From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id 6A50BA0C3F; Thu, 15 Apr 2021 14:37:37 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 5329D161F0C; Thu, 15 Apr 2021 14:37:37 +0200 (CEST) Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [216.205.24.124]) by mails.dpdk.org (Postfix) with ESMTP id B5616161F02 for ; Thu, 15 Apr 2021 14:37:35 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1618490254; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=ebpIYtd0Cwr4FuQG4qlKt66kwGO17GwGgQ2VWpetd+M=; b=OvsH7pPuWB5i9PNxx2YBgWr3HsuILpe2Kt054d+dXJBR9nxcPSRynXL4DzSTHqNprByWmY 8a+RAMsmI/4w1iKelka+UhFMzVjOwOBqjwADYEkhcHUOFk0ETBliBKFacv11oCEVOQUlMd VYIezQ4MQpKk4gtOV6HDsSnTnaERfjU= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-207-8kt4jtJrN_GvmQrx6dDipQ-1; Thu, 15 Apr 2021 08:37:33 -0400 X-MC-Unique: 8kt4jtJrN_GvmQrx6dDipQ-1 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 3616BBBEE8 for ; Thu, 15 Apr 2021 12:37:32 +0000 (UTC) Received: from [10.36.110.28] (unknown [10.36.110.28]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 50ADBE17B; Thu, 15 Apr 2021 12:37:31 +0000 (UTC) To: Balazs Nemeth , dev@dpdk.org References: From: Maxime Coquelin Message-ID: Date: Thu, 15 Apr 2021 14:37:28 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.8.1 MIME-Version: 1.0 In-Reply-To: X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 Authentication-Results: relay.mimecast.com; auth=pass smtp.auth=CUSA124A263 smtp.mailfrom=maxime.coquelin@redhat.com X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 8bit Subject: Re: [dpdk-dev] [PATCH v2 1/4] vhost: move allocation of mbuf outside of packet enqueue X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" Hi Balazs, Hint for future revisions, please add a cover letter when multiple patches, it makes the series handling for the maintainer easier. Also, please use the MAINTAINERS file to add the other maintainers. On 4/7/21 12:17 PM, Balazs Nemeth wrote: > In preparation for subsequent patches, move mbuf allocation out and > rename virtio_dev_pktmbuf_alloc to virtio_dev_pktmbuf_prep. This > function now receives an already allocated mbuf pointer. > > Signed-off-by: Balazs Nemeth > --- > lib/librte_vhost/virtio_net.c | 54 ++++++++++++++++++++++++++--------- > 1 file changed, 40 insertions(+), 14 deletions(-) > > diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c > index 7f621fb6d..666e7fdb8 100644 > --- a/lib/librte_vhost/virtio_net.c > +++ b/lib/librte_vhost/virtio_net.c > @@ -2166,6 +2166,23 @@ virtio_dev_pktmbuf_alloc(struct virtio_net *dev, struct rte_mempool *mp, > return NULL; > } > > +static __rte_always_inline int > +virtio_dev_pktmbuf_prep(struct virtio_net *dev, struct rte_mbuf *pkt, > + uint32_t data_len) > +{ > + if (rte_pktmbuf_tailroom(pkt) >= data_len) > + return 0; > + > + /* attach an external buffer if supported */ > + if (dev->extbuf && !virtio_dev_extbuf_alloc(pkt, data_len)) > + return 0; > + > + /* check if chained buffers are allowed */ > + if (!dev->linearbuf) > + return 0; Add new line here. > + return 1; Maybe return a negative value for consistency. > +} > + > static __rte_noinline uint16_t > virtio_dev_tx_split(struct virtio_net *dev, struct vhost_virtqueue *vq, > struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count) > @@ -2259,7 +2276,6 @@ virtio_dev_tx_split(struct virtio_net *dev, struct vhost_virtqueue *vq, > static __rte_always_inline int > vhost_reserve_avail_batch_packed(struct virtio_net *dev, > struct vhost_virtqueue *vq, > - struct rte_mempool *mbuf_pool, > struct rte_mbuf **pkts, > uint16_t avail_idx, > uintptr_t *desc_addrs, > @@ -2304,8 +2320,7 @@ vhost_reserve_avail_batch_packed(struct virtio_net *dev, > } > > vhost_for_each_try_unroll(i, 0, PACKED_BATCH_SIZE) { > - pkts[i] = virtio_dev_pktmbuf_alloc(dev, mbuf_pool, lens[i]); > - if (!pkts[i]) > + if (virtio_dev_pktmbuf_prep(dev, pkts[i], lens[i])) > goto free_buf; > } > > @@ -2326,16 +2341,12 @@ vhost_reserve_avail_batch_packed(struct virtio_net *dev, > return 0; > > free_buf: It is no more freeing here, so better to rename the label. > - for (i = 0; i < PACKED_BATCH_SIZE; i++) > - rte_pktmbuf_free(pkts[i]); > - > return -1; > } > > static __rte_always_inline int > virtio_dev_tx_batch_packed(struct virtio_net *dev, > struct vhost_virtqueue *vq, > - struct rte_mempool *mbuf_pool, > struct rte_mbuf **pkts) > { > uint16_t avail_idx = vq->last_avail_idx; > @@ -2345,8 +2356,8 @@ virtio_dev_tx_batch_packed(struct virtio_net *dev, > uint16_t ids[PACKED_BATCH_SIZE]; > uint16_t i; > > - if (vhost_reserve_avail_batch_packed(dev, vq, mbuf_pool, pkts, > - avail_idx, desc_addrs, ids)) > + if (vhost_reserve_avail_batch_packed(dev, vq, pkts, avail_idx, > + desc_addrs, ids)) > return -1; > > vhost_for_each_try_unroll(i, 0, PACKED_BATCH_SIZE) > @@ -2396,8 +2407,8 @@ vhost_dequeue_single_packed(struct virtio_net *dev, > VHOST_ACCESS_RO) < 0)) > return -1; > > - *pkts = virtio_dev_pktmbuf_alloc(dev, mbuf_pool, buf_len); > - if (unlikely(*pkts == NULL)) { > + > + if (unlikely(virtio_dev_pktmbuf_prep(dev, *pkts, buf_len))) { > if (!allocerr_warned) { > VHOST_LOG_DATA(ERR, > "Failed mbuf alloc of size %d from %s on %s.\n", > @@ -2416,7 +2427,6 @@ vhost_dequeue_single_packed(struct virtio_net *dev, > dev->ifname); > allocerr_warned = true; > } > - rte_pktmbuf_free(*pkts); > return -1; > } > > @@ -2459,22 +2469,38 @@ virtio_dev_tx_packed(struct virtio_net *dev, > { > uint32_t pkt_idx = 0; > uint32_t remained = count; > + uint16_t i; > > do { > rte_prefetch0(&vq->desc_packed[vq->last_avail_idx]); > > if (remained >= PACKED_BATCH_SIZE) { > - if (!virtio_dev_tx_batch_packed(dev, vq, mbuf_pool, > + vhost_for_each_try_unroll(i, 0, PACKED_BATCH_SIZE) { > + pkts[pkt_idx + i] = > + rte_pktmbuf_alloc(mbuf_pool); No check on whether the alloc succeeded? Also, we recently move to up to 100 chars lines, so maybe it can fit in a single one here now. > + } > + > + if (!virtio_dev_tx_batch_packed(dev, vq, > &pkts[pkt_idx])) { Ditto > pkt_idx += PACKED_BATCH_SIZE; > remained -= PACKED_BATCH_SIZE; > + > continue; > + } else { > + vhost_for_each_try_unroll(i, 0, > + PACKED_BATCH_SIZE) { Same here > + rte_pktmbuf_free(pkts[pkt_idx + i]); > + } > } > } > > + pkts[pkt_idx] = rte_pktmbuf_alloc(mbuf_pool); Here also you may want to ensure the allocation succeeded. > + > if (virtio_dev_tx_single_packed(dev, vq, mbuf_pool, > - &pkts[pkt_idx])) > + &pkts[pkt_idx])) { > + rte_pktmbuf_free(pkts[pkt_idx]); > break; > + } > pkt_idx++; > remained--; > > Maxime