[june24] regenerate all processes - only banner.py has changed

epochX/cudacpp/ee_mumu.mad/CODEGEN_mad_ee_mumu_log.txt

@@ -62,7 +62,7 @@ generate e+ e- > mu+ mu-
 No model currently active, so we import the Standard Model
 INFO: load particles 
 INFO: load vertices 
-DEBUG: model prefixing  takes 0.00567936897277832 
+DEBUG: model prefixing  takes 0.005466938018798828 
 INFO: Restrict model sm with file models/sm/restrict_default.dat . 
 DEBUG: Simplifying conditional expressions 
 DEBUG: remove interactions: u s w+ at order: QED=1 
@@ -154,7 +154,7 @@ INFO: Checking for minimal orders which gives processes.
 INFO: Please specify coupling orders to bypass this step. 
 INFO: Trying process: e+ e- > mu+ mu- WEIGHTED<=4 @1  
 INFO: Process has 2 diagrams 
-1 processes with 2 diagrams generated in 0.005 s
+1 processes with 2 diagrams generated in 0.004 s
 Total: 1 processes with 2 diagrams
 output madevent_simd ../TMPOUT/CODEGEN_mad_ee_mumu --hel_recycling=False --vector_size=32
 Load PLUGIN.CUDACPP_OUTPUT
@@ -164,10 +164,10 @@ It has been validated for the last time with version: 3.5.2
 Addition matrix-element will be done with PLUGIN: CUDACPP_OUTPUT
 Output will be done with PLUGIN: CUDACPP_OUTPUT
 DEBUG:  cformat =  standalone_simd [export_cpp.py at line 3070] 
-DEBUG:  Entering PLUGIN_ProcessExporter.__init__ (initialise the exporter) [output.py at line 162] 
+DEBUG:  Entering PLUGIN_ProcessExporter.__init__ (initialise the exporter) [output.py at line 165] 
 INFO: initialize a new directory: CODEGEN_mad_ee_mumu 
 INFO: remove old information in CODEGEN_mad_ee_mumu 
-DEBUG:  Entering PLUGIN_ProcessExporter.copy_template (initialise the directory) [output.py at line 167] 
+DEBUG:  Entering PLUGIN_ProcessExporter.copy_template (initialise the directory) [output.py at line 170] 
 WARNING: File exists /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu 
 INFO: Creating subdirectories in directory /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu 
 WARNING: File exists /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu/Cards 
@@ -185,24 +185,20 @@ INFO: Created files CPPProcess.h and CPPProcess.cc in directory ./.
 DEBUG:  config_map =  [1, 2] [export_cpp.py at line 711] 
 DEBUG:  subproc_number =  0 [export_cpp.py at line 712] 
 DEBUG:  Done [export_cpp.py at line 713] 
-DEBUG:  vector, subproc_group,self.opt['vector_size'] =  False True 32 [export_v4.py at line 1877] 
-DEBUG:  vector, subproc_group,self.opt['vector_size'] =  False True 32 [export_v4.py at line 1877] 
-DEBUG:  vector, subproc_group,self.opt['vector_size'] =  32 True 32 [export_v4.py at line 1877] 
 WARNING: vector code for lepton pdf not implemented. We removed the option to run dressed lepton 
-DEBUG:  vector, subproc_group,self.opt['vector_size'] =  32 True 32 [export_v4.py at line 1877] 
 INFO: Generating Feynman diagrams for Process: e+ e- > mu+ mu- WEIGHTED<=4 @1 
 INFO: Finding symmetric diagrams for subprocess group epem_mupmum 
-DEBUG:  os.getcwd() =  /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu/SubProcesses/P1_epem_mupmum [export_v4.py at line 6444] 
-DEBUG:  len(subproc_diagrams_for_config) =  2 [model_handling.py at line 1520] 
-DEBUG:  iconfig_to_diag =  {1: 1, 2: 2} [model_handling.py at line 1544] 
-DEBUG:  diag_to_iconfig =  {1: 1, 2: 2} [model_handling.py at line 1545] 
+DEBUG:  os.getcwd() =  /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu/SubProcesses/P1_epem_mupmum [export_v4.py at line 6494] 
+DEBUG:  len(subproc_diagrams_for_config) =  2 [model_handling.py at line 1523] 
+DEBUG:  iconfig_to_diag =  {1: 1, 2: 2} [model_handling.py at line 1547] 
+DEBUG:  diag_to_iconfig =  {1: 1, 2: 2} [model_handling.py at line 1548] 
 Generated helas calls for 1 subprocesses (2 diagrams) in 0.004 s
-Wrote files for 8 helas calls in 0.107 s
+Wrote files for 8 helas calls in 0.081 s
 ALOHA: aloha starts to compute helicity amplitudes
 ALOHA: aloha creates FFV1 routines
 ALOHA: aloha creates FFV2 routines
 ALOHA: aloha creates FFV4 routines
-ALOHA: aloha creates 3 routines in  0.202 s
+ALOHA: aloha creates 3 routines in  0.203 s
 ALOHA: aloha starts to compute helicity amplitudes
 ALOHA: aloha creates FFV1 routines
 ALOHA: aloha creates FFV2 routines
@@ -236,22 +232,20 @@ DEBUG: cd /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu
 patching file Source/genps.inc
 patching file SubProcesses/makefile
 DEBUG: cd /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu/SubProcesses/P1_epem_mupmum; patch -p6 -i /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/CODEGEN/PLUGIN/CUDACPP_SA_OUTPUT/MG5aMC_patches/PROD/patch.P1
-patching file auto_dsig1.f
-Hunk #1 succeeded at 496 (offset 12 lines).
 patching file driver.f
 patching file matrix1.f
-Hunk #2 succeeded at 229 (offset 9 lines).
-DEBUG:  p.returncode =  0 [output.py at line 255] 
+Hunk #2 succeeded at 242 (offset 9 lines).
+DEBUG:  p.returncode =  0 [output.py at line 258] 
 Output to directory /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu done.
 Type "launch" to generate events from this process, or see
 /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_mad_ee_mumu/README
 Run "open index.html" to see more information about this process.
 quit
 
