[RFC] Proposal to merge the Patatrack development into CMSSW 10.5.x [squashed] #25647
Conversation
cmsbuild
added
alca-pending
code-checks-pending
comparison-pending
core-pending
db-pending
dqm-pending
fastsim-pending
geometry-pending
new-package-pending
operations-pending
orp-pending
pdmv-pending
pending-signatures
reconstruction-pending
simulation-pending
tests-pending
upgrade-pending
labels
|
The code-checks are being triggered in jenkins. |
|
enable GPU |
|
@cmsbuild, please test |
|
-code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-25647/7978
Code check has found code style and quality issues which could be resolved by applying a patch in https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-25647/7978/git-diff.patch You can run |
fwyzard
force-pushed
the
Patatrack_developments_104x
branch
from
4cc0f76
to
451ec31
Compare
| try { | ||
| auto error = cudaGetErrorName(status); | ||
| auto message = cudaGetErrorString(status); | ||
| throw cms::Exception("CUDAError") << "Callback of CUDA stream " << streamId << " in device " << deviceId << " error " << error << ": " << message; |
There was a problem hiding this comment.
I think this can be replaced with std::make_exception_ptr.
There was a problem hiding this comment.
we were using that before, but I changed to a throw in order to be able to catch throw it in GDB
There was a problem hiding this comment.
You could change catch(..) to catch(cms::Exception const&) instead, since you know the type.
There was a problem hiding this comment.
ah, I see what you mean - sure, that makes sense.
I think we could suppress it using https://clang.llvm.org/extra/clang-tidy/#id3 |
|
For the moment I have added a whitelist to out python checks, same as for the |
|
by the way @kpedro88, do you have any idea why I do not get the same error when testing locally, on top of the latest IB ? |
@Dr15Jones not that I know of. |
|
@fwyzard I am not sure, the PR that added .7 workflows has been in IBs since |
Take into account the proper number of modules: increment by one pathsAndConsumes.allModules().size(), because it does not include the "source" module. Add asserts to check that all ranges are properly closed. Optionally, delay starting the profiler until after the first event on each stream has completed. This requires running 'nvprof' with the '--profile-from-start off' option.
Squash git@github.com:cms-patatrack/cmssw.git/CMSSW_10_4_X_Patatrack on top of CMSSW_10_5_X_2019-01-23-2300 .
fwyzard
force-pushed
the
Patatrack_developments_104x
branch
from
fdb21bd
to
57d562b
Compare
|
The code-checks are being triggered in jenkins. |
|
-code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-25647/8173
Code check has found code style and quality issues which could be resolved by applying a patch in https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-25647/8173/git-diff.patch You can run |
|
Will try again... |
This should be equivalent to #25353 after fixing conflicts and squashing on top of CMSSW_10_4_0 .
Currently, not all the developments we would like to integrate have been merged in the development branch: see cms-patatrack#200 for the status of the Patatrack branch with respect to readiness for merging upstream.