The tensor module uses its own random generator which is by default initialized based on the system's random generator and the current system time.
This makes reproducing/debugging failing unit-tests harder than necessary.
Generally, I think, random number generation should be shared between the Core module and the Tensor module.
But at least during unit-tests there should be no seeds based on the current system time (except in the main() method, when no seed is passed).
-- GitLab Migration Automatic Message --
This bug has been migrated to gitlab.com's GitLab instance and has been closed from further activity.
You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.com/libeigen/eigen/issues/1602.