-real	0m2.783s
-user	0m1.820s
-sys	0m0.261s
-Code generation completed in 3 seconds
+real	0m2.059s
+user	0m1.802s
+sys	0m0.254s
+Code generation completed in 2 seconds
 ************************************************************
 *                                                          *
 *                      W E L C O M E to                    *

epochX/cudacpp/ee_mumu.mad/bin/internal/banner.py

@@ -3157,6 +3157,7 @@ def get_value_from_include(self, path, list_of_params, output_dir):
         if path does not exists return the current value in self for all parameter"""
 
         #WARNING DOES NOT HANDLE LIST/DICT so far
+        misc.sprint(output_dir, path)
         # handle case where file is missing
         if not os.path.exists(pjoin(output_dir,path)):
             misc.sprint("include file not existing", pjoin(output_dir,path))
@@ -3478,8 +3479,10 @@ def write_include_file(self, output_dir, output_file=None):
         #ensusre that system only parameter are correctly set
         self.update_system_parameter_for_include()
 
-        value_in_old_include = self.get_last_value_include(output_dir)
-
+        if output_dir: #output_dir is set to None in some unittest
+            value_in_old_include = self.get_last_value_include(output_dir)
+        else:
+           value_in_old_include = {} 
 
         if output_dir:
             self.write_autodef(output_dir, output_file=None)
@@ -3496,7 +3499,6 @@ def write_include_file(self, output_dir, output_file=None):
     def write_one_include_file(self, output_dir, incname, output_file=None):
         """write one include file at the time"""
 
-        misc.sprint(incname)
         if incname is True:
             pathinc = self.default_include_file
         elif incname is False:
@@ -3911,6 +3913,7 @@ def remove_all_cut(self):
    %(aloha_flag)s  = aloha_flag ! fortran optimization flag for aloha function. Suggestions: '-ffast-math'
    %(matrix_flag)s = matrix_flag ! fortran optimization flag for matrix.f function. Suggestions: '-O3'
    %(vector_size)s = vector_size ! size of fortran arrays allocated in the multi-event API for SIMD/GPU (VECSIZE_MEMMAX)
+   %(nb_warp)s = nb_warp ! total number of warp/frontwave
 """
 
 template_off = '# To see advanced option for Phase-Space optimization: type "update psoptim"'
@@ -4314,7 +4317,10 @@ def default_setup(self):
         self.add_param('matrix_flag', '', include=False, hidden=True, comment='fortran compilation flag	for the	matrix-element files, suggestion -O3',
                        fct_mod=(self.make_Ptouch, ('matrix'),{}))        
         self.add_param('vector_size', 1, include='vector.inc', hidden=True, comment='lockstep size for parralelism run', 
-                       fortran_name='VECSIZE_MEMMAX', fct_mod=(self.reset_simd,(),{}))
+                       fortran_name='WARP_SIZE', fct_mod=(self.reset_simd,(),{}))
+        self.add_param('nb_warp', 1, include='vector.inc', hidden=True, comment='number of warp for parralelism run', 
+                       fortran_name='NB_WARP', fct_mod=(self.reset_simd,(),{}))
+        self.add_param('vecsize_memmax', 0, include='vector.inc', system=True)
 
         # parameter allowing to define simple cut via the pdg
         # Special syntax are related to those. (can not be edit directly)
@@ -4605,7 +4611,7 @@ def update_system_parameter_for_include(self):
             self['mxxmin4pdg'] =[0.] 
             self['mxxpart_antipart'] = [False]
 
-                    
+        self['vecsize_memmax'] = self['nb_warp'] * self['vector_size']       
 
     def create_default_for_process(self, proc_characteristic, history, proc_def):
         """Rules

epochX/cudacpp/ee_mumu.sa/CODEGEN_cudacpp_ee_mumu_log.txt

@@ -62,7 +62,7 @@ generate e+ e- > mu+ mu-
 No model currently active, so we import the Standard Model
 INFO: load particles 
 INFO: load vertices 
-DEBUG: model prefixing  takes 0.005564689636230469 
+DEBUG: model prefixing  takes 0.0058383941650390625 
 INFO: Restrict model sm with file models/sm/restrict_default.dat . 
 DEBUG: Simplifying conditional expressions 
 DEBUG: remove interactions: u s w+ at order: QED=1 
@@ -154,25 +154,25 @@ INFO: Checking for minimal orders which gives processes.
 INFO: Please specify coupling orders to bypass this step. 
 INFO: Trying process: e+ e- > mu+ mu- WEIGHTED<=4 @1  
 INFO: Process has 2 diagrams 
-1 processes with 2 diagrams generated in 0.005 s
+1 processes with 2 diagrams generated in 0.004 s
 Total: 1 processes with 2 diagrams
 output standalone_cudacpp ../TMPOUT/CODEGEN_cudacpp_ee_mumu
 Load PLUGIN.CUDACPP_OUTPUT
 Plugin PLUGIN.CUDACPP_OUTPUT has marked as NOT being validated with this version: 3.5.3_lo_vect. 
 It has been validated for the last time with version: 3.5.2
 Output will be done with PLUGIN: CUDACPP_OUTPUT
 DEBUG:  cformat =  plugin [export_cpp.py at line 3070] 
-DEBUG:  Entering PLUGIN_ProcessExporter.__init__ (initialise the exporter) [output.py at line 162] 
-DEBUG:  Entering PLUGIN_ProcessExporter.copy_template (initialise the directory) [output.py at line 167] 
+DEBUG:  Entering PLUGIN_ProcessExporter.__init__ (initialise the exporter) [output.py at line 165] 
+DEBUG:  Entering PLUGIN_ProcessExporter.copy_template (initialise the directory) [output.py at line 170] 
 INFO: Creating subdirectories in directory /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_cudacpp_ee_mumu 
 INFO: Organizing processes into subprocess groups 
 INFO: Generating Helas calls for process: e+ e- > mu+ mu- WEIGHTED<=4 @1 
 INFO: Processing color information for process: e+ e- > mu+ mu- @1 
-DEBUG:  Entering PLUGIN_ProcessExporter.generate_subprocess_directory (create the directory) [output.py at line 209] 
-DEBUG:    type(subproc_group)=<class 'madgraph.core.helas_objects.HelasMatrixElement'> [output.py at line 210] 
-DEBUG:    type(fortran_model)=<class 'PLUGIN.CUDACPP_OUTPUT.model_handling.PLUGIN_GPUFOHelasCallWriter'> [output.py at line 211] 
-DEBUG:    type(me)=<class 'int'> me=0 [output.py at line 212] 
-DEBUG:  "need to link", self.to_link_in_P =  need to link ['nvtx.h', 'timer.h', 'timermap.h', 'ompnumthreads.h', 'GpuRuntime.h', 'GpuAbstraction.h', 'MemoryAccessHelpers.h', 'MemoryAccessVectors.h', 'MemoryAccessMatrixElements.h', 'MemoryAccessMomenta.h', 'MemoryAccessRandomNumbers.h', 'MemoryAccessWeights.h', 'MemoryAccessAmplitudes.h', 'MemoryAccessWavefunctions.h', 'MemoryAccessGs.h', 'MemoryAccessCouplingsFixed.h', 'MemoryAccessNumerators.h', 'MemoryAccessDenominators.h', 'EventStatistics.h', 'CommonRandomNumbers.h', 'CrossSectionKernels.cc', 'CrossSectionKernels.h', 'MatrixElementKernels.cc', 'MatrixElementKernels.h', 'RamboSamplingKernels.cc', 'RamboSamplingKernels.h', 'RandomNumberKernels.h', 'CommonRandomNumberKernel.cc', 'CurandRandomNumberKernel.cc', 'HiprandRandomNumberKernel.cc', 'Bridge.h', 'BridgeKernels.cc', 'BridgeKernels.h', 'fbridge.cc', 'fbridge.inc', 'fsampler.cc', 'fsampler.inc', 'MadgraphTest.h', 'runTest.cc', 'testmisc.cc', 'testxxx_cc_ref.txt', 'valgrind.h', 'cudacpp.mk', 'testxxx.cc', 'MemoryBuffers.h', 'MemoryAccessCouplings.h', 'perf.py', 'profile.sh'] [output.py at line 213] 
+DEBUG:  Entering PLUGIN_ProcessExporter.generate_subprocess_directory (create the directory) [output.py at line 212] 
+DEBUG:    type(subproc_group)=<class 'madgraph.core.helas_objects.HelasMatrixElement'> [output.py at line 213] 
+DEBUG:    type(fortran_model)=<class 'PLUGIN.CUDACPP_OUTPUT.model_handling.PLUGIN_GPUFOHelasCallWriter'> [output.py at line 214] 
+DEBUG:    type(me)=<class 'int'> me=0 [output.py at line 215] 
+DEBUG:  "need to link", self.to_link_in_P =  need to link ['nvtx.h', 'timer.h', 'timermap.h', 'ompnumthreads.h', 'GpuRuntime.h', 'GpuAbstraction.h', 'MemoryAccessHelpers.h', 'MemoryAccessVectors.h', 'MemoryAccessMatrixElements.h', 'MemoryAccessMomenta.h', 'MemoryAccessRandomNumbers.h', 'MemoryAccessWeights.h', 'MemoryAccessAmplitudes.h', 'MemoryAccessWavefunctions.h', 'MemoryAccessGs.h', 'MemoryAccessCouplingsFixed.h', 'MemoryAccessNumerators.h', 'MemoryAccessDenominators.h', 'MemoryAccessChannelIds.h', 'EventStatistics.h', 'CommonRandomNumbers.h', 'CrossSectionKernels.cc', 'CrossSectionKernels.h', 'MatrixElementKernels.cc', 'MatrixElementKernels.h', 'RamboSamplingKernels.cc', 'RamboSamplingKernels.h', 'RandomNumberKernels.h', 'CommonRandomNumberKernel.cc', 'CurandRandomNumberKernel.cc', 'HiprandRandomNumberKernel.cc', 'Bridge.h', 'BridgeKernels.cc', 'BridgeKernels.h', 'fbridge.cc', 'fbridge.inc', 'fsampler.cc', 'fsampler.inc', 'MadgraphTest.h', 'runTest.cc', 'testmisc.cc', 'testxxx_cc_ref.txt', 'valgrind.h', 'cudacpp.mk', 'testxxx.cc', 'MemoryBuffers.h', 'MemoryAccessCouplings.h', 'perf.py', 'profile.sh'] [output.py at line 216] 
 INFO: Creating files in directory /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_cudacpp_ee_mumu/SubProcesses/P1_Sigma_sm_epem_mupmum 
 FileWriter <class 'PLUGIN.CUDACPP_OUTPUT.model_handling.PLUGIN_CPPWriter'> for /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_cudacpp_ee_mumu/SubProcesses/P1_Sigma_sm_epem_mupmum/./CPPProcess.h
 FileWriter <class 'PLUGIN.CUDACPP_OUTPUT.model_handling.PLUGIN_CPPWriter'> for /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_cudacpp_ee_mumu/SubProcesses/P1_Sigma_sm_epem_mupmum/./CPPProcess.cc
@@ -183,7 +183,7 @@ ALOHA: aloha creates FFV1 routines
 ALOHA: aloha creates FFV2 routines
 ALOHA: aloha creates FFV4 routines
 ALOHA: aloha creates FFV2_4 routines
-ALOHA: aloha creates 4 routines in  0.266 s
+ALOHA: aloha creates 4 routines in  0.269 s
 <class 'aloha.create_aloha.AbstractRoutine'> FFV1
 <class 'aloha.create_aloha.AbstractRoutine'> FFV1
 <class 'aloha.create_aloha.AbstractRoutine'> FFV2
@@ -202,7 +202,7 @@ INFO: Created files Parameters_sm.h and Parameters_sm.cc in directory
 INFO: /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_cudacpp_ee_mumu/src/. and /data/avalassi/GPU2023/madgraph4gpuX/MG5aMC/TMPOUT/CODEGEN_cudacpp_ee_mumu/src/. 
 quit
 
