From patchwork Thu Mar 10 12:30:27 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Paul Durrant X-Patchwork-Id: 8556971 Return-Path: X-Original-To: patchwork-xen-devel@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork1.web.kernel.org (Postfix) with ESMTP id 405829F6E4 for ; Thu, 10 Mar 2016 12:43:25 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 79FA32026F for ; Thu, 10 Mar 2016 12:43:19 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 4EE3220306 for ; Thu, 10 Mar 2016 12:43:13 +0000 (UTC) Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xen.org with esmtp (Exim 4.84) (envelope-from ) id 1adztD-0004xo-U0; Thu, 10 Mar 2016 12:40:27 +0000 Received: from mail6.bemta14.messagelabs.com ([193.109.254.103]) by lists.xen.org with esmtp (Exim 4.84) (envelope-from ) id 1adztC-0004wA-8v for xen-devel@lists.xenproject.org; Thu, 10 Mar 2016 12:40:26 +0000 Received: from [193.109.254.147] by server-7.bemta-14.messagelabs.com id 1B/5A-04065-A3B61E65; Thu, 10 Mar 2016 12:40:26 +0000 X-Env-Sender: prvs=870045bc7=Paul.Durrant@citrix.com X-Msg-Ref: server-14.tower-27.messagelabs.com!1457613621!17999737!3 X-Originating-IP: [66.165.176.63] X-SpamReason: No, hits=0.0 required=7.0 tests=sa_preprocessor: VHJ1c3RlZCBJUDogNjYuMTY1LjE3Ni42MyA9PiAzMDYwNDg=\n, received_headers: No Received headers X-StarScan-Received: X-StarScan-Version: 8.11; banners=-,-,- X-VirusChecked: Checked Received: (qmail 17212 invoked from network); 10 Mar 2016 12:40:25 -0000 Received: from smtp02.citrix.com (HELO SMTP02.CITRIX.COM) (66.165.176.63) by server-14.tower-27.messagelabs.com with RC4-SHA encrypted SMTP; 10 Mar 2016 12:40:25 -0000 X-IronPort-AV: E=Sophos;i="5.24,315,1454976000"; d="scan'208";a="344615511" From: Paul Durrant To: , Date: Thu, 10 Mar 2016 12:30:27 +0000 Message-ID: <1457613028-10381-3-git-send-email-paul.durrant@citrix.com> X-Mailer: git-send-email 2.1.4 In-Reply-To: <1457613028-10381-1-git-send-email-paul.durrant@citrix.com> References: <1457613028-10381-1-git-send-email-paul.durrant@citrix.com> MIME-Version: 1.0 X-DLP: MIA1 Cc: Paul Durrant , Wei Liu Subject: [Xen-devel] [PATCH net-next 2/3] xen-netback: support multiple extra info fragments passed from frontend X-BeenThere: xen-devel@lists.xen.org X-Mailman-Version: 2.1.18 Precedence: list List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xen.org Sender: "Xen-devel" X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP The code does not currently support a frontend passing multiple extra info fragments to the backend in a tx request. The xenvif_get_extras() function handles multiple extra_info fragments but make_tx_response() assumes there is only ever a single extra info fragment. This patch modifies xenvif_get_extras() to pass back a count of extra info fragments, which is then passed to make_tx_response() (after possibly being stashed in pending_tx_info for deferred responses). Signed-off-by: Paul Durrant Cc: Wei Liu Acked-by: Wei Liu --- drivers/net/xen-netback/common.h | 1 + drivers/net/xen-netback/netback.c | 65 +++++++++++++++++++++++++-------------- 2 files changed, 43 insertions(+), 23 deletions(-) diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h index 1128252..f44b388 100644 --- a/drivers/net/xen-netback/common.h +++ b/drivers/net/xen-netback/common.h @@ -52,6 +52,7 @@ typedef unsigned int pending_ring_idx_t; struct pending_tx_info { struct xen_netif_tx_request req; /* tx request */ + unsigned int extra_count; /* Callback data for released SKBs. The callback is always * xenvif_zerocopy_callback, desc