Skip to content
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

AMReX_FLATTEN_FOR #3855

Merged
merged 2 commits into from
Mar 28, 2024
Merged

Conversation

WeiqunZhang
Copy link
Member

A new build option to force flattening of ParallelFor and similar functions for host device. The default is yes unless it's a debug build. The CMake option is AMReX_FLATTEN_FOR, whereas for GNU make, it's USE_FLATTEN_FOR.

A new build option to force flattening of ParallelFor and similar functions
for host device. The default is yes unless it's a debug build. The CMake
option is AMReX_FLATTEN_FOR, whereas for GNU make, it's USE_FLATTEN_FOR.
@WeiqunZhang WeiqunZhang requested a review from atmyers March 27, 2024 21:00
@WeiqunZhang WeiqunZhang linked an issue Mar 27, 2024 that may be closed by this pull request
Copy link
Member

@ax3l ax3l left a comment

Choose a reason for hiding this comment

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

That looks good! 👍

@WeiqunZhang
Copy link
Member Author

A GCC CI failed with g++-8: fatal error: Killed signal terminated program cc1plus. I cannot reproduce it with GCC 9. May have to disable flattening in that test.

@WeiqunZhang WeiqunZhang enabled auto-merge (squash) March 28, 2024 01:45
@WeiqunZhang WeiqunZhang merged commit 8d57ebc into AMReX-Codes:development Mar 28, 2024
68 of 69 checks passed
asalmgren pushed a commit that referenced this pull request Mar 29, 2024
Flattening (#3855) makes compilation too slow for some codes. For
example, on my desktop, the compilation time for Castro_react.cpp
increased from 3 minutes to 49 minutes.
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.

Unexpected Performance penality when using amrex::ParallelFor
2 participants