From mboxrd@z Thu Jan 1 00:00:00 1970 From: Asias He Subject: Re: [PATCH V4 0/2] tcm_vhost hotplug Date: Mon, 25 Mar 2013 16:20:09 +0800 Message-ID: <20130325082009.GC20435@hj.localdomain> References: <1363930745-7520-1-git-send-email-asias@redhat.com> <20130324152054.GC8657@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20130324152054.GC8657@redhat.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: virtualization-bounces@lists.linux-foundation.org Errors-To: virtualization-bounces@lists.linux-foundation.org To: "Michael S. Tsirkin" Cc: kvm@vger.kernel.org, virtualization@lists.linux-foundation.org, target-devel@vger.kernel.org, Stefan Hajnoczi , Paolo Bonzini List-Id: virtualization@lists.linuxfoundation.org On Sun, Mar 24, 2013 at 05:20:54PM +0200, Michael S. Tsirkin wrote: > On Fri, Mar 22, 2013 at 01:39:03PM +0800, Asias He wrote: > > Asias He (2): > > tcm_vhost: Introduce tcm_vhost_check_feature() > > tcm_vhost: Add hotplug/hotunplug support > > So this work should stay out of tree until we have > 2 users for vhost-scsi, but I sent some comment to > help you make progress. Actually, I do not think the real reason you want to blcok/delay this is that there is only one user. I believe the real reason is simply that qemu is not using it. If there were already 2 users other than qemu (qemu is still not using it), what would you say, you want 3 users of this? Would you say the same 2 user thing, if a feature is only used by qemu and other users are not using it? Also, can you comment on Paolo's comment in the other thread. """ Il 19/03/2013 14:45, Michael S. Tsirkin ha scritto: > On Tue, Mar 19, 2013 at 10:36:42AM +0100, Paolo Bonzini wrote: >> Il 18/03/2013 22:53, Michael S. Tsirkin ha scritto: >>> Sorry, no, I'd prefer we get userspace support in qemu in first. >>> If there's only a single user for this driver (kvmtool), >>> then it was a mistake to merge it, the right thing would be to >>> freeze it >>> and look at whether we can drop it completely. >> >> I'm still not sure why this matters for this patch, since it does not >> change the userspace ABI. > > It enables a new feature bit. How does that matter if userspace is not supposed to do anything special with that bit? Paolo """ > > drivers/vhost/tcm_vhost.c | 224 ++++++++++++++++++++++++++++++++++++++++++++-- > > drivers/vhost/tcm_vhost.h | 10 +++ > > 2 files changed, 229 insertions(+), 5 deletions(-) > > > > -- > > 1.8.1.4 -- Asias