-real	0m0.666s
-user	0m0.592s
-sys	0m0.058s
-Code generation completed in 0 seconds
+real	0m0.667s
+user	0m0.596s
+sys	0m0.059s
+Code generation completed in 1 seconds

epochX/cudacpp/gg_tt.mad/CODEGEN_mad_gg_tt_log.txt

@@ -62,7 +62,7 @@ generate g g > t t~
 No model currently active, so we import the Standard Model
 INFO: load particles 
 INFO: load vertices 
-DEBUG: model prefixing  takes 0.005675077438354492 
+DEBUG: model prefixing  takes 0.005654811859130859 
 INFO: Restrict model sm with file models/sm/restrict_default.dat . 
 DEBUG: Simplifying conditional expressions 
 DEBUG: remove interactions: u s w+ at order: QED=1 
@@ -197,11 +197,11 @@ Wrote files for 10 helas calls in 0.084 s
 ALOHA: aloha starts to compute helicity amplitudes
 ALOHA: aloha creates VVV1 set of routines with options: P0
 ALOHA: aloha creates FFV1 routines
-ALOHA: aloha creates 2 routines in  0.149 s
+ALOHA: aloha creates 2 routines in  0.145 s
 ALOHA: aloha starts to compute helicity amplitudes
 ALOHA: aloha creates VVV1 set of routines with options: P0
 ALOHA: aloha creates FFV1 routines
-ALOHA: aloha creates 4 routines in  0.133 s
+ALOHA: aloha creates 4 routines in  0.135 s
 <class 'aloha.create_aloha.AbstractRoutine'> VVV1
 <class 'aloha.create_aloha.AbstractRoutine'> FFV1
 <class 'aloha.create_aloha.AbstractRoutine'> FFV1
@@ -234,9 +234,9 @@ Type "launch" to generate events from this process, or see
 Run "open index.html" to see more information about this process.
 quit
 
-real	0m1.905s
+real	0m2.031s
 user	0m1.645s
-sys	0m0.257s
+sys	0m0.254s
 Code generation completed in 2 seconds
 ************************************************************
 *                                                          *