From nobody Thu Apr 2 06:10:57 2026 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=quarantine dis=none) header.from=redhat.com ARC-Seal: i=1; a=rsa-sha256; t=1774873828; cv=none; d=zohomail.com; s=zohoarc; b=hitE7OC5H0pk/IUy3AW+dD4cTjYZU7Z5RccoxyKVbxoBR9aWYxc2JGqDiUlygt5y8LHD4cwg4dnd7FEpGfQ1M1XzFK1a+pA5BI3cG8+YAwnMXEWKLLbuFvoK168q9i+w5cwccEsL26/IcHCLKXaLlCaOP6Nlb26TTYD9DF0intY= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1774873828; h=Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=d09OnqQttl2o/9T3TEgKhyfsEQgk4MnQ9HxKJ0C8/iM=; b=ko8WdrjRYrhxDRVXiwBvEEOAkSKMl+4VfCs+G/YkDsipHpzXlSlS+MGkICJMeT7N8xX1hL/shgHNFSOhrPiTsAMTS4NdY0nITVeJbcBC3Oqq4lr1AWBjP1L6Cs/6N8/3YW+bKYcEo4WMGTYdB370f2zfDSyjmt2Rfla6zi+hQ7M= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=quarantine dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 177487382898558.04342793493606; Mon, 30 Mar 2026 05:30:28 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1w7BkR-0000ZJ-OK; Mon, 30 Mar 2026 08:29:41 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1w7BkK-0000Yf-38 for qemu-devel@nongnu.org; Mon, 30 Mar 2026 08:29:33 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.129.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1w7BkF-0004kj-VC for qemu-devel@nongnu.org; Mon, 30 Mar 2026 08:29:31 -0400 Received: from mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-323-l7qTyIGDMyiU6Tx9SP-iaA-1; Mon, 30 Mar 2026 08:29:23 -0400 Received: from mx-prod-int-01.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-01.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.4]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id CC0091955D8B for ; Mon, 30 Mar 2026 12:29:22 +0000 (UTC) Received: from kdutta-thinkpadp1gen7.bengluru.csb (unknown [10.74.64.235]) by mx-prod-int-01.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 055A130001A2; Mon, 30 Mar 2026 12:29:19 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1774873764; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding; bh=d09OnqQttl2o/9T3TEgKhyfsEQgk4MnQ9HxKJ0C8/iM=; b=ZZYrz9xw/qsponSOwg42+KQtcW2geE7TPPRQEQURNah7EDySiMnMJTBEPhyLN4ZxpxIMy7 jW9xU3j2pqj4YY5HE2qXzjYvFV2WnBRIJkmZ+rQKeyA3NnIhwsBjOQFvfJqtAfMoPd5lMS dKwWwyB7L8pr6yQynp6xCxIqzeIr/QA= X-MC-Unique: l7qTyIGDMyiU6Tx9SP-iaA-1 X-Mimecast-MFC-AGG-ID: l7qTyIGDMyiU6Tx9SP-iaA_1774873762 From: Koushik Dutta To: qemu-devel@nongnu.org Cc: "Michael S. Tsirkin" , Stefano Garzarella , Jason Wang , =?UTF-8?q?Eugenio=20P=C3=A9rez?= Subject: [PATCH v4] hw/net/virtio-net: add support for notification coalescing Date: Mon, 30 Mar 2026 17:59:17 +0530 Message-ID: <20260330122917.468225-1-kdutta@redhat.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Scanned-By: MIMEDefang 3.4.1 on 10.30.177.4 Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=170.10.129.124; envelope-from=kdutta@redhat.com; helo=us-smtp-delivery-124.mimecast.com X-Spam_score_int: 27 X-Spam_score: 2.7 X-Spam_bar: ++ X-Spam_report: (2.7 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.54, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_SBL_CSS=3.335, RCVD_IN_VALIDITY_CERTIFIED_BLOCKED=1, RCVD_IN_VALIDITY_RPBL_BLOCKED=1, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001 