-
Notifications
You must be signed in to change notification settings - Fork 33
New issue
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
Option to produce multi-build (fortran, cuda, cpp) gridpack #945
Comments
This was referenced Aug 7, 2024
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…ILDDIR=1 to make_opts (part of madgraph5#945) Then regenerate gg_tt.mad: this only affects launch_plugin.py itself Launching lauX.sh on ggtt fails as follows Error detected in "generate_events -f" write debug file /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gg_tt.mad/run_01_tag_1_debug.log If you need help with this issue please contact us on https://answers.launchpad.net/mg5amcnlo str : A compilation Error occurs when trying to compile /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gg_tt.mad/SubProcesses/P1_gg_ttx. The compilation fails with the following output message: make USEGTEST=0 BACKEND=cuda build.auto_m_inl0_hrd0/madevent_cuda make[1]: Entering directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gg_tt.mad/SubProcesses/P1_gg_ttx' make[1]: *** No rule to make target 'build.auto_m_inl0_hrd0/madevent_cuda'. Stop. make[1]: Leaving directory '/data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gg_tt.mad/SubProcesses/P1_gg_ttx' make: *** [makefile:163: madevent_cuda_link] Error 2 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
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
… madevent_xxx_link' in args[0][0] to support USEBUILDDIR=1 (part of madgraph5#945) Then regenerate gg_tt.mad However lauX.sh still fails
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…][1] to 'make madevent_xxx_link' to support USEBUILDDIR=1 (part of madgraph5#945) Then regenerate gg_tt.mad
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…two patches, will fix gg_tt.mad makefile instead Revert "[runcard] in CODEGEN launch_plugin.py, append 'BACKEND=xxx' as args[0][1] to 'make madevent_xxx_link' to support USEBUILDDIR=1 (part of madgraph5#945)" This reverts commit ce67fb5. Revert "[runcard] in CODEGEN launch_plugin.py, prepend 'BACKEND=xxx' to 'make madevent_xxx_link' in args[0][0] to support USEBUILDDIR=1 (part of madgraph5#945)" This reverts commit b51a59a.
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…UILDDIR=1' (part of madgraph5#945)
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…targets instead of having this in a python script (part of madgraph5#945)
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…_opts configuration, this will be hardcoded in link targets (part of madgraph5#945)" Revert "[runcard] in CODEGEN launch_plugin.py, configure compile to add USEBUILDDIR=1 to make_opts (part of madgraph5#945)" This reverts commit 853a20e.
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…(and propagate FPTYPE from make_opts) in _link targets for madgraph5#945 The only files that still need to be patched are - 3 in patch.common: Source/makefile, Source/genps.inc, SubProcesses/makefile - 3 in patch.P1: auto_dsig1.f, driver.f, matrix1.f ./CODEGEN/generateAndCompare.sh gg_tt --mad --nopatch git diff --no-ext-diff -R gg_tt.mad/Source/makefile gg_tt.mad/Source/genps.inc gg_tt.mad/SubProcesses/makefile > CODEGEN/PLUGIN/CUDACPP_SA_OUTPUT/MG5aMC_patches/PROD/patch.common git diff --no-ext-diff -R gg_tt.mad/SubProcesses/P1_gg_ttx/auto_dsig1.f gg_tt.mad/SubProcesses/P1_gg_ttx/driver.f gg_tt.mad/SubProcesses/P1_gg_ttx/matrix1.f > CODEGEN/PLUGIN/CUDACPP_SA_OUTPUT/MG5aMC_patches/PROD/patch.P1 git checkout gg_tt.mad
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…ards madgraph5#945, then regenerate gg_tt.mad
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…s and remove it from _link targets madgraph5#945 (Also fix the cpp targets where I had forgotten one part of my previous patch) The only files that still need to be patched are - 3 in patch.common: Source/makefile, Source/genps.inc, SubProcesses/makefile - 3 in patch.P1: auto_dsig1.f, driver.f, matrix1.f ./CODEGEN/generateAndCompare.sh gg_tt --mad --nopatch git diff --no-ext-diff -R gg_tt.mad/Source/makefile gg_tt.mad/Source/genps.inc gg_tt.mad/SubProcesses/makefile > CODEGEN/PLUGIN/CUDACPP_SA_OUTPUT/MG5aMC_patches/PROD/patch.common git diff --no-ext-diff -R gg_tt.mad/SubProcesses/P1_gg_ttx/auto_dsig1.f gg_tt.mad/SubProcesses/P1_gg_ttx/driver.f gg_tt.mad/SubProcesses/P1_gg_ttx/matrix1.f > CODEGEN/PLUGIN/CUDACPP_SA_OUTPUT/MG5aMC_patches/PROD/patch.P1 git checkout gg_tt.mad
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
…to grid Fix conflicts: epochX/cudacpp/tlau/lauX.sh
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Aug 7, 2024
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Oct 7, 2024
…ranch grid This includes in summary - madgraph5#700 add cudacpp_helinl, cudacpp_hrdcod, rename cudacpp_fptype - madgraph5#945 add cudacpp_bldall for multi-backend gridpacks - madgraph5#957 enhance tlau tests: instrument python code for gridpack profiling, add test logs and scripts to parse them
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
I am looking at #943: why does CMS see speedups in DY+4jets but not DY+3 jets
The way I would like to address this is to produce a multi-build (fortran, cuda, cpp) gridpack
That is to say:
This is related to #674 about madevent_fortran, madevent_cuda, madevent_cpp.
The text was updated successfully, but these errors were encountered: