From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [RFC 0/4] Virtio uses DMA API for all devices Date: Mon, 6 Aug 2018 23:18:14 -0700 Message-ID: <20180807061814.GC32709__7769.56701088256$1533622580$gmane$org@infradead.org> References: <20180803220443-mutt-send-email-mst@kernel.org> <051fd78e15595b414839fa8f9d445b9f4d7576c6.camel@kernel.crashing.org> <20180805031046-mutt-send-email-mst@kernel.org> <20180806164106-mutt-send-email-mst@kernel.org> <20180806233024-mutt-send-email-mst@kernel.org> <0967fc30001323e6e38ed12c8dba8ee3d1aa13f5.camel@kernel.crashing.org> <20180807002857-mutt-send-email-mst@kernel.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20180807002857-mutt-send-email-mst@kernel.org> 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: robh@kernel.org, srikar@linux.vnet.ibm.com, Benjamin Herrenschmidt , Will Deacon , linux-kernel@vger.kernel.org, linuxram@us.ibm.com, virtualization@lists.linux-foundation.org, Christoph Hellwig , jean-philippe.brucker@arm.com, paulus@samba.org, marc.zyngier@arm.com, mpe@ellerman.id.au, joe@perches.com, robin.murphy@arm.com, david@gibson.dropbear.id.au, linuxppc-dev@lists.ozlabs.org, elfring@users.sourceforge.net, haren@linux.vnet.ibm.com, Anshuman Khandual List-Id: virtualization@lists.linuxfoundation.org On Tue, Aug 07, 2018 at 12:46:34AM +0300, Michael S. Tsirkin wrote: > Well we have the RFC for that - the switch to using DMA ops unconditionally isn't > problematic itself IMHO, for now that RFC is blocked > by its perfromance overhead for now but Christoph says > he's trying to remove that for direct mappings, > so we should hopefully be able to get there in X weeks. The direct calls to dma_direct_ops aren't going to help you with legacy virtio, given that virtio is specified to deal with physical addresses, while dma-direct is not in many cases. It would however help with the case where qemu always sets the platform dma flag, as we'd avoid the indirect calls for that.