On Fri, 2021-01-15 at 16:58 +0000, [email protected] wrote:
> On Fri, Jan 15, 2021 at 04:48:13PM +0000, Richard Purdie wrote:
> > On Fri, 2021-01-15 at 12:48 +0200, Mikko Rapeli wrote:
> > > https://nvd.nist.gov/vuln/detail/CVE-2018-13410 is disputed and
> > > also Debian considers it not a vulnerability:
> > > 
> > > https://security-tracker.debian.org/tracker/CVE-2018-13410
> > > 
> > > http://seclists.org/fulldisclosure/2018/Jul/24
> > > "Negligible security impact, would involve that a untrusted party 
> > > controls the -TT value."
> > > 
> > > https://nvd.nist.gov/vuln/detail/CVE-2018-13684 is not for zip, also 
> > > Debian concludes this:
> > > 
> > > https://security-tracker.debian.org/tracker/CVE-2018-13684
> > > 
> > > "NOT-FOR-US: smart contract implementation for ZIP"
> > > 
> > > Signed-off-by: Mikko Rapeli <[email protected]>
> > > ---
> > >  meta/recipes-extended/zip/zip_3.0.bb | 3 +++
> > >  1 file changed, 3 insertions(+)
> > > 
> > > diff --git a/meta/recipes-extended/zip/zip_3.0.bb 
> > > b/meta/recipes-extended/zip/zip_3.0.bb
> > > index c00a932763..47e6fc5278 100644
> > > --- a/meta/recipes-extended/zip/zip_3.0.bb
> > > +++ b/meta/recipes-extended/zip/zip_3.0.bb
> > > @@ -19,6 +19,9 @@ UPSTREAM_VERSION_UNKNOWN = "1"
> > >  SRC_URI[md5sum] = "7b74551e63f8ee6aab6fbc86676c0d37"
> > >  SRC_URI[sha256sum] = 
> > > "f0e8bb1f9b7eb0b01285495a2699df3a4b766784c1765a8f1aeedf63c0806369"
> > > 
> > > 
> > > +CVE_CHECK_WHITELIST += "CVE-2018-13410"
> > > +CVE_CHECK_WHITELIST += "CVE-2018-13684"
> > > +
> > 
> > Where we're adding these can we put a small comment in as well just
> > saying why we're whitelisting it?
> 
> Sure, but lets try to be consistent then.

I am trying to be, I did add comments with the recently whitelist
entries I added! :)

There are probably some older ones which don't have it but we should
probably try and improve that over time if we can too.
> 

> > I appreciate the info is in the commit but I think its important enough
> > to list in the recipe as a comment too.
> 
> I agree.
> 
> Though it's not clear if these should be removed when updating recipe 
> versions.
> Sometimes NVD data is good, sometimes bad, sometimes it's not clear howto 
> fix...
> Doing local builds with
> 
> INHERIT += "cve-check"
> 
> is the only way to find out, I guess.

Yes, if upstream are no longer listing it as being relevant, we could
drop the WHITELIST entry but in general I think they'll be in cases
where the upstream entry can't be changed. That means for recipe
upgrading, we probably end up keeping them.

Cheers,

Richard

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#146841): 
https://lists.openembedded.org/g/openembedded-core/message/146841
Mute This Topic: https://lists.openembedded.org/mt/79698852/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub 
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to