From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx1.redhat.com (mx1.redhat.com [209.132.183.28]) by dpdk.org (Postfix) with ESMTP id 0ACDD968 for ; Thu, 31 Aug 2017 11:50:59 +0200 (CEST) Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 30F657E43F; Thu, 31 Aug 2017 09:50:58 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 30F657E43F Authentication-Results: ext-mx03.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com Authentication-Results: ext-mx03.extmail.prod.ext.phx2.redhat.com; spf=fail smtp.mailfrom=maxime.coquelin@redhat.com Received: from localhost.localdomain (ovpn-112-32.ams2.redhat.com [10.36.112.32]) by smtp.corp.redhat.com (Postfix) with ESMTP id 02201841C1; Thu, 31 Aug 2017 09:50:46 +0000 (UTC) From: Maxime Coquelin To: dev@dpdk.org, yliu@fridaylinux.org, jfreiman@redhat.com, tiwei.bie@intel.com Cc: mst@redhat.com, vkaplans@redhat.com, jasowang@redhat.com, Maxime Coquelin Date: Thu, 31 Aug 2017 11:50:02 +0200 Message-Id: <20170831095023.21037-1-maxime.coquelin@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.12 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.27]); Thu, 31 Aug 2017 09:50:58 +0000 (UTC) Subject: [dpdk-dev] [PATCH 00/21] Vhost-user: Implement device IOTLB support X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 31 Aug 2017 09:50:59 -0000 This first non-RFC, which targets v17.11, adds support for VIRTIO_F_IOMMU_PLATFORM feature, by implementing device IOTLB in the vhost-user backend. It improves the guest safety by enabling the possibility to isolate the Virtio device. It makes possible to use Virtio PMD in guest with using VFIO driver without enable_unsafe_noiommu_mode parameter set, so that the DPDK application on guest can only access memory its has been allowed to, and preventing malicious/buggy DPDK application in guest to make vhost-user backend write random guest memory. Note that Virtio-net Kernel driver also support IOMMU. The series depends on Qemu's "vhost-user: Specify and implement device IOTLB support" [0], available upstream and which will be part of Qemu v2.10 release. Performance-wise, even if this RFC has still room for optimizations, no performance degradation is noticed with static mappings (i.e. DPDK on guest) with PVP benchmark: Traffic Generator: Moongen (lua-trafficgen) Acceptable Loss: 0.005% Validation run time: 1 min Guest DPDK version/commit: v17.05 QEMU version/commit: master (6db174aed1fd) Virtio features: default CPU: Intel(R) Xeon(R) CPU E5-2667 v4 @ 3.20GHz NIC: 2 x X710 Page size: 1G host/1G guest Results (bidirectional, total of the two flows): - base: 18.8Mpps - base + IOTLB series, IOMMU OFF: 18.8Mpps - base + IOTLB series, IOMMU ON: 18.8Mpps (14.5Mpps w/o PATCH 21/21) This is explained because IOTLB misses, which are very costly, only happen at startup time. Indeed, once used, the buffers are not invalidated, so if the IOTLB cache is large enough, there will be only cache hits. Also, the use of 1G huge pages improves the IOTLB cache searching time by reducing the number of entries. With 2M hugepages, a performance degradation is seen with IOMMU on: Traffic Generator: Moongen (lua-trafficgen) Acceptable Loss: 0.005% Validation run time: 1 min Guest DPDK version/commit: v17.05 QEMU version/commit: master (6db174aed1fd) Virtio features: default CPU: Intel(R) Xeon(R) CPU E5-2667 v4 @ 3.20GHz NIC: 2 x X710 Page size: 2M host/2M guest Results (bidirectional, total of the two flows): - base: 18.8Mpps - base + IOTLB series, IOMMU OFF: 18.8Mpps - base + IOTLB series, IOMMU ON: 13.5Mpps (12.4Mpps wo PATCH 21/21) A possible improvement would be to merge contiguous IOTLB entries sharing the same permissions. A very rough patch implementing this idea fixes the performance degradation (18.8Mpps), but the required work to clean it would delay this series after v17.11. With dynamic mappings (i.e. Virtio-net kernel driver), this is another story. The performance is so poor it makes it almost unusable. Indeed, since the Kernel driver unmaps the buffers as soon as they are handled, almost all descriptors buffers addresses translations result in an IOTLB miss. There is not much that can be done on DPDK side, except maybe batching IOTLB miss requests no to break bursts, but it