Re: [RFC PATCH-for-5.2 4/5] qom: Let ObjectPropertyGet functions return a boolean value

[Peter Maydell]

On Mon, 7 Sep 2020 at 15:26, Markus Armbruster <armbru@redhat.com> wrote:
> Making more functions taking Error ** return bool (commit
> b6d7e9b66f..a43770df5d) created a number of false positive
> CHECKED_RETURN Coverity issues.  Peter notified me back then.  This
> series could help (we'd have to double-check), but it covers only the
> getters, not the setters.
>
> Peter, are you still interested in reducing the false positives (as
> opposed to marking them)?

I don't have a strong view on whether we should make the change
in this series, though I do think overall we should have a
decision on whether we like (Error** + void) or (Error** + bool)
and try to aim to eventually converge on being consistent about it.

As far as the Coverity issues are concerned, I don't think we
should let them drive our technical decisions. We should
decide what we want to do with our APIs on other grounds,
and then either that causes the issues to go away or we
mark them as false-positives. I would ideally like us to
do one or the other relatively soon simply because there are
a lot of these issues currently piled up in the Coverity UI
and it makes it harder to see the interesting issues.

thanks
-- PMM