRDCOD=0 (and, not surprisingly, for HRDCOD=1 too). See the upcoming MR #706. So there is no problem with nobm models, there is only a problem with loop_nobm models.

I will rename this issue, again.

[valassi]

I think that this issue was mixing too many things, HRDCOD=0 and HRDCOD=1, and no_b_mass and loop_no_b_mass.

The no_b_mass builds of pp_ttW are fixed after a long debugging (FPEs etc), for both HRDCOD=0 and HRDCOD=1.

There are a few issues remaining specifically for loop_no_b_mass instead (COND and reglog), but these are better followed up in #696.

I close this one for simplicity and I rename it.