From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 79978C4332F for ; Tue, 8 Nov 2022 17:09:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234256AbiKHRJP (ORCPT ); Tue, 8 Nov 2022 12:09:15 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:45252 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S234164AbiKHRJK (ORCPT ); Tue, 8 Nov 2022 12:09:10 -0500 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7AF4454B06 for ; Tue, 8 Nov 2022 09:08:08 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1667927287; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding; bh=Dd037o29bpcihfgqtjZoF+DSwDz8ncx6U1ijrgWEaU4=; b=RiQQ/p02CEYw4ok+IkkiTSs+MP2yb9/prn2gRvmr0awjHemI0hv/HSPhxgDlZiagfd7LZ1 WBSrVnZcJsE3mqELQRGcVwPNq+/5TIAWyGBnQxXykUEf5UmkOEPsBuTwWMlXYZaBzBV1MG s7FJKYA5pLLHbV3m/+rruBFAHEWn1bQ= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-102-WuBb2cXMMRm2nnfn9ULGsQ-1; Tue, 08 Nov 2022 12:08:03 -0500 X-MC-Unique: WuBb2cXMMRm2nnfn9ULGsQ-1 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.rdu2.redhat.com [10.11.54.8]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 58CE81012460; Tue, 8 Nov 2022 17:08:02 +0000 (UTC) Received: from eperezma.remote.csb (unknown [10.39.193.47]) by smtp.corp.redhat.com (Postfix) with ESMTP id 5F22DC15BB5; Tue, 8 Nov 2022 17:07:58 +0000 (UTC) From: =?UTF-8?q?Eugenio=20P=C3=A9rez?= To: qemu-devel@nongnu.org Cc: Parav Pandit , Stefan Hajnoczi , Si-Wei Liu , Laurent Vivier , Harpreet Singh Anand , "Michael S. Tsirkin" , Gautam Dawar , Liuxiangdong , Stefano Garzarella , Jason Wang , Cindy Lu , Eli Cohen , Cornelia Huck , Zhu Lingshan , kvm@vger.kernel.org, "Gonglei (Arei)" , Paolo Bonzini Subject: [PATCH v6 00/10] ASID support in vhost-vdpa net Date: Tue, 8 Nov 2022 18:07:45 +0100 Message-Id: <20221108170755.92768-1-eperezma@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable X-Scanned-By: MIMEDefang 3.1 on 10.11.54.8 Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Control VQ is the way net devices use to send changes to the device state, = like=0D the number of active queues or its mac address.=0D =0D QEMU needs to intercept this queue so it can track these changes and is abl= e to=0D migrate the device. It can do it from 1576dbb5bbc4 ("vdpa: Add x-svq to=0D NetdevVhostVDPAOptions"). However, to enable x-svq implies to shadow all Vi= rtIO=0D device's virtqueues, which will damage performance.=0D =0D This series adds address space isolation, so the device and the guest=0D communicate directly with them (passthrough) and CVQ communication is split= in=0D two: The guest communicates with QEMU and QEMU forwards the commands to the= =0D device.=0D =0D Comments are welcome. Thanks!=0D =0D v6:=0D - Do not allocate SVQ resources like file descriptors if SVQ cannot be used= .=0D - Disable shadow CVQ if the device does not support it because of net=0D features.=0D =0D v5:=0D - Move vring state in vhost_vdpa_get_vring_group instead of using a=0D parameter.=0D - Rename VHOST_VDPA_NET_CVQ_PASSTHROUGH to VHOST_VDPA_NET_DATA_ASID=0D =0D v4:=0D - Rebased on last CVQ start series, that allocated CVQ cmd bufs at load=0D - Squash vhost_vdpa_cvq_group_is_independent.=0D - Do not check for cvq index on vhost_vdpa_net_prepare, we only have one=0D that callback registered in that NetClientInfo.=0D - Add comment specifying behavior if device does not support _F_ASID=0D - Update headers to a later Linux commit to not to remove SETUP_RNG_SEED=0D =0D v3:=0D - Do not return an error but just print a warning if vdpa device initializa= tion=0D returns failure while getting AS num of VQ groups=0D - Delete extra newline=0D =0D v2:=0D - Much as commented on series [1], handle vhost_net backend through=0D NetClientInfo callbacks instead of directly.=0D - Fix not freeing SVQ properly when device does not support CVQ=0D - Add BIT_ULL missed checking device's backend feature for _F_ASID.=0D =0D Eugenio P=C3=A9rez (10):=0D vdpa: Use v->shadow_vqs_enabled in vhost_vdpa_svqs_start & stop=0D vhost: set SVQ device call handler at SVQ start=0D vhost: Allocate SVQ device file descriptors at device start=0D vdpa: add vhost_vdpa_net_valid_svq_features=0D vdpa: move SVQ vring features check to net/=0D vdpa: Allocate SVQ unconditionally=0D vdpa: Add asid parameter to vhost_vdpa_dma_map/unmap=0D vdpa: Store x-svq parameter in VhostVDPAState=0D vdpa: Add listener_shadow_vq to vhost_vdpa=0D vdpa: Always start CVQ in SVQ mode=0D =0D include/hw/virtio/vhost-vdpa.h | 10 +-=0D hw/virtio/vhost-shadow-virtqueue.c | 35 +-----=0D hw/virtio/vhost-vdpa.c | 114 ++++++++++---------=0D net/vhost-vdpa.c | 171 ++++++++++++++++++++++++++---=0D hw/virtio/trace-events | 4 +-=0D 5 files changed, 222 insertions(+), 112 deletions(-)=0D =0D -- =0D 2.31.1=0D =0D