Quoting Bruce Perens (bruce@???):

> I've been getting credible reports that grsecurity.net is infringing the
> kernel by preventing customers from redistributing the GPL code for their
> patch.

They would be infringing the copyright on the Linux kernel if it were
establishable as a judiciable fact that their patchsets fail the fair
use tests as derivative works -- but that is not clearly the case.
(At least in US law, fair use is an affirmative defence, so you find out
whether it's applicable to a particular case only when it's decided.)

Back when qmail, djbdns, etc. were still proprietary code, I heard legal
commentary on the wide distribution of patchsets, and recall hearing
lawyerly opinion that such patchsets fairly easily satisfy the test of
being commentary on the copyrighted work
(http://fairuse.stanford.edu/overview/fair-use/what-is-fair-use/) as a
'transformative' purpose. So, for example the consensus was that Russ
Nelson & friends' 'netqmail' patchset did not violate DJB's copyright in
qmail.

> What is it we want Google to do? They usually listen when I ask...

Brad Spengler seemeed to want Google to cease using old grsecurity
releases on old kernel codebases (for, I think, Chrome OS) and still
letting the world think that the results reflected on the quality of
grsecurity. Their complaint was that this, in their view, damaged their
brand. (A more cynical take would be that they just wanted to get
paid.)

This much water having gone over the dam, I personally doubt asking
Google to do/not do something at this late date would mend things.