-
Notifications
You must be signed in to change notification settings - Fork 33
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Port post-generation patches (patchMad.sh) upstream to mg5amcnlo or to the plugin #656
Comments
Note, this is part of point 2 in #576 |
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 2, 2023
…NAL) The plan is to eventually copy these files from here rather than using patch.P1 and patch.common madgraph5#656
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 2, 2023
…ified Fortran makefile if this is madevent + cudacpp madgraph5#656
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 2, 2023
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 2, 2023
This was referenced Jun 2, 2023
Closed
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 3, 2023
…NAL) The plan is to eventually copy these files from here rather than using patch.P1 and patch.common madgraph5#656
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 3, 2023
…ified Fortran makefile if this is madevent + cudacpp madgraph5#656
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 3, 2023
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 3, 2023
I have started a WIP patch for this in MR #675 |
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 12, 2023
…g a modified MG5AMC for codegen I used it to understand that run_card.dat is created by export_v4.py at the very end of everything... It seems very difficult to imagine I can modify the runcard in the cudacpp plugin: to add the -O3 global flags, I will have to do it in patchMad madgraph5#656 - and tell users to do it themselves By the way I still need to test that the fastmath actually does not change much... madgraph5#252 I will therefore move both make_opts and runcard to patch.common and patch.P1...
valassi
added a commit
to valassi/madgraph4gpu
that referenced
this issue
Jun 12, 2023
…5 more files in patch.common ./CODEGEN/generateAndCompare.sh gg_tt --mad --nopatch git diff --no-ext-diff -R gg_tt.mad/Source/dsample.f gg_tt.mad/Source/genps.inc gg_tt.mad/Source/vector.inc gg_tt.mad/SubProcesses/makefile > CODEGEN/MG5aMC_patches/PROD/patch.common git diff --no-ext-diff -R gg_tt.mad/Cards/run_card.dat gg_tt.mad/Source/make_opts gg_tt.mad/bin/internal/banner.py gg_tt.mad/bin/internal/gen_ximprove.py gg_tt.mad/bin/internal/madevent_interface.py >> CODEGEN/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/MG5aMC_patches/PROD/patch.P1 git checkout gg_tt.mad Clearly, this will make the removal of patchMad.sh more complex madgraph5#656
The strategy for integrating patchMad.sh patches upstream has changed:
This will be closed when we merge #747 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Presently code generation in cudacpp goes through the patchMad.sh script, currently:
https://github.com/madgraph5/madgraph4gpu/blob/3a4850184f1de483bbab1bff5a6a322590692291/epochX/cudacpp/CODEGEN/patchMad.sh
This applies post-generation patches to the code generated using the python madgraph machinery. For a release, these patches must be instead moved upstream to mg5amcnlo (or to the plugin itself).
The main ingredients of this are the following (I may be forgetting a few):
Assigning this to me, cc @roiser and @oliviermattelaer
The text was updated successfully, but these errors were encountered: