Skip to content

Conversation

@LiamOS
Copy link
Member

@LiamOS LiamOS commented Aug 29, 2024

While trying to build duneanaobj with spack I noticed that it completed rather easily, later realising that the CMake files effectively depend on UPS/CET and do something between nothing and something useful when not found.

I've made some changes to the CMake files to build and install correctly. I haven't validated this fully, more changes may come if/when I find something I missed.

LiamOS added 30 commits August 27, 2024 15:34
@LiamOS LiamOS requested a review from chenel August 29, 2024 15:56
@LiamOS LiamOS changed the title Minimal edit spack build Generalising CMake Aug 29, 2024
Copy link
Collaborator

@chenel chenel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we will need to wait until @tomjunk comes back from Fermilab shutdown to review this---I'm completely unclear on exactly what's good or not good from a Spack perspective.

@chenel chenel requested a review from tomjunk August 29, 2024 16:04
@LiamOS
Copy link
Member Author

LiamOS commented Aug 30, 2024

Spack is pretty flexible; as long as the build system is configurable in a somewhat normal way, the spack packaging can handle any quirks. The main expectation is that it build and install in a somewhat similar way to how most software would be built with CMake, which this would not if it could not find UPS stuff.

I think the changes I've made shouldn't break the SL7 builds, but would need someone to validate. Although this may become a non-issue when (...if?) spack becomes our everyday driver.

There's also no rush on merging this, for testing I'm telling spack to pull from this branch rather than main.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants