Skip to content

Conversation

@linuswagner
Copy link
Contributor

See #83 .
We are currently moving in the other extreme, because there are other schemes that we might be able to remove safely.

Looking at my project, I see the following override relations:

<"cpp+functionSet","cpp+method">,
<"cpp+destructor","cpp+destructor">,
<"cpp+destructor","cpp+functionSpecialization">,
<"cpp+method","cpp+method">,
<"cpp+method","cpp+functionSpecialization">,
<"cpp+functionSpecialization","cpp+destructor">,
<"cpp+functionSpecialization","cpp+method">,
<"cpp+functionSpecialization","cpp+functionSpecialization">,
<"cpp+new","cpp+functionSpecialization">,
<"cpp+new","cpp+constructor">,
<"cpp+new","cpp+functionInstance">,
<"cpp+new","problem">

(left is overwritten by right).

Without too much research, it looks like most of them should stay in the call graph even after expanding it.
The first one might be removable (see here, functionSet seems to be an intermediate representation).

@jurgenvinju also pointed out that we might want to look at individual virtual methods and filter out the ones that do not have an implementation.

@linuswagner linuswagner linked an issue Feb 20, 2024 that may be closed by this pull request
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Underapproximation in closeOverriddenVirtualCalls

1 participant