-
-
Notifications
You must be signed in to change notification settings - Fork 32
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
Work around world age error for Duals in exp_generic #43
Conversation
generated functions that causes world age errors in exp_pade_p if the function has been called before ForwardDiff was loaded.
Please hold off a bit. I need to define the zero branch differently since
|
Fixed it so I believe this one should be good to go. |
N = randn(n, n) | ||
@test exp(N) ≈ exp_generic(N) | ||
|
||
exp(n) ≈ exp_generic(n) | ||
end | ||
end | ||
|
||
# The issue was only triggered if exp_generic had been called before loading |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This should be incorrect. What matters is the load order, while it's a bug if the call order is observable.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
how'd you find this? 😆 I think we can just straight avoid the issue for the most common case (#45) and handle the more general thing later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Someone put out the bat signal 🙃
Superseded by #45 |
and fix
exp_generic(0.0)
.Closes #42 but not really #41. It's more a temporary workaround than an actual fix. I'm not sure how to fix the problem in full generality.