Skip to content

Conversation

@davidrohr
Copy link
Collaborator

@MichaelLettrich : First attempt to improve my CMake for the GPU Standalone build:

  • It now requires CMake 3.14 (since I have been using 3.14 syntax, as you figured out...)
  • It should auto-detect ms_gsl.
  • Disabling the standalone event display in the config file should work now, also for Vc / fmt / OpenMP.
  • Disabling ROOT doesn't work right now (Still works with the old makefile). It is too deeply integrated in our CMake. Perhaps I'll reevaluate next time I have remote access to a GPU machine without ROOT, but for now this is not needed.
  • For CUB / FMT, you'll still need to provide CMake options -Dfmt_DIR=[SW_DIR]/fmt/latest/lib64/cmake/fmt/ -Dcub_ROOT=.... (Unfortunately, I cannot autodetect them, since our alidist recipes do not set environment variables for them, and I don't want to mess with the recipes for my standalone tool... If they are installed systemwide, they should be detected correctly.)
  • The rest I just tested in alidock, and it worked for me.

@davidrohr davidrohr requested review from a team and ktf as code owners July 31, 2019 23:02
@MichaelLettrich
Copy link
Collaborator

Does that mean we should bump CMake as well?

@davidrohr
Copy link
Collaborator Author

We will globally bump CMake to 3.15 once that is out I think because the next FairROOT/DDS will need the new CMake.

@davidrohr
Copy link
Collaborator Author

@Barthelemy : It seems this QC test breaks the O2 CI for my PR:
5/17 Test #6: testTaskInterface ................***Timeout 30.06 sec

@davidrohr davidrohr merged commit f29447a into AliceO2Group:dev Aug 1, 2019
EmilGorm pushed a commit to EmilGorm/AliceO2 that referenced this pull request Apr 15, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Development

Successfully merging this pull request may close these issues.

2 participants