Conversation
Signed-off-by: Alejandro Hernández Cordero <ahcorde@gmail.com>
|
Why this change? The context is initialized in the |
fujitatomoya
left a comment
There was a problem hiding this comment.
lgtm. the fix itself makes sense to me that we call shutdown explicitly once it completes the all test cases with this fixture.
i would also like to know @alsora 's comment above. i think default_context will be destructed when the process space is destroyed, then context will be shutdown. What kind of problem did you have with rmw_zenoh?
|
@ahcorde any chance you opened this PR because you ran into the If so I wonder if my comment here explains why we need this PR. |
There was a problem hiding this comment.
Please see ros2/rmw_zenoh#170 (comment) for why this is needed.
TLDR; Leaving the context's destructor to shutdown the context at program termiantion causes Zenoh to panic since Thread local storage gets cleared earlier.
|
CI started with this repos file. |
Signed-off-by: Yadunund <yadunund@intrinsic.ai>
4b50b61 to
2fa9c92
Compare
Signed-off-by: Yadunund <yadunund@intrinsic.ai>
|
I took the liberty to adopt the change into other tests in @fujitatomoya @alsora kindly review this PR whenever time permits! Re-run CI after pushing changes: |
Signed-off-by: Yadunund <yadunund@intrinsic.ai>
clalancette
left a comment
There was a problem hiding this comment.
Generally, these look good to me, but there is one that I think we should revamp.
Signed-off-by: Alejandro Hernández Cordero <ahcorde@gmail.com>
|
Still failing here |
clalancette
left a comment
There was a problem hiding this comment.
Looks good to me with green CI.
|
Pulls: #2633 |
This is something we need to fix in @ahcorde looks like uncrustify is failing. |
Signed-off-by: Alejandro Hernández Cordero <ahcorde@gmail.com>
|
Merging with green CI! |
No description provided.