From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga14.intel.com (mga14.intel.com [192.55.52.115]) by dpdk.org (Postfix) with ESMTP id 396205677 for ; Thu, 28 Apr 2016 21:04:14 +0200 (CEST) Received: from orsmga003.jf.intel.com ([10.7.209.27]) by fmsmga103.fm.intel.com with ESMTP; 28 Apr 2016 12:04:04 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.24,547,1455004800"; d="scan'208";a="794483793" Received: from dpdk06.sh.intel.com ([10.239.128.225]) by orsmga003.jf.intel.com with ESMTP; 28 Apr 2016 12:04:02 -0700 From: Jianfeng Tan To: dev@dpdk.org Cc: huawei.xie@intel.com, yuanhan.liu@linux.intel.com, Jianfeng Tan Date: Thu, 28 Apr 2016 19:04:01 +0000 Message-Id: <1461870241-83333-3-git-send-email-jianfeng.tan@intel.com> X-Mailer: git-send-email 2.1.4 In-Reply-To: <1461870241-83333-1-git-send-email-jianfeng.tan@intel.com> References: <1461673932-128879-1-git-send-email-jianfeng.tan@intel.com> <1461870241-83333-1-git-send-email-jianfeng.tan@intel.com> Subject: [dpdk-dev] [PATCH v2 2/2] virtio: fix memory leak of virtqueue memzones X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 28 Apr 2016 19:04:14 -0000 Issue: When virtio was proposed in DPDK, there is no API to free memzones. But this has changed since rte_memzone_free() has been implemented by commit ff909fe21f0a ("mem: introduce memzone freeing"). This patch is to make sure memzones in struct virtqueue, like mz and virtio_net_hdr_mz, are freed when queue is released or setup fails. Signed-off-by: Jianfeng Tan --- drivers/net/virtio/virtio_ethdev.c | 21 ++++++++++++++------- drivers/net/virtio/virtqueue.h | 2 ++ 2 files changed, 16 insertions(+), 7 deletions(-) diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c index 0553b67..a14cc88 100644 --- a/drivers/net/virtio/virtio_ethdev.c +++ b/drivers/net/virtio/virtio_ethdev.c @@ -260,12 +260,18 @@ virtio_set_multiple_queues(struct rte_eth_dev *dev, uint16_t nb_queues) } void -virtio_dev_queue_release(struct virtqueue *vq) { +virtio_dev_queue_release(struct virtqueue *vq) +{ struct virtio_hw *hw; if (vq) { hw = vq->hw; - hw->vtpci_ops->del_queue(hw, vq); + if (vq->started) + hw->vtpci_ops->del_queue(hw, vq); + + rte_memzone_free(vq->mz); + if (vq->virtio_net_hdr_mz) + rte_memzone_free(vq->virtio_net_hdr_mz); rte_free(vq->sw_ring); rte_free(vq); @@ -330,7 +336,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, socket_id); if (!vq->sw_ring) { PMD_INIT_LOG(ERR, "Can not allocate RX soft ring"); - rte_free(vq); + virtio_dev_queue_release(vq); return -ENOMEM; } } @@ -358,7 +364,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, if (rte_errno == EEXIST) mz = rte_memzone_lookup(vq_name); if (mz == NULL) { - rte_free(vq); + virtio_dev_queue_release(vq); return -ENOMEM; } } @@ -370,7 +376,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, */ if ((mz->phys_addr + vq->vq_ring_size - 1) >> (VIRTIO_PCI_QUEUE_ADDR_SHIFT + 32)) { PMD_INIT_LOG(ERR, "vring address shouldn't be above 16TB!"); - rte_free(vq); + virtio_dev_queue_release(vq); return -ENOMEM; } @@ -402,7 +408,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, if (rte_errno == EEXIST) hdr_mz = rte_memzone_lookup(vq_name); if (hdr_mz == NULL) { - rte_free(vq); + virtio_dev_queue_release(vq); return -ENOMEM; } } @@ -436,7 +442,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, vq->virtio_net_hdr_mz = rte_memzone_lookup(vq_name); if (vq->virtio_net_hdr_mz == NULL) { - rte_free(vq); + virtio_dev_queue_release(vq); return -ENOMEM; } } @@ -447,6 +453,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, hw->vtpci_ops->setup_queue(hw, vq); + vq->started = 1; *pvq = vq; return 0; } diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h index 4e9239e..83d89ca 100644 --- a/drivers/net/virtio/virtqueue.h +++ b/drivers/net/virtio/virtqueue.h @@ -201,6 +201,8 @@ struct virtqueue { uint16_t *notify_addr; + int started; + struct vq_desc_extra { void *cookie; uint16_t ndescs; -- 2.1.4