On Wed, 31 May 2017, Igor Stoppa wrote: > On 30/05/17 13:32, James Morris wrote: > > > This seems like pointless churn in security-critical code in anticipation > > of features which are still in development and may not be adopted. > > > > Is there a compelling reason to merge this now? (And I don't mean worrying > > about non-existent compliers). > > I propose to take this patch as part of those I will be submitting. > It took me some unplanned time to add support for hardened user copy, > but now it's done - at least to a point that I can test it without failures. > > So I'm back on track to provide an example of the smalloc api and I can > also use Tetsuo's work (thanks again, btw). > This patch would be sandwiched between the smalloc ones and the LSM rework. > > It can get merged when the rest (hopefully) is merged. > > But I have a more prosaic question: since smalloc is affecting the > memory subsystem, can it still be merged through the security tree?
It needs acks from the maintainers of the affected subsystems. -- James Morris <jmor...@namei.org>