Bugzilla, lightning talks, and future summits [LWN.net]

Bugzilla, lightning talks, and future summits

Posted Oct 30, 2013 12:48 UTC (Wed) by johill (subscriber, #25196)
Parent article: Bugzilla, lightning talks, and future summits

FWIW, after taking a look, it seems that wireless is leading the list because coverity is getting confused by 'enum nl80211_band' vs. 'enum ieee80211_band' and flagging every single use of any of those ...

to post comments

Bugzilla, lightning talks, and future summits

Posted Oct 30, 2013 16:59 UTC (Wed) by blackwood (guest, #44174) [Link]

Yes, it seems to not be terribly clever with handling enums in general. In drm/i915 there's lots of reports about enum mismatches despite that we have explicit casts for all (or at least most) of these cases. Often with a comment that on a given hw generation pipes and transcoders are wired up in a fixed way, so casting is the Right Thing. Still coverity doesn't seem to get it. I'll plan to file those all as false positives and hope they improve the tool a bit.