Patch Detail
get:
Show a patch.
patch:
Update a patch.
put:
Update a patch.
GET /api/patches/110645/?format=api
{ "id": 110645, "url": "https://patches.dpdk.org/api/patches/110645/?format=api", "web_url": "https://patches.dpdk.org/project/dpdk/patch/20220505062335.25445-2-xuan.ding@intel.com/", "project": { "id": 1, "url": "https://patches.dpdk.org/api/projects/1/?format=api", "name": "DPDK", "link_name": "dpdk", "list_id": "dev.dpdk.org", "list_email": "dev@dpdk.org", "web_url": "http://core.dpdk.org", "scm_url": "git://dpdk.org/dpdk", "webscm_url": "http://git.dpdk.org/dpdk", "list_archive_url": "https://inbox.dpdk.org/dev", "list_archive_url_format": "https://inbox.dpdk.org/dev/{}", "commit_url_format": "" }, "msgid": "<20220505062335.25445-2-xuan.ding@intel.com>", "list_archive_url": "https://inbox.dpdk.org/dev/20220505062335.25445-2-xuan.ding@intel.com", "date": "2022-05-05T06:23:31", "name": "[v4,1/5] vhost: prepare sync for descriptor to mbuf refactoring", "commit_ref": null, "pull_url": null, "state": "superseded", "archived": true, "hash": "39991b6a2bd117a5ebf6a9cf06e2abf4b783b330", "submitter": { "id": 1401, "url": "https://patches.dpdk.org/api/people/1401/?format=api", "name": "Ding, Xuan", "email": "xuan.ding@intel.com" }, "delegate": { "id": 2642, "url": "https://patches.dpdk.org/api/users/2642/?format=api", "username": "mcoquelin", "first_name": "Maxime", "last_name": "Coquelin", "email": "maxime.coquelin@redhat.com" }, "mbox": "https://patches.dpdk.org/project/dpdk/patch/20220505062335.25445-2-xuan.ding@intel.com/mbox/", "series": [ { "id": 22784, "url": "https://patches.dpdk.org/api/series/22784/?format=api", "web_url": "https://patches.dpdk.org/project/dpdk/list/?series=22784", "date": "2022-05-05T06:23:30", "name": "vhost: support async dequeue data path", "version": 4, "mbox": "https://patches.dpdk.org/series/22784/mbox/" } ], "comments": "https://patches.dpdk.org/api/patches/110645/comments/", "check": "warning", "checks": "https://patches.dpdk.org/api/patches/110645/checks/", "tags": {}, "related": [], "headers": { "Return-Path": "<dev-bounces@dpdk.org>", "X-Original-To": "patchwork@inbox.dpdk.org", "Delivered-To": "patchwork@inbox.dpdk.org", "Received": [ "from mails.dpdk.org (mails.dpdk.org [217.70.189.124])\n\tby inbox.dpdk.org (Postfix) with ESMTP id 0A1ACA00BE;\n\tThu, 5 May 2022 08:31:01 +0200 (CEST)", "from [217.70.189.124] (localhost [127.0.0.1])\n\tby mails.dpdk.org (Postfix) with ESMTP id 1291C42821;\n\tThu, 5 May 2022 08:30:52 +0200 (CEST)", "from mga02.intel.com (mga02.intel.com [134.134.136.20])\n by mails.dpdk.org (Postfix) with ESMTP id 1808F40C35\n for <dev@dpdk.org>; Thu, 5 May 2022 08:30:46 +0200 (CEST)", "from orsmga007.jf.intel.com ([10.7.209.58])\n by orsmga101.