On Wed, Jan 16, 2019 at 04:54:49PM +1200, Linus Torvalds wrote: > On Wed, Jan 16, 2019 at 11:45 AM Dave Chinner <david@xxxxxxxxxxxxx> wrote: > > > > I'm assuming that you can invalidate the page cache reliably by a > > means that does not repeated require probing to detect invalidation > > has occurred. I've mentioned one method in this discussion > > already... > > Yes. And it was made clear to you that it was a bug in xfs dio and > what the right thing to do was. > > And you ignored that, and claimed it was a feature. Linus, either you aren't listening or you're being intentionally provocative. So, for the *third* time this thread: we can probably remove this code but first we need to be sure it doesn't cause unexpected regressions before we commit such a change. We are not cowboys who test userspace behavioural changes on users without review or discussion. Indeed, I wrote a patch to remove the invalidation /several days ago/ and put it into my test trees, and it's been there since. Just because you don't see immediate changes doesn't mean it isn't happening. > Either you care or you don't. If you don't care (and so far everything > you said seems to imply you don't), Linus, this is just a personal attack and IMO a violation of the CoC. It's straight out wrong, insulting, totally unprofessional and completely uncalled for. This is most definitely not a useful technical response to the issues I raised. i.e you cut out all the context of my response about whether "no probing necessary" page cache invalidation attacks are something we need to care about in the future. We don't need you to shout about existing "no probing necessary" page cache invalidation attacks that are already being addressed, we need to determine if it's going to be a recurring problem in future because that directly affects the mitigation strategies we can implement. -Dave. -- Dave Chinner david@xxxxxxxxxxxxx