would require a big rework. In Qemu, we may consider enabling IOMMU MAP notifications, so that DPDK receives the IOTLB updates without having to send IOTLB miss request. Regarding the design choices: - I initially intended to use userspace RCU library[1] for the cache implementation, but it would have added an external dependency, and the lib is not available in all distros. Qemu for example got rid of this dependency by copying some of the userspace RCU lib parts into Qemu tree, but this is not possible with DPDK due to licensing issues (RCU lib is LGPL v2). Thanks to Jason advice, I implemented the cache using rd/wr locks. - I initially implemented a per-device IOTLB cache, but the concurrent acccesses on the IOTLB lock had huge impact on performance (~-40% in bidirectionnal, expect even worse with multiqueue). I move to a per- virtqueue IOTLB design, which prevents this concurrency. - The slave IOTLB miss request supports reply-ack feature in spec, but this version doesn't block or busy-wait for the corresponding update so that other queues sharing the same lcore can be processed in the meantime. For those who would like to test the series, I made it available on gitlab[2] (vhost_user_iotlb_v1 tag). The guest kernel command line requires the intel_iommu=on parameter, and the guest should be started with and iommu device attached to the virtio-net device. For example: ./qemu-system-x86_64 \ -enable-kvm -m 4096 -smp 2 \ -M q35,kernel-irqchip=split \ -cpu host \ -device intel-iommu,device-iotlb=on,intremap \ -device ioh3420,id=root.1,chassis=1 \ -chardev socket,id=char0,path=/tmp/vhost-user1 \ -netdev type=vhost-user,id=hn2,chardev=char0 \ -device virtio-net-pci,netdev=hn2,id=v0,mq=off,mac=$MAC,bus=root.1,disable-modern=off,disable-legacy=on,iommu_platform=on,ats=on \ ... [0]: https://lists.gnu.org/archive/html/qemu-devel/2017-06/msg00520.html [1]: http://liburcu.org/ [2]: https://gitlab.com/mcoquelin/dpdk-next-virtio/commits/vhost_user_iotlb_v1 Changes since RFC: ================== - Fix memory leak in error patch reported by Jens - Rework wait for IOTLB update by stopping the burst to let other queues to be processed, if any. It implies the introduction an iotlb_pending_list, so that iotlb miss requests aren't sent multiple times for a same address. - Optimize iotlb lock usage to recover to same as IOMMU off performance - Fix device locking issue in rte_vhost_dequeue_burst() error path - Change virtio_dev_rx error handling for consistency with mergeable rx, and to ease returning in case of IOTLB misses. - Fix checkpatch warnings reported by checkpatch@dpdk.org Maxime Coquelin (21): Revert "vhost: workaround MQ fails to startup" vhost: make error handling consistent in rx path vhost: protect virtio_net device struct vhost: prepare send_vhost_message() to slave requests vhost: add support to slave requests channel vhost: declare missing IOMMU-related definitions for old kernels vhost: add iotlb helper functions vhost: iotlb: add pending miss request list and helpers vhost-user: add support to IOTLB miss slave requests vhost: initialize vrings IOTLB caches vhost-user: handle IOTLB update and invalidate requests vhost: introduce guest IOVA to backend VA helper vhost: use the guest IOVA to host VA helper vhost: enable rings at the right time vhost: don't dereference invalid dev pointer after its reallocation vhost: postpone rings addresses translation vhost-user: translate ring addresses when IOMMU enabled vhost-user: iommu: postpone device creation until ring are mapped vhost: iommu: Invalidate vring in case of matching IOTLB invalidate vhost: enable IOMMU support vhost: iotlb: reduce iotlb read lock usage lib/librte_vhost/Makefile | 4 +- lib/librte_vhost/iotlb.c | 315 +++++++++++++++++++++++++++++++++++ lib/librte_vhost/iotlb.h | 64 +++++++ lib/librte_vhost/vhost.c | 340 ++++++++++++++++++++++++++++++++------ lib/librte_vhost/vhost.h | 53 +++++- lib/librte_vhost/vhost_user.c | 376 ++++++++++++++++++++++++++++++++---------- lib/librte_vhost/vhost_user.h | 20 ++- lib/librte_vhost/virtio_net.c | 129 +++++++++++---- 8 files changed, 1130 insertions(+), 171 deletions(-) create mode 100644 lib/librte_vhost/iotlb.c create mode 100644 lib/librte_vhost/iotlb.h -- 2.13.3