From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:41519) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1eS1kM-0002Lh-Dt for qemu-devel@nongnu.org; Thu, 21 Dec 2017 09:22:55 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1eS1kH-0001Qo-T4 for qemu-devel@nongnu.org; Thu, 21 Dec 2017 09:22:54 -0500 Received: from mx1.redhat.com ([209.132.183.28]:49110) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1eS1kH-0001Q4-Na for qemu-devel@nongnu.org; Thu, 21 Dec 2017 09:22:49 -0500 Date: Thu, 21 Dec 2017 16:22:44 +0200 From: "Michael S. Tsirkin" Message-ID: <20171221162200-mutt-send-email-mst@kernel.org> References: <20171217125457.3429-1-marcel@redhat.com> <20171219194951-mutt-send-email-mst@kernel.org> <77710384-30d0-cb76-9dbc-6195b831f050@redhat.com> <20171220200550-mutt-send-email-mst@kernel.org> <20171221072750.GC2931@yuvallap> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20171221072750.GC2931@yuvallap> Subject: Re: [Qemu-devel] [PATCH V2 0/5] hw/pvrdma: PVRDMA device implementation List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Yuval Shaia Cc: Marcel Apfelbaum , qemu-devel@nongnu.org, ehabkost@redhat.com, imammedo@redhat.com, pbonzini@redhat.com On Thu, Dec 21, 2017 at 09:27:51AM +0200, Yuval Shaia wrote: > > > > > > > What happens if guest attempts to register all its memory? > > > > > > > > > > Then we loose, is not different from bare metal, reg_mr will pin all the RAM. > > > > We need to find a way to communicate to guests about amount > > of memory they can pin. > > dev_caps.max_mr_size is the way device limits guest driver. > This value is controlled by the command line argument dev-caps-max-mr-size > so we should be fine (btw, default value is 1<<32). Isn't that still leaving the option for guest to register all memory, just in chunks? > > > > > However this is only one scenario, and hopefully not much used > > > for RoCE. (I know IPoIB does that, but it doesn't make sense to use it with RoCE). > > > > SRP does it too AFAIK. > >