-
Notifications
You must be signed in to change notification settings - Fork 47
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
Complete init/shutdown API test coverage. #107
Conversation
Signed-off-by: Michel Hidalgo <michel@ekumenlabs.com>
These tests need ros2/rmw#242 and connected PRs to pass. |
rmw_context_t context = rmw_get_zero_initialized_context(); | ||
rmw_init_options_t options = rmw_get_zero_initialized_init_options(); | ||
rmw_ret_t ret = rmw_init_options_init(&options, rcutils_get_default_allocator()); | ||
EXPECT_NE(RMW_RET_OK, rmw_shutdown(&context)); |
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.
Is there a specific error you are expecting? RMW_RET_ERROR?
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.
Based on rmw_shutdown()
API contract, an RMW implementation could equally return RMW_RET_INVALID_ARGUMENT
or RMW_RET_INCORRECT_IMPLEMENTATION
. That's why I don't check for a specific return code.
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.
Since you already test the incorrect_implementation error below, wouldn't it make more sense to test the invalid_argument error here? Set context.implementation_identifier to a known valid identifier, then check that it returns invalid argument?
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.
Hmm, problem is that we get into bad initialization territory. Here I deal with the four (4) known context states: zero-initialized, initialized and not shutdown, initialized and shutdown, finalized. Anything outside that is UB. So if we want to be specific about the assertion we have to further clarify API contracts i.e. what it means for a context to be invalid for all rmw
implementations.
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.
If you don't want to monkey around with the identifier field to force the invalid argument, then I recommend to at least narrow down the expect statement to only two errors you expect. Something like:
EXPECT_TRUE((ret == RMW_RET_INVALID_ARGUMENT ||
ret == RMW_RET_INCORRECT_IMPLEMENTATION))
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.
Sure, we can do that. See 9bff7f2
Signed-off-by: Michel Hidalgo <michel@ekumenlabs.com>
Signed-off-by: Michel Hidalgo <michel@ekumenlabs.com>
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.
lgtm
Hold on, as tests may change with ros2/rmw#243. |
Signed-off-by: Michel Hidalgo <michel@ekumenlabs.com>
Ready for re-review ! |
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.
LGTM
Does this need to be retested on CI? |
More than that. RMW implementations have to be released into |
Same as in #108, I'll hold off merging this until I've re-released RMWs so that rolling PR jobs pass. |
@ros-pull-request-builder retest this please |
Alright, all's green! Going in. |
Signed-off-by: Michel Hidalgo <michel@ekumenlabs.com>
Signed-off-by: Michel Hidalgo <michel@ekumenlabs.com>
Precisely what the title says. Follow-up after #106.