From patchwork Wed Jul 6 22:28:41 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Shirley Ma X-Patchwork-Id: 103587 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id D9F6FB6F74 for ; Thu, 7 Jul 2011 08:29:06 +1000 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752933Ab1GFW2r (ORCPT ); Wed, 6 Jul 2011 18:28:47 -0400 Received: from e3.ny.us.ibm.com ([32.97.182.143]:55908 "EHLO e3.ny.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751514Ab1GFW2p (ORCPT ); Wed, 6 Jul 2011 18:28:45 -0400 Received: from d01relay04.pok.ibm.com (d01relay04.pok.ibm.com [9.56.227.236]) by e3.ny.us.ibm.com (8.14.4/8.13.1) with ESMTP id p66M5aCK021279; Wed, 6 Jul 2011 18:05:36 -0400 Received: from d03av04.boulder.ibm.com (d03av04.boulder.ibm.com [9.17.195.170]) by d01relay04.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id p66MSiFh155166; Wed, 6 Jul 2011 18:28:44 -0400 Received: from d03av04.boulder.ibm.com (loopback [127.0.0.1]) by d03av04.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id p66GShmH015803; Wed, 6 Jul 2011 10:28:44 -0600 Received: from [9.49.135.143] (sig-9-49-135-143.mts.ibm.com [9.49.135.143]) by d03av04.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id p66GSgfm015718; Wed, 6 Jul 2011 10:28:42 -0600 Subject: [PATCH V8 4/4 net-next] vhost: vhost TX zero-copy support From: Shirley Ma To: David Miller , mst@redhat.com Cc: netdev@vger.kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org Date: Wed, 06 Jul 2011 15:28:41 -0700 Message-ID: <1309991321.10209.26.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Evolution 2.28.3 (2.28.3-1.fc12) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org This patch maintains the outstanding userspace buffers in the sequence it is delivered to vhost. The outstanding userspace buffers will be marked as done once the lower device buffers DMA has finished. This is monitored through last reference of kfree_skb callback. Two buffer index are used for this purpose. The vhost passes the userspace buffers info to lower device skb through message control. Since there will be some done DMAs when entering vhost handle_tx. The worse case is all buffers in the vq are in pending/done status, so we need to notify guest to release DMA done buffers first before get any new buffers from the vq. The busywait is waiting for fix in clean up and ring changes. Signed-off-by: Shirley --- drivers/vhost/net.c | 45 ++++++++++++++++++++++++++++++++++++++++++++- drivers/vhost/vhost.c | 48 ++++++++++++++++++++++++++++++++++++++++++++++++ drivers/vhost/vhost.h | 15 +++++++++++++++ 3 files changed, 107 insertions(+), 1 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c index e224a92..7de0c6e 100644 --- a/drivers/vhost/net.c +++ b/drivers/vhost/net.c @@ -32,6 +32,10 @@ * Using this limit prevents one virtqueue from starving others. */ #define VHOST_NET_WEIGHT 0x80000 +/* MAX number of TX used buffers for outstanding zerocopy */ +#define VHOST_MAX_PEND 128 +#define VHOST_GOODCOPY_LEN 256 + enum { VHOST_NET_VQ_RX = 0, VHOST_NET_VQ_TX = 1, @@ -151,6 +155,10 @@ static void handle_tx(struct vhost_net *net) hdr_size = vq->vhost_hlen; for (;;) { + /* Release DMAs done buffers first */ + if (atomic_read(&vq->refcnt) > VHOST_MAX_PEND) + vhost_zerocopy_signal_used(vq); + head = vhost_get_vq_desc(&net->dev, vq, vq->iov, ARRAY_SIZE(vq->iov), &out, &in, @@ -166,6 +174,12 @@ static void handle_tx(struct vhost_net *net) set_bit(SOCK_ASYNC_NOSPACE, &sock->flags); break; } + /* If more outstanding DMAs, queue the work */ + if (atomic_read(&vq->refcnt) > VHOST_MAX_PEND) { + tx_poll_start(net, sock); + set_bit(SOCK_ASYNC_NOSPACE, &sock->flags); + break; + } if (unlikely(vhost_enable_notify(&net->dev, vq))) { vhost_disable_notify(&net->dev, vq); continue; @@ -188,6 +202,26 @@ static void handle_tx(struct vhost_net *net) iov_length(vq->hdr, s), hdr_size); break; } + /* use msg_control to pass vhost zerocopy ubuf info to skb */ + if (sock_flag(sock->sk, SOCK_ZEROCOPY)) { + vq->heads[vq->upend_idx].id = head; + if (len < VHOST_GOODCOPY_LEN) + /* copy don't need to wait for DMA done */ + vq->heads[vq->upend_idx].len = + VHOST_DMA_DONE_LEN; + else { + struct ubuf_info *ubuf = &vq->ubuf_info[head]; + + vq->heads[vq->upend_idx].len = len; + ubuf->callback = vhost_zerocopy_callback; + ubuf->arg = vq; + ubuf->desc = vq->upend_idx; + msg.msg_control = ubuf; + msg.msg_controllen = sizeof(ubuf); + } + atomic_inc(&vq->refcnt); + vq->upend_idx = (vq->upend_idx + 1) % UIO_MAXIOV; + } /* TODO: Check specific error and bomb out unless ENOBUFS? */ err = sock->ops->sendmsg(NULL, sock, &msg, len); if (unlikely(err < 0)) { @@ -198,12 +232,21 @@ static