Skip to content

Conversation

@haampie
Copy link
Collaborator

@haampie haampie commented Apr 3, 2020

The pipeline definition is stored here https://gitlab.com/cscs-ci/eth-cscs/cosma-ci

CI should be triggered by bors, let's see if that works as well.

@haampie
Copy link
Collaborator Author

haampie commented Apr 3, 2020

bors try

@bors
Copy link
Contributor

bors bot commented Apr 3, 2020

🔒 Permission denied

Existing reviewers: click here to make haampie a reviewer

@haampie
Copy link
Collaborator Author

haampie commented Apr 3, 2020

Ok... I don't have access 😅, @kabicm, can you either comment bors try or give me push access temporarily?

@kabicm
Copy link
Collaborator

kabicm commented Apr 3, 2020

bors try

@bors
Copy link
Contributor

bors bot commented Apr 3, 2020

try

Configuration problem

bors.toml: not found

@haampie
Copy link
Collaborator Author

haampie commented Apr 3, 2020

bors try

bors bot added a commit that referenced this pull request Apr 3, 2020
@bors
Copy link
Contributor

bors bot commented Apr 3, 2020

try

Build succeeded

@haampie
Copy link
Collaborator Author

haampie commented Apr 3, 2020

So, what this did is:

  • it created a build environment with CUDA, MKL and MPICH here

  • and compiled COSMA for this PR and bundled it with its dependencies to a small docker image here

    Try

    docker run -it --rm registry.gitlab.com/cscs-ci/eth-cscs/cosma/deploy:9232a10d37da252ae5627826609dbcb1ef46ae9a /bin/bash
    

    to see the contents.

  • then pulled this image on Daint here, but it's not running it yet. That should still be implemented here.

The second time the build environment is cached as long as you don't bump MKL, CUDA Toolkit or MPICH, so let's see if that actually works as expected.

bors try

bors bot added a commit that referenced this pull request Apr 3, 2020
@bors
Copy link
Contributor

bors bot commented Apr 3, 2020

try

Build succeeded

@teonnik teonnik requested review from kabicm and teonnik April 3, 2020 18:25
@kabicm kabicm added the feature label Apr 3, 2020
@haampie
Copy link
Collaborator Author

haampie commented Apr 7, 2020

bors try

bors bot added a commit that referenced this pull request Apr 7, 2020
@bors
Copy link
Contributor

bors bot commented Apr 7, 2020

try

Build failed

@haampie
Copy link
Collaborator Author

haampie commented Apr 7, 2020

bors try

bors bot added a commit that referenced this pull request Apr 7, 2020
@bors
Copy link
Contributor

bors bot commented Apr 7, 2020

try

Build failed

@haampie
Copy link
Collaborator Author

haampie commented Apr 7, 2020

Last attempt hopefully? 😅

bors try

bors bot added a commit that referenced this pull request Apr 7, 2020
@bors
Copy link
Contributor

bors bot commented Apr 7, 2020

try

Build failed

@haampie
Copy link
Collaborator Author

haampie commented Apr 7, 2020

@kabicm @teonnik can you maybe comment on what's going wrong here?

https://gitlab.com/cscs-ci/eth-cscs/COSMA/-/jobs/501552983

@kabicm
Copy link
Collaborator

kabicm commented Apr 7, 2020

@haampie Hard to say, is this using the latest commit on master?

@haampie
Copy link
Collaborator Author

haampie commented Apr 9, 2020

Currently I'm only building the GPU version. I could add a CPU version as well!

COPY . /COSMA

RUN mkdir /COSMA/build && cd /COSMA/build && \
CC=mpicc CXX=mpicxx FC=mpif90 FCCPP=cpp cmake .. \
Copy link
Collaborator

Choose a reason for hiding this comment

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

I think FC=mpif90 FCCPP=cpp are not necessary since we don't use fortran compilers.

@haampie
Copy link
Collaborator Author

haampie commented Apr 9, 2020

now with CPU and GPU both:

bors try

bors bot added a commit that referenced this pull request Apr 9, 2020
@bors
Copy link
Contributor

bors bot commented Apr 9, 2020

try

Build failed

@haampie
Copy link
Collaborator Author

haampie commented Apr 9, 2020

bors try

bors bot added a commit that referenced this pull request Apr 9, 2020
@bors
Copy link
Contributor

bors bot commented Apr 9, 2020

try

Build failed

@haampie
Copy link
Collaborator Author

haampie commented Apr 9, 2020

Now with new env variables for tile sizes

bors try

bors bot added a commit that referenced this pull request Apr 9, 2020
@bors
Copy link
Contributor

bors bot commented Apr 9, 2020

try

Build failed

@haampie
Copy link
Collaborator Author

haampie commented Apr 9, 2020

Forgot to recursively clone the repo; another try

bors try

bors bot added a commit that referenced this pull request Apr 9, 2020
@bors
Copy link
Contributor

bors bot commented Apr 9, 2020

try

Build succeeded

@haampie
Copy link
Collaborator Author

haampie commented Apr 9, 2020

😌

@kabicm
Copy link
Collaborator

kabicm commented Apr 9, 2020

This looks great @haampie! If @teonnik agrees, I think this is ready for merging, I am really looking forward to it!

Copy link
Collaborator

@teonnik teonnik left a comment

Choose a reason for hiding this comment

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

Looks good to me! :)

@haampie
Copy link
Collaborator Author

haampie commented Apr 10, 2020

Thanks @teonnik :) I've pushed the changes

@teonnik
Copy link
Collaborator

teonnik commented Apr 10, 2020

bors merge

@bors
Copy link
Contributor

bors bot commented Apr 10, 2020

Build succeeded

@bors bors bot merged commit c8f020a into eth-cscs:master Apr 10, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants