From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:38600) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cWikB-0000Hf-BZ for qemu-devel@nongnu.org; Thu, 26 Jan 2017 07:01:36 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cWik7-00017P-BF for qemu-devel@nongnu.org; Thu, 26 Jan 2017 07:01:35 -0500 Received: from mx1.redhat.com ([209.132.183.28]:56300) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1cWik7-00017C-5V for qemu-devel@nongnu.org; Thu, 26 Jan 2017 07:01:31 -0500 Date: Thu, 26 Jan 2017 20:01:22 +0800 From: Peter Xu Message-ID: <20170126120122.GL5151@pxdev.xzpeter.org> References: <1485253571-19058-1-git-send-email-peterx@redhat.com> <1485253571-19058-3-git-send-email-peterx@redhat.com> <20170124092905.41832531@t450s.home> <9ef03816-0bef-f54b-63fe-daf27eab4a40@redhat.com> <20170126071244.GJ5151@pxdev.xzpeter.org> <11032fa7-1e5f-dac6-f0d8-56b41121361b@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <11032fa7-1e5f-dac6-f0d8-56b41121361b@redhat.com> Subject: Re: [Qemu-devel] [PATCH v5 02/18] vfio: introduce vfio_get_vaddr() List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini Cc: Alex Williamson , tianyu.lan@intel.com, kevin.tian@intel.com, mst@redhat.com, jan.kiszka@siemens.com, jasowang@redhat.com, qemu-devel@nongnu.org, bd.aviv@gmail.com On Thu, Jan 26, 2017 at 11:55:22AM +0100, Paolo Bonzini wrote: > > > On 26/01/2017 08:12, Peter Xu wrote: > > > > /* > > * Here, we need to have the lock not only for vfio_get_vaddr(), > > * but also needs to make sure that the vaddr will be valid for > > * further operations. > > * > > * When we map new pages, we need the lock to make sure that vaddr > > * is valid along the way we build up the IO page table (via > > * vfio_dma_map()). Then, as long as the mapping is set up, we can > > * unlock since those pages will be pinned in kernel (which makes > > * sure that the RAM backend of vaddr will always be there, even > > * if the memory object is destroyed and RAM released). > > * > > * For unmapping case, we don't really need the protection since > > * the pages are in all cases locked in kernel, so we'll probably > > * be safe even without the lock. However, it won't hurt we have > > * the lock as well here. > > */ > > Even simpler, just before the definition of vfio_get_vaddr: > > /* Called with rcu_read_lock held. */ > > and just before the vfio_dma_map call: > > /* vaddr is only valid until rcu_read_unlock(). But after > * vfio_dma_map has set up the mapping the pages will be pinned > * by the kernel. This makes sure that the RAM backend of vaddr > * will always be there, even if the memory object is destroyed > * and its backing memory munmap-ed. > */ > > I'm not sure that you can get rid of the lock for the unmapping case. > Better remove that part of the comment. Sure. Let me take yours. Thanks! -- peterx