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384;\n 04 May 2022 23:27:09 -0700", "from npg-dpdk-xuan-cbdma.sh.intel.com ([10.67.110.228])\n by orsmga007.jf.intel.com with ESMTP; 04 May 2022 23:27:07 -0700" ], "DKIM-Signature": "v=1; a=rsa-sha256; c=relaxed/simple;\n d=intel.com; i=@intel.com; q=dns/txt; s=Intel;\n t=1651732247; x=1683268247;\n h=from:to:cc:subject:date:message-id:in-reply-to: references;\n bh=99kj52PLeozOwqPkaQjO1C7XmTqmmJ2BHLpQzt8s5q8=;\n b=G1Ul3oQw58ChyOFM8bxcPN/615RRcVmu2OVq2pp6akTkWazezGBgj8Pk\n fCG8xCx8mqTaGWhGgy9JXvHBZg8wLaxvoNa2Wtq/s6Y+9zy3/9PbStAg8\n jbcSpY22ng3XA36MV/JZrBWbVi8l7kPmtY8leRmu9p/YneQB1MlCmS8N/\n xScuxSaass3k7SXpDxPd25mLpHquEkCaNyW+2URd5ugGxxDmF2sIwqVSk\n dnzLHtdyhTMMLK0WTQk9nvJinIXozUDOa5uAMlmm+HTbcDQSSXUPUXvIC\n 4W1zQXOrde99N+3F0kh0CNcwCQceJBwJQZTRDJqlT70hOpjxcfEn1MJMt w==;", "X-IronPort-AV": [ "E=McAfee;i=\"6400,9594,10337\"; a=\"255488227\"", "E=Sophos;i=\"5.91,200,1647327600\"; d=\"scan'208\";a=\"255488227\"", "E=Sophos;i=\"5.91,200,1647327600\"; d=\"scan'208\";a=\"563092940\"" ], "X-ExtLoop1": "1", "From": "xuan.ding@intel.com", "To": "maxime.coquelin@redhat.com,\n\tchenbo.xia@intel.com", "Cc": "dev@dpdk.org, jiayu.hu@intel.com, cheng1.jiang@intel.com,\n sunil.pai.g@intel.com, liangma@liangbit.com,\n Xuan Ding <xuan.ding@intel.com>", "Subject": "[PATCH v4 1/5] vhost: prepare sync for descriptor to mbuf refactoring", "Date": "Thu, 5 May 2022 06:23:31 +0000", "Message-Id": "<20220505062335.25445-2-xuan.ding@intel.com>", "X-Mailer": "git-send-email 2.17.1", "In-Reply-To": "<20220505062335.25445-1-xuan.ding@intel.com>", "References": "<20220407152546.38167-1-xuan.ding@intel.com>\n <20220505062335.25445-1-xuan.ding@intel.com>", "X-BeenThere": "dev@dpdk.org", "X-Mailman-Version": "2.1.29", "Precedence": "list", "List-Id": "DPDK patches and discussions <dev.dpdk.org>", "List-Unsubscribe": "<https://mails.dpdk.org/options/dev>,\n <mailto:dev-request@dpdk.org?subject=unsubscribe>", "List-Archive": "<http://mails.dpdk.org/archives/dev/>", "List-Post": "<mailto:dev@dpdk.org>", "List-Help": "<mailto:dev-request@dpdk.org?subject=help>", "List-Subscribe": "<https://mails.dpdk.org/listinfo/dev>,\n <mailto:dev-request@dpdk.org?subject=subscribe>", "Errors-To": "dev-bounces@dpdk.org" }, "content": "From: Xuan Ding <xuan.ding@intel.com>\n\nThis patch extracts the descriptors to buffers filling from\ncopy_desc_to_mbuf() into a dedicated function. Besides, enqueue\nand dequeue path are refactored to use the same function\nsync_fill_seg() for preparing batch elements, which simplifies\nthe code without performance degradation.\n\nSigned-off-by: Xuan Ding <xuan.ding@intel.com>\nReviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>\n---\n lib/vhost/virtio_net.c | 78 ++++++++++++++++++++----------------------\n 1 file changed, 38 insertions(+), 40 deletions(-)", "diff": "diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c\nindex 5f432b0d77..d4c94d2a9b 100644\n--- a/lib/vhost/virtio_net.c\n+++ b/lib/vhost/virtio_net.c\n@@ -1030,23 +1030,36 @@ async_mbuf_to_desc_seg(struct virtio_net *dev, struct vhost_virtqueue *vq,\n }\n \n static __rte_always_inline void\n-sync_mbuf_to_desc_seg(struct virtio_net *dev, struct vhost_virtqueue *vq,\n+sync_fill_seg(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \t\tstruct rte_mbuf *m, uint32_t mbuf_offset,\n-\t\tuint64_t buf_addr, uint64_t buf_iova, uint32_t cpy_len)\n+\t\tuint64_t buf_addr, uint64_t buf_iova, uint32_t cpy_len, bool to_desc)\n {\n \tstruct batch_copy_elem *batch_copy = vq->batch_copy_elems;\n \n \tif (likely(cpy_len > MAX_BATCH_LEN || vq->batch_copy_nb_elems >= vq->size)) {\n-\t\trte_memcpy((void *)((uintptr_t)(buf_addr)),\n+\t\tif (to_desc) {\n+\t\t\trte_memcpy((void *)((uintptr_t)(buf_addr)),\n \t\t\t\trte_pktmbuf_mtod_offset(m, void *, mbuf_offset),\n \t\t\t\tcpy_len);\n+\t\t} else {\n+\t\t\trte_memcpy(rte_pktmbuf_mtod_offset(m, void *, mbuf_offset),\n+\t\t\t\t(void *)((uintptr_t)(buf_addr)),\n+\t\t\t\tcpy_len);\n+\t\t}\n \t\tvhost_log_cache_write_iova(dev, vq, buf_iova, cpy_len);\n \t\tPRINT_PACKET(dev, (uintptr_t)(buf_addr), cpy_len, 0);\n \t} else {\n-\t\tbatch_copy[vq->batch_copy_nb_elems].dst =\n-\t\t\t(void *)((uintptr_t)(buf_addr));\n-\t\tbatch_copy[vq->batch_copy_nb_elems].src =\n-\t\t\trte_pktmbuf_mtod_offset(m, void *, mbuf_offset);\n+\t\tif (to_desc) {\n+\t\t\tbatch_copy[vq->batch_copy_nb_elems].dst =\n+\t\t\t\t(void *)((uintptr_t)(buf_addr));\n+\t\t\tbatch_copy[vq->batch_copy_nb_elems].src =\n+\t\t\t\trte_pktmbuf_mtod_offset(m, void *, mbuf_offset);\n+\t\t} else {\n+\t\t\tbatch_copy[vq->batch_copy_nb_elems].dst =\n+\t\t\t\trte_pktmbuf_mtod_offset(m, void *, mbuf_offset);\n+\t\t\tbatch_copy[vq->batch_copy_nb_elems].src =\n+\t\t\t\t(void *)((uintptr_t)(buf_addr));\n+\t\t}\n \t\tbatch_copy[vq->batch_copy_nb_elems].log_addr = buf_iova;\n \t\tbatch_copy[vq->batch_copy_nb_elems].len = cpy_len;\n \t\tvq->batch_copy_nb_elems++;\n@@ -1158,9 +1171,9 @@ mbuf_to_desc(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \t\t\t\t\t\tbuf_iova + buf_offset, cpy_len) < 0)\n \t\t\t\tgoto error;\n \t\t} else {\n-\t\t\tsync_mbuf_to_desc_seg(dev, vq, m, mbuf_offset,\n-\t\t\t\t\tbuf_addr + buf_offset,\n-\t\t\t\t\tbuf_iova + buf_offset, cpy_len);\n+\t\t\tsync_fill_seg(dev, vq, m, mbuf_offset,\n+\t\t\t\t buf_addr + buf_offset,\n+\t\t\t\t buf_iova + buf_offset, cpy_len, true);\n \t\t}\n \n \t\tmbuf_avail -= cpy_len;\n@@ -2473,8 +2486,8 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \t\t struct rte_mbuf *m, struct rte_mempool *mbuf_pool,\n \t\t bool legacy_ol_flags)\n {\n-\tuint32_t buf_avail, buf_offset;\n-\tuint64_t buf_addr, buf_len;\n+\tuint32_t buf_avail, buf_offset, buf_len;\n+\tuint64_t buf_addr, buf_iova;\n \tuint32_t mbuf_avail, mbuf_offset;\n \tuint32_t cpy_len;\n \tstruct rte_mbuf *cur = m, *prev = m;\n@@ -2482,16 +2495,13 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \tstruct virtio_net_hdr *hdr = NULL;\n \t/* A counter