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.

Sign up for GitHub

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

Jump to bottom

Regression test problems with GNU compilers #265

Closed

climbfuji opened this issue Jun 10, 2019 · 1 comment

Collaborator

climbfuji commented Jun 10, 2019

This issue is a placeholder for regression test / other bit-for-bit testing issues with the GNU compiler.

Known problems at the time of creation:

fv3_ccpp_2threads: on Cheyenne with GNU 6.3.0, this test gives different results for the following files for the dynamic CCPP build than for the static CCPP build and the IPD build (the latter two are identical) in REPRO mode:

 Comparing atmos_4xdaily.tile1.nc ............NOT OK
 Comparing atmos_4xdaily.tile2.nc ............NOT OK
 Comparing atmos_4xdaily.tile3.nc ............NOT OK
 Comparing atmos_4xdaily.tile4.nc ............NOT OK
 Comparing atmos_4xdaily.tile5.nc ............NOT OK
 Comparing atmos_4xdaily.tile6.nc ............NOT OK

fv3_ccpp_gfdlmp_32bit gives different results for IPD build than for the two CCPP builds (the latter two are identical) on macOS with GNU 8.1.0

The text was updated successfully, but these errors were encountered:

climbfuji mentioned this issue

Correct way to set 32-bit compiler flags for fast physics #212

Closed

Collaborator Author

climbfuji commented Dec 10, 2019

These issues have been solved in #212 and by upgrading to a newer version of the GNU compiler on Cheyenne (8.3.0).

climbfuji closed this as completed

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