Re: apparent bug about check_free_strict

On Tue, Nov 25, 2025 at 10:34:49AM -0500, Oleg Drokin wrote:
> On Tue, 2025-11-25 at 17:04 +0200, Toomas Soome wrote:
> > For positive side, the current smatch has been able to detect many
> > bugs the previous versions had missed - its definitely good progress.
> > While there are some issues, we can disable checks on such cases.
> > Only problem is that where previously —disable=check_free_strict did
> > work, —disable=check_free does not seem to:)
> 
> The way I found to be the most impactful is to just keep a database of
> the warnings and then:
> 1. mark the clearly invalid ones so they don't pop up again
> 2. every time there's a new patch - see if anything new gets added.
> 
> the new additions are the primary focus of new triage then since all
> the old ones were presumably checked already.

Yeah.  This is my favorite approach too.

I feel like you're more in favour of a higher false positive ratio
than I am, but I do think looking at false positives one time only
and then ignoring it forever is the best approach.  It lets you check
for more things, faster.

regards,
dan carpenter