From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Layton Subject: Re: btime and change_attr features in kclient Date: Mon, 24 Oct 2016 12:30:33 -0400 Message-ID: <1477326633.14828.13.camel@redhat.com> References: <1477325965.14828.8.camel@redhat.com> <1477326343.14828.11.camel@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8bit Return-path: Received: from mail-qt0-f170.google.com ([209.85.216.170]:35580 "EHLO mail-qt0-f170.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S941334AbcJXQah (ORCPT ); Mon, 24 Oct 2016 12:30:37 -0400 Received: by mail-qt0-f170.google.com with SMTP id f6so134199686qtd.2 for ; Mon, 24 Oct 2016 09:30:36 -0700 (PDT) In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Sage Weil Cc: zyan@redhat.com, ceph-devel@vger.kernel.org On Mon, 2016-10-24 at 16:28 +0000, Sage Weil wrote: > On Mon, 24 Oct 2016, Jeff Layton wrote: > > On Mon, 2016-10-24 at 12:19 -0400, Jeff Layton wrote: > > > On Mon, 2016-10-24 at 16:16 +0000, Sage Weil wrote: > > > > > > > > Hi, > > > > > > > > This PR that changes the entity_addr encoding is finally ready to merge > > > > > > > >         https://github.com/ceph/ceph/pull/9825 > > > > > > > > but it defines an MSG_ADDR2 feature bit indicating whether we support the > > > > new addr encoding.  Right now it's defined with the same value as the > > > > FS_CHANGE_ATTR and FS_BTIME feature bit.  Are those upstream yet?  If not, > > > > we should get the addr2 support in at the same time so that we can > > > > conserve a feature bit.  If not, we'll have to consume another bit.  :( > > > > > > > > sage > > > > > > > > > > That support is merged into master -- is that what you mean by > > > "upstream"? > > > > > > > Sorry, I misread the subject... > > > > No, those are not yet supported or recognized in kclient yet. In which > > kernel are you planning to merge the addr2 patches? > > That's good news!  :)  The kclient patches for addr2 haven't been written > yet, though.  There is no particular rush, but we should plan to merge > all three together so that we can use the same feature bit. > > Thanks! > sage Makes sense to me and I agree that there's no rush: The kernel doesn't care about the btime. I think we'll probably just discard it until we have a way to expose it to userland. The change_attr could be useful when exporting via knfsd, but we've lived without it so far. -- Jeff Layton