-
Notifications
You must be signed in to change notification settings - Fork 769
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
Fix Windows Tests #1553
Fix Windows Tests #1553
Conversation
@varunagrawal You asked my help for these tests. Good luck to hunt more bugs! Edit: |
When compiling with ninja, it make the error more clear:
|
By the way, I am not sure if you notice. Even that the ci pass, some of the tests are not compile with missing symbols. |
testBasisFactors.cpp.obj : error LNK2019: unresolved external symbol "class Eigen::Matrix<double,-1,-1,0,-1,-1> __cdecl gtsam::kroneckerProductIdentity(unsigned __int64,class Eigen::Matrix<double,1,-1,1,1,-1> const &)" (?kroneckerProductIdentity@gtsam@@ya?AV?$Matrix@N$0?0$0?0$0A@$0?0$0?0@Eigen@@_KAEBV?$Matrix@N$00$0?0$00$00$0?0@3@@z) referenced in function "protected: void __cdecl gtsam::Basis::VectorDerivativeFunctor::calculateJacobian(void)" (?calculateJacobian@VectorDerivativeFunctor@?$Basis@VChebyshev2@gtsam@@@gtsam@@IEAAXXZ) |
Same reasoning as above. :)
Nice catch! |
Yup I am aware. That's the reason this PR is still a draft. :) |
@varunagrawal Thank you the explanation. I was did a second review check, and I understand it my own that it ok. I edit my comment, and leave the original msg with strikes. |
@talregev @mikesheffler @dellaert I fixed all the |
@varunagrawal Are you interesting to deal with these errors? |
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.
You should squash the commits. It not recommended to have all these commits and merges to dev branch.
We prefer to keep the history so we can go back and identify bad commits and undo them. This is why we don't squash or force-push. |
Why is it using |
@talregev can you please approve this PR so I can merge it? The CI should pass which is a good indicator that the fixes work. |
This is another bug you have in the cmake when I compile it with vcpkg. But the debug assertion still happen on your side. Are you interesting to solve it? |
I have never had this issue. I have been compiling and debugging using the Release config for the last week and it seems to be fine. Maybe the issue is in your vcpkg settings?
I do not get this error on my end. I re-ran the tests again and they pass without problems. |
Let's deal with this later. It not for this PR.
Of course there isn't errors. It not enable. I can enable it and you will see these errors. |
Sorry I don't understand. What is not enabled? |
The flag that enable the debug assertion that enable extra checks on run time. |
Can you please share the changes here? I can run them locally. |
I will. It will be later. |
Then can you please approve this PR? |
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.
Cool! If this is ready to merge, go for it :-)
My attempt at fixing #1541.
size_t
is interpreted since themultiple symbols defined
error comes from additional declarations ofstd::map<Key, size_t>
. This might be a bug in MSVC 16.gtsam_unstable
tests.There is still quite a bit of work to do to get all the tests passing on Windows, but hopefully this is a good start point.