Re: [RFC] Are you good with Lockdep?

[Byungchul Park]

On Thu, Nov 12, 2020 at 03:15:32PM +0900, Byungchul Park wrote:
> > If on the other hand there's some bug in lockdep itself that causes 
> > excessive false positives, it's better to limit the number of reports 
> > to one per bootup, so that it's not seen as a nuisance debugging 
> > facility.
> > 
> > Or if lockdep gets extended that causes multiple previously unreported 
> > (but very much real) bugs to be reported, it's *still* better to 
> > handle them one by one: because lockdep doesn't know whether it's real 
> 
> Why do you think we cannot handle them one by one with multi-reporting?
> We can handle them with the first one as we do with single-reporting.
> And also that's how we work, for example, when building the kernel or
> somethinig.

Let me add a little bit more. I just said the fact that we are able to
handle the bugs one by one as if we do with single-reporting.

But the thing is multi-reporting could be more useful in some cases.
More precisely speaking, bugs not caused by IRQ state will be reported
without annoying nuisance. I bet you have experienced a ton of nuisances
when multi-reporting Lockdep detected a deadlock by IRQ state.

For some cases, multi-reporting is as useful as single-reporting, while
for the other cases, multi-reporting is more useful. Then I think we
have to go with mutil-reporting if there's no technical issue.

Thanks,
Byungchul