autolearn=no autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: qemu development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @redhat.com) X-ZM-MESSAGEID: 1774873831610158500 Content-Type: text/plain; charset="utf-8" Implement VirtIO Network Notification Coalescing (Bit 53). This allows the guest to manage interrupt frequency using ethtool -C for both RX and TX paths. - Added VIRTIO_NET_F_NOTF_COAL to host features. - Implemented VIRTIO_NET_CTRL_NOTF_COAL class handling in virtio_net_handle_ctrl_iov. - Added logic to store and apply rx/tx usecs and max_packets. - Added packet counters and threshold logic for both RX and TX data paths. - Dynamic Dispatcher: Implemented a dispatcher mechanism that dynamically switches/activates the notification callback logic only after the guest enables TX coalescing via ethtool. vhost-vdpa: add support for SVQ interrupt coalescing. This reduces interrupt overhead by batching notifications based on either a packet count or a time-based threshold. Signed-off-by: Koushik Dutta --- hw/net/virtio-net.c | 137 +++++++++++++++++++++++++++++---- include/hw/virtio/virtio-net.h | 8 ++ net/vhost-vdpa.c | 2 + 3 files changed, 131 insertions(+), 16 deletions(-) diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c index eccb48ad42..16f197cbdb 100644 --- a/hw/net/virtio-net.c +++ b/hw/net/virtio-net.c @@ -131,6 +131,8 @@ static const VirtIOConfigSizeParams cfg_size_params =3D= { .feature_sizes =3D feature_sizes }; =20 +static void virtio_net_tx_timer(void *opaque); + static VirtIONetQueue *virtio_net_get_subqueue(NetClientState *nc) { VirtIONet *n =3D qemu_get_nic_opaque(nc); @@ -157,6 +159,15 @@ static void flush_or_purge_queued_packets(NetClientSta= te *nc) * - we could suppress RX interrupt if we were so inclined. */ =20 +static void virtio_net_rx_notify(void *opaque) +{ + VirtIONet *n =3D opaque; + VirtIODevice *vdev =3D VIRTIO_DEVICE(n); + + n->rx_pkt_cnt =3D 0; + virtio_notify(vdev, n->vqs[0].rx_vq); +} + static void virtio_net_get_config(VirtIODevice *vdev, uint8_t *config) { VirtIONet *n =3D VIRTIO_NET(vdev); @@ -1081,6 +1092,43 @@ static int virtio_net_handle_offloads(VirtIONet *n, = uint8_t cmd, } } =20 +static int virtio_net_handle_coal(VirtIONet *n, uint8_t cmd, + struct iovec *iov, unsigned int iov_cnt) +{ + struct virtio_net_ctrl_coal coal; + VirtIONetQueue *q; + size_t s; + int i; + + s =3D iov_to_buf(iov, iov_cnt, 0, &coal, sizeof(coal)); + if (s !=3D sizeof(coal)) { + return VIRTIO_NET_ERR; + } + + if (cmd =3D=3D VIRTIO_NET_CTRL_NOTF_COAL_RX_SET) { + n->rx_coal_usecs =3D le32_to_cpu(coal.max_usecs); + n->rx_coal_packets =3D le32_to_cpu(coal.max_packets); + if (!n->rx_index_timer) { + n->rx_index_timer =3D timer_new_us(QEMU_CLOCK_VIRTUAL, + virtio_net_rx_notify, n); + } + } else if (cmd =3D=3D VIRTIO_NET_CTRL_NOTF_COAL_TX_SET) { + n->tx_coal_usecs =3D le32_to_cpu(coal.max_usecs); + n->tx_coal_packets =3D le32_to_cpu(coal.max_packets); + n->tx_timeout =3D n->tx_coal_usecs * 1000; + for (i =3D 0; i < n->max_queue_pairs; i++) { + q =3D &n->vqs[i]; + if (!q->tx_timer) { + q->tx_timer =3D