void handle_tx(struct vhost_net *net) if (err != len) pr_debug("Truncated TX packet: " " len %d != %zd\n", err, len); - vhost_add_used_and_signal(&net->dev, vq, head, 0); + if (!sock_flag(sock->sk, SOCK_ZEROCOPY)) + vhost_add_used_and_signal(&net->dev, vq, head, 0); total_len += len; if (unlikely(total_len >= VHOST_NET_WEIGHT)) { vhost_poll_queue(&vq->poll); break; } + /* if upend_idx is full, then wait for free more */ +/* + if (unlikely(vq->upend_idx == vq->done_idx)) { + tx_poll_start(net, sock); + set_bit(SOCK_ASYNC_NOSPACE, &sock->flags); + break; + } +*/ } mutex_unlock(&vq->mutex); diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index ea966b3..db242b1 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c @@ -179,6 +179,9 @@ static void vhost_vq_reset(struct vhost_dev *dev, vq->call_ctx = NULL; vq->call = NULL; vq->log_ctx = NULL; + vq->upend_idx = 0; + vq->done_idx = 0; + atomic_set(&vq->refcnt, 0); } static int vhost_worker(void *data) @@ -237,6 +240,8 @@ static long vhost_dev_alloc_iovecs(struct vhost_dev *dev) GFP_KERNEL); dev->vqs[i].heads = kmalloc(sizeof *dev->vqs[i].heads * UIO_MAXIOV, GFP_KERNEL); + dev->vqs[i].ubuf_info = kmalloc(sizeof *dev->vqs[i].ubuf_info * + UIO_MAXIOV, GFP_KERNEL); if (!dev->vqs[i].indirect || !dev->vqs[i].log || !dev->vqs[i].heads) @@ -249,6 +254,7 @@ err_nomem: kfree(dev->vqs[i].indirect); kfree(dev->vqs[i].log); kfree(dev->vqs[i].heads); + kfree(dev->vqs[i].ubuf_info); } return -ENOMEM; } @@ -390,6 +396,30 @@ long vhost_dev_reset_owner(struct vhost_dev *dev) return 0; } +/* In case of DMA done not in order in lower device driver for some reason. + * upend_idx is used to track end of used idx, done_idx is used to track head + * of used idx. Once lower device DMA done contiguously, we will signal KVM + * guest used idx. + */ +void vhost_zerocopy_signal_used(struct vhost_virtqueue *vq) +{ + int i, j = 0; + + for (i = vq->done_idx; i != vq->upend_idx; i = (i + 1) % UIO_MAXIOV) { + if ((vq->heads[i].len == VHOST_DMA_DONE_LEN)) { + vq->heads[i].len = VHOST_DMA_CLEAR_LEN; + vhost_add_used_and_signal(vq->dev, vq, + vq->heads[i].id, 0); + ++j; + } else + break; + } + if (j) { + vq->done_idx = i; + atomic_sub(j, &vq->refcnt); + } +} + /* Caller should have device mutex */ void vhost_dev_cleanup(struct vhost_dev *dev) { @@ -400,6 +430,9 @@ void vhost_dev_cleanup(struct vhost_dev *dev) vhost_poll_stop(&dev->vqs[i].poll); vhost_poll_flush(&dev->vqs[i].poll); } + /* Wait for all lower device DMAs done (busywait FIXME) */ + while (atomic_read(&dev->vqs[i].refcnt)) + vhost_zerocopy_signal_used(&dev->vqs[i]); if (dev->vqs[i].error_ctx) eventfd_ctx_put(dev->vqs[i].error_ctx); if (dev->vqs[i].error) @@ -612,6 +645,11 @@ static long vhost_set_vring(struct vhost_dev *d, int ioctl, void __user *argp) mutex_lock(&vq->mutex); + /* clean up lower device outstanding DMAs, before setting ring + busywait FIXME */ + while (atomic_read(&vq->refcnt)) + vhost_zerocopy_signal_used(vq); + switch (ioctl) { case VHOST_SET_VRING_NUM: /* Resizing ring with an active backend? @@ -1486,3 +1524,13 @@ void vhost_disable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq) &vq->used->flags, r); } } + +void vhost_zerocopy_callback(void *arg) +{ + struct ubuf_info *ubuf = (struct ubuf_info *)arg; + struct vhost_virtqueue *vq; + + vq = (struct vhost_virtqueue *)ubuf->arg; + /* set len = 1 to mark this desc buffers done DMA */ + vq->heads[ubuf->desc].len = VHOST_DMA_DONE_LEN; +} diff --git a/drivers/vhost/vhost.h b/drivers/vhost/vhost.h index 8e03379..883688c 100644 --- a/drivers/vhost/vhost.h +++ b/drivers/vhost/vhost.h @@ -13,6 +13,11 @@ #include #include +/* This is for zerocopy, used buffer len is set to 1 when lower device DMA + * done */ +#define VHOST_DMA_DONE_LEN 1 +#define VHOST_DMA_CLEAR_LEN 0 + struct vhost_device; struct vhost_work; @@ -114,6 +119,14 @@ struct vhost_virtqueue { /* Log write descriptors */ void __user *log_base; struct vhost_log *log; + /* vhost zerocopy support */ + atomic_t refcnt; /* num of outstanding zerocopy DMAs */ + /* last used idx for outstanding DMA zerocopy buffers */ + int upend_idx; + /* first used idx for DMA done zerocopy buffers */ + int done_idx; + /* an array of userspace buffers info */ + struct ubuf_info *ubuf_info; }; struct vhost_dev { @@ -160,6 +173,8 @@ bool vhost_enable_notify(struct vhost_dev *, struct vhost_virtqueue *); int vhost_log_write(struct vhost_virtqueue *vq, struct vhost_log *log, unsigned int log_num, u64 len); +void vhost_zerocopy_callback(void *arg); +void vhost_zerocopy_signal_used(struct vhost_virtqueue *vq); #define vq_err(vq, fmt, ...) do { \ pr_debug(pr_fmt(fmt), ##__VA_ARGS__); \