add time when timer expired to timer events #1130
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Adjusted the test to compare the expected time to when the timer expired (callback added to the queue) instead of when the callback is actually called... Instead of adding a new |
|
I went ahead and merged #1132 for now since it addresses the problem by only changing the test. If you think the changes in this patch are still useful we can keep this PR open. Currently the patch is ABI-incompatible since it e.g. inserts new arguments before existing ones. |
When the a timer expires, the callback is added to the queue. So the callback is then then poped from the queue at some later point (for which the time is recored as current_real in the timer event). This adds the time of when the timer actually expired to the timer event info, in the hope that this will make debugging easier and provide a better base to test the timers.
Instead of checking when the callback was actually called, check for when the timer expired (the callback was added to the callback queue) using the new current_expired time. This *should* make the test more reliable.
|
Some numbers running this on my i5 laptop with As somewhat expected the main source of the time difference is not that the timer expired too late, but that the callback is popped too late from the queue. In the late cases the time difference between timer expiration (callback gets added to queue) and when the callback is actually called is almost always over 0.1s. |
|
I would say that is expected behavior on a non-realtime kernel. Since the patch still changes the API I would think it shouldn't be merged into already released distros. Therefore I would suggest to merge it for ROS M-turtle. |
mikepurvis
added a commit
that referenced
this pull request
Nov 3, 2017
mikepurvis
added a commit
that referenced
this pull request
Nov 15, 2017
|
@ros-pull-request-builder retest this please |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
When the a timer expires, the callback is added to the queue.
So the callback is then then popped from the queue at some later point (for which the time is recored as current_real in the timer event).
This adds the time of when the timer actually expired to the timer event info,
in the hope that this will make debugging easier and provide a better base to test the timers.
I'm not really sure if it would be ok to add something to the timer events...
This is mainly meant as a base for discussion (see #1120 and #1129) and the possibility to test if the timer test can be improved using the
current_expiredtime instead ofcurrent_real.