to avoid desc dead loop chain */\n \tuint16_t vec_idx = 0;\n-\tstruct batch_copy_elem *batch_copy = vq->batch_copy_elems;\n-\tint error = 0;\n \n \tbuf_addr = buf_vec[vec_idx].buf_addr;\n+\tbuf_iova = buf_vec[vec_idx].buf_iova;\n \tbuf_len = buf_vec[vec_idx].buf_len;\n \n-\tif (unlikely(buf_len < dev->vhost_hlen && nr_vec <= 1)) {\n-\t\terror = -1;\n-\t\tgoto out;\n-\t}\n+\tif (unlikely(buf_len < dev->vhost_hlen && nr_vec <= 1))\n+\t\treturn -1;\n \n \tif (virtio_net_with_host_offload(dev)) {\n \t\tif (unlikely(buf_len < sizeof(struct virtio_net_hdr))) {\n@@ -2515,11 +2525,12 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \t\tbuf_offset = dev->vhost_hlen - buf_len;\n \t\tvec_idx++;\n \t\tbuf_addr = buf_vec[vec_idx].buf_addr;\n+\t\tbuf_iova = buf_vec[vec_idx].buf_iova;\n \t\tbuf_len = buf_vec[vec_idx].buf_len;\n \t\tbuf_avail = buf_len - buf_offset;\n \t} else if (buf_len == dev->vhost_hlen) {\n \t\tif (unlikely(++vec_idx >= nr_vec))\n-\t\t\tgoto out;\n+\t\t\tgoto error;\n \t\tbuf_addr = buf_vec[vec_idx].buf_addr;\n \t\tbuf_len = buf_vec[vec_idx].buf_len;\n \n@@ -2539,22 +2550,9 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \twhile (1) {\n \t\tcpy_len = RTE_MIN(buf_avail, mbuf_avail);\n \n-\t\tif (likely(cpy_len > MAX_BATCH_LEN ||\n-\t\t\t\t\tvq->batch_copy_nb_elems >= vq->size ||\n-\t\t\t\t\t(hdr && cur == m))) {\n-\t\t\trte_memcpy(rte_pktmbuf_mtod_offset(cur, void *,\n-\t\t\t\t\t\tmbuf_offset),\n-\t\t\t\t\t(void *)((uintptr_t)(buf_addr +\n-\t\t\t\t\t\t\tbuf_offset)), cpy_len);\n-\t\t} else {\n-\t\t\tbatch_copy[vq->batch_copy_nb_elems].dst =\n-\t\t\t\trte_pktmbuf_mtod_offset(cur, void *,\n-\t\t\t\t\t\tmbuf_offset);\n-\t\t\tbatch_copy[vq->batch_copy_nb_elems].src =\n-\t\t\t\t(void *)((uintptr_t)(buf_addr + buf_offset));\n-\t\t\tbatch_copy[vq->batch_copy_nb_elems].len = cpy_len;\n-\t\t\tvq->batch_copy_nb_elems++;\n-\t\t}\n+\t\tsync_fill_seg(dev, vq, cur, mbuf_offset,\n+\t\t\t buf_addr + buf_offset,\n+\t\t\t buf_iova + buf_offset, cpy_len, false);\n \n \t\tmbuf_avail -= cpy_len;\n \t\tmbuf_offset += cpy_len;\n@@ -2567,6 +2565,7 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \t\t\t\tbreak;\n \n \t\t\tbuf_addr = buf_vec[vec_idx].buf_addr;\n+\t\t\tbuf_iova = buf_vec[vec_idx].buf_iova;\n \t\t\tbuf_len = buf_vec[vec_idx].buf_len;\n \n \t\t\tbuf_offset = 0;\n@@ -2585,8 +2584,7 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \t\t\tif (unlikely(cur == NULL)) {\n \t\t\t\tVHOST_LOG_DATA(ERR, \"(%s) failed to allocate memory for mbuf.\\n\",\n \t\t\t\t\t\tdev->ifname);\n-\t\t\t\terror = -1;\n-\t\t\t\tgoto out;\n+\t\t\t\tgoto error;\n \t\t\t}\n \n \t\t\tprev->next = cur;\n@@ -2606,9 +2604,9 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq,\n \tif (hdr)\n \t\tvhost_dequeue_offload(dev, hdr, m, legacy_ol_flags);\n \n-out:\n-\n-\treturn error;\n+\treturn 0;\n+error:\n+\treturn -1;\n }\n \n static void\n", "prefixes": [ "v4", "1/5" ] }