Closed valassi closed 7 months ago
I think I have essentially fixed all of these issues. Now cross checking them one by one.
- In "'import model MSSM_SLHA2; generate g g > t t~ t t~", both builds succeed, Strangely however the ME results are different in HRDCOD=1 with respect to the reference file produced with HRDCOD=0
This is fixed. The HRDCOD=1 results were right, the HRDCOD=0 results were wrong but are now fixed (some value assignments were missing).
- In "import model MSSM_SLHA2; generate g g > go go", the HRDCOD=1 build succeeds, but the HRDCOD=0 build fails
This is also fixed
- In "import model MSSM_SLHA2; generate g g > t1 t1~", HRDCOD=0 builds fail
- In "import model MSSM_SLHA2; generate g g > t1 t1~", HRDCOD=1 builds also fail
These are also fixed. Since this is a true and simple SUSY process, I aded this to the repo (and I removed susy_gg_tt, which just includes susy contrubtions to an existing SM process)
- In "import model MSSM_SLHA2; generate g g > ul ul~", code generation fails
This is also fixed.
All codegen, builds and basic tests for the processes above are all succeeding and giving internally consistent results.
--
Summary: this is fixed in PR #824 - closing for simplicity
The builds for one simple process gg>tt~ in the SUSY model is now fixed by PR #625 (for both HRDCOD=0 and HRDCOD=1).
I tried to investigate a few more SUSY processes, and unfortunately I still found other issues (which sometimes look very similar to those I already fixed).
In "'import model MSSM_SLHA2; generate g g > t t~ t t~", both builds succeed, Strangely however the ME results are different in HRDCOD=1 with respect to the reference file produced with HRDCOD=0
In "import model MSSM_SLHA2; define q = u c d s u~ c~ d~ s~; generate g q > t t~ q", the HRDCOD=0 and HRDCOD=1 builds succeed and give the same results as each other. This is true in both P1 subdirectories.
In "import model MSSM_SLHA2; generate g g > go go", the HRDCOD=1 build succeeds, but the HRDCOD=0 build fails
In "import model MSSM_SLHA2; generate g g > t1 t1~", HRDCOD=0 builds fail
In "import model MSSM_SLHA2; generate g g > t1 t1~", HRDCOD=1 builds also fail
Parameters_MSSM_SLHA2.h(835): error: no operator "" matches these operands operand types are: thrust::complex const mgOnGpu::cxsmpl
Parameters_MSSM_SLHA2.h(836): error: no operator "" matches these operands operand types are: thrust::complex const mgOnGpu::cxsmpl
Parameters_MSSM_SLHA2.h(836): error: no operator "" matches these operands operand types are: thrust::complex const mgOnGpu::cxsmpl
Parameters_MSSM_SLHA2.h(837): error: no operator "" matches these operands operand types are: thrust::complex const mgOnGpu::cxsmpl
Parameters_MSSM_SLHA2.h(837): error: no operator "" matches these operands operand types are: thrust::complex const mgOnGpu::cxsmpl
File "/data/avalassi/GPU2023/madgraph4gpuBis/MG5aMC/mg5amcnlo/PLUGIN/CUDACPP_OUTPUT/model_handling.py", line 1543, in super_write_p rocess_cc_file replace_dict = super(PLUGIN_export_cpp.OneProcessExporterGPU, self).write_process_cc_file(False) File "/data/avalassi/GPU2023/madgraph4gpuBis/MG5aMC/mg5amcnlo/madgraph/iolibs/export_cpp.py", line 785, in write_process_cc_file self.get_process_function_definitions() File "/data/avalassi/GPU2023/madgraph4gpuBis/MG5aMC/mg5amcnlo/PLUGIN/CUDACPP_OUTPUT/model_handling.py", line 1154, in get_process_f unction_definitions params[pos] = para IndexError: list assignment index out of range