Re: [PATCH 2/2] Documentation: security-bugs.rst: clarify CVE handling

[Kees Cook <keescook@chromium.org>]

On Mon, Jul 03, 2023 at 09:05:15PM +0200, Greg Kroah-Hartman wrote:
> On Mon, Jul 03, 2023 at 11:35:37AM -0700, Kees Cook wrote:
> > On Mon, Jul 03, 2023 at 05:00:15PM +0200, Greg Kroah-Hartman wrote:
> > > On Mon, Jul 03, 2023 at 06:08:00AM +0200, Willy Tarreau wrote:
> > > >   The security team does not assign CVEs, nor do we require them for
> > > >   reports or fixes, as this can needlessly complicate the process and may
> > > >   delay the bug handling.  If a reporter wishes to have a CVE identifier
> > > >   assigned, they should find one by themselves, for example by contacting
> > > >   MITRE directly.  However under no circumstances will a patch inclusion
> > > >   be delayed to wait for a CVE identifier to arrive.
> > > > 
> > > > This puts the responsibility for finding one in time on the reporter
> > > > depending on what they expect, and if they want it in the commit
> > > > message, they'd rather have one before reporting the problem.
> > > 
> > > Oh, nice wording, let me steal that!  :)
> > 
> > Yeah, this is good. The last sentence is a little hard to parse, so how
> > about this, with a little more rationale expansion:
> > 
> > However under no circumstances will patch publication be delayed for
> > CVE identifier assignment. Getting fixes landed takes precedence; the
> > CVE database entry will already reference the commit, so there is no loss
> > of information if the CVE is assigned later.
> 
> "simple is better" should be the key here, reading a wall of text is
> hard for people, so let me just keep the one new sentance that Willy
> proposed and if people still struggle with the whole CVEs and
> security@k.o mess in the future, we can revise it again.
> 
> Also, there is not really a "CVE database", I think that's what NVD from
> NIST does and CNNVD from China does, and "Something to be named in the
> future soon" will do for the EU.  There is a "CVE List" at cve.org, but
> that thing is always out of date, and for all of this I don't want to
> have to try to explain it in our document as that's nothing we want to
> mess with :)

Okay, fair, though please include something about it in the commit
log so that other folks with concerns similar to Mathias Krause's will
be answered:
https://infosec.exchange/@minipli/110632971830936754

I still think this version of the sentence is more readable:

However under no circumstances will patch publication be delayed for
CVE identifier assignment.

"patch inclusion" is less clear to me that "publication", and "be
delayed to wait for" is redundant: a delay is a wait, and "to arrive"
is just the assignment, which is the subject of the paragraph, so better
to keep the language for that consistent.

-Kees

-- 
Kees Cook