timer_new_ns(QEMU_CLOCK_VIRTUAL, + virtio_net_tx_timer, + q); + } + } + } + + return VIRTIO_NET_OK; +} + static int virtio_net_handle_mac(VirtIONet *n, uint8_t cmd, struct iovec *iov, unsigned int iov_cnt) { @@ -1582,6 +1630,8 @@ size_t virtio_net_handle_ctrl_iov(VirtIODevice *vdev, status =3D virtio_net_handle_mq(n, ctrl.cmd, iov, out_num); } else if (ctrl.class =3D=3D VIRTIO_NET_CTRL_GUEST_OFFLOADS) { status =3D virtio_net_handle_offloads(n, ctrl.cmd, iov, out_num); + } else if (ctrl.class =3D=3D VIRTIO_NET_CTRL_NOTF_COAL) { + status =3D virtio_net_handle_coal(n, ctrl.cmd, iov, out_num); } =20 s =3D iov_from_buf(in_sg, in_num, 0, &status, sizeof(status)); @@ -2041,7 +2091,22 @@ static ssize_t virtio_net_receive_rcu(NetClientState= *nc, const uint8_t *buf, } =20 virtqueue_flush(q->rx_vq, i); - virtio_notify(vdev, q->rx_vq); + + /* rx coalescing */ + n->rx_pkt_cnt +=3D i; + if (n->rx_coal_usecs =3D=3D 0 || n->rx_pkt_cnt >=3D n->rx_coal_packets= ) { + if (n->rx_index_timer) { + timer_del(n->rx_index_timer); + } + virtio_net_rx_notify(n); + } else { + if (n->rx_index_timer) { + if (!timer_pending(n->rx_index_timer)) { + timer_mod(n->rx_index_timer, + qemu_clock_get_us(QEMU_CLOCK_VIRTUAL) + n->rx_co= al_usecs); + } + } + } =20 return size; =20 @@ -2818,7 +2883,6 @@ detach: return -EINVAL; } =20 -static void virtio_net_tx_timer(void *opaque); =20 static void virtio_net_handle_tx_timer(VirtIODevice *vdev, VirtQueue *vq) { @@ -2900,6 +2964,12 @@ static void virtio_net_tx_timer(void *opaque) if (ret =3D=3D -EBUSY || ret =3D=3D -EINVAL) { return; } + if (n->tx_pkt_cnt < ret) { + n->tx_pkt_cnt =3D 0; + } else { + n->tx_pkt_cnt -=3D ret; + } + /* * If we flush a full burst of packets, assume there are * more coming and immediately rearm @@ -2919,6 +2989,7 @@ static void virtio_net_tx_timer(void *opaque) ret =3D virtio_net_flush_tx(q); if (ret > 0) { virtio_queue_set_notification(q->tx_vq, 0); + n->tx_pkt_cnt -=3D ret; q->tx_waiting =3D 1; timer_mod(q->tx_timer, qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) + n->tx_timeout); @@ -2974,6 +3045,32 @@ static void virtio_net_tx_bh(void *opaque) } } =20 +static void virtio_net_handle_tx_dispatch(VirtIODevice *vdev, VirtQueue *v= q) +{ + VirtIONet *n =3D VIRTIO_NET(vdev); + VirtIONetQueue *q =3D &n->vqs[vq2q(virtio_get_queue_index(vq))]; + bool use_timer =3D n->tx_timer_activate || n->tx_coal_usecs > 0 || + n->tx_coal_packets > 0; + bool pkt_limit =3D (n->tx_coal_packets > 0); + + if (use_timer) { + n->tx_pkt_cnt++; + if (!pkt_limit || n->tx_pkt_cnt < n->tx_coal_packets) { + if (q->tx_timer) { + virtio_net_handle_tx_timer(vdev, vq); + return; + } + } + n->tx_pkt_cnt =3D 0; + if (q->tx_timer) { + timer_del(q->tx_timer); + } + virtio_net_handle_tx_bh(vdev, vq); + } else { + virtio_net_handle_tx_bh(vdev, vq); + } +} + static void virtio_net_add_queue(VirtIONet *n, int index) { VirtIODevice *vdev =3D VIRTIO_DEVICE(n); @@ -2981,20 +3078,15 @@ static void virtio_net_add_queue(VirtIONet *n, int = index) n->vqs[index].rx_vq =3D virtio_add_queue(vdev, n->net_conf.rx_queue_si= ze, virtio_net_handle_rx); =20 - if (n->net_conf.tx && !strcmp(n->net_conf.tx, "timer")) { - n->vqs[index].tx_vq =3D - virtio_add_queue(vdev, n->net_conf.tx_queue_size, - virtio_net_handle_tx_timer); - n->vqs[index].tx_timer =3D timer_new_ns(QEMU_CLOCK_VIRTUAL, - virtio_net_tx_timer, - &n->vqs[index]); - } else { - n->vqs[index].tx_vq =3D - virtio_add_queue(vdev, n->net_conf.tx_queue_size, - virtio_net_handle_tx_bh); - n->vqs[index].tx_bh =3D qemu_bh_new_guarded(virtio_net_tx_bh, &n->= vqs[index], - &DEVICE(vdev)->mem_reent= rancy_guard); - } + n->vqs[index].tx_vq =3D + virtio_add_queue(vdev, + n->net_conf.tx_queue_size, + virtio_net_handle_tx_dispatch); + + n->vqs[index].tx_bh =3D + qemu_bh_new_guarded(virtio_net_tx_bh, + &n->vqs[index], + &DEVICE(vdev)->mem_reentrancy_guard); =20 n->vqs[index].tx_waiting =3D 0; n->vqs[index].n =3D n; @@ -3972,6 +4064,10 @@ static void virtio_net_device_realize(DeviceState *d= ev, Error **errp) error_printf("Defaulting to \"bh\""); } =20 + if (n->net_conf.tx && strcmp(n->net_conf.tx, "timer")) { + n->tx_timer_activate =3D true; + } + n->net_conf.tx_queue_size =3D MIN(virtio_net_max_tx_queue_size(n), n->net_conf.tx_queue_size); =20 @@ -4048,6 +4144,13 @@ static void virtio_net_device_realize(DeviceState *d= ev, Error **errp) n->rss_data.specified_hash_types.on_bits | n->rss_data.specified_hash_types.auto_bits; } + n->rx_pkt_cnt =3D 0; + n->tx_pkt_cnt =3D 0; + n->rx_coal_usecs =3D 0; + n->tx_coal_usecs =3D 0; + n->rx_coal_packets =3D 0; + n->tx_coal_packets =3D 0; + n->rx_index_timer =3D NULL; } =20 static void virtio_net_device_unrealize(DeviceState *dev) @@ -4262,6 +4365,8 @@ static const Property virtio_net_properties[] =3D { VIRTIO_NET_F_GUEST_USO6, true), DEFINE_PROP_BIT64("host_uso", VirtIONet, host_features, VIRTIO_NET_F_HOST_USO, true), + DEFINE_PROP_BIT64("vq_notf_coal", VirtIONet, host_features, + VIRTIO_NET_F_NOTF_COAL, true), DEFINE_PROP_ON_OFF_AUTO_BIT64("hash-ipv4", VirtIONet, rss_data.specified_hash_types, VIRTIO_NET_HASH_REPORT_IPv4 - 1, diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virtio-net.h index 5b8ab7bda7..024501ed37 100644 --- a/include/hw/virtio/virtio-net.h +++ b/include/hw/virtio/virtio-net.h @@ -231,6 +231,14 @@ struct VirtIONet { struct EBPFRSSContext ebpf_rss; uint32_t nr_ebpf_rss_fds; char **ebpf_rss_fds; + QEMUTimer *rx_index_timer; + uint32_t rx_coal_usecs; + uint32_t rx_coal_packets; + uint32_t rx_pkt_cnt; + uint32_t tx_coal_usecs; + uint32_t tx_coal_packets; + uint32_t tx_pkt_cnt; + bool tx_timer_activate; }; =20 size_t virtio_net_handle_ctrl_iov(VirtIODevice *vdev, diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c index 3df6091274..a20db78b81 100644 --- a/net/vhost-vdpa.c +++ b/net/vhost-vdpa.c @@ -70,6 +70,7 @@ static const int vdpa_feature_bits[] =3D { VIRTIO_NET_F_CTRL_RX, VIRTIO_NET_F_CTRL_RX_EXTRA, VIRTIO_NET_F_CTRL_VLAN, + VIRTIO_NET_F_NOTF_COAL, VIRTIO_NET_F_CTRL_VQ, VIRTIO_NET_F_GSO, VIRTIO_NET_F_GUEST_CSUM, @@ -115,6 +116,7 @@ static const uint64_t vdpa_svq_device_features =3D BIT_ULL(VIRTIO_NET_F_HOST_UFO) | BIT_ULL(VIRTIO_NET_F_MRG_RXBUF) | BIT_ULL(VIRTIO_NET_F_STATUS) | + BIT_ULL(VIRTIO_NET_F_NOTF_COAL) | BIT_ULL(VIRTIO_NET_F_CTRL_VQ) | BIT_ULL(VIRTIO_NET_F_GSO) | BIT_ULL(VIRTIO_NET_F_CTRL_RX) | --=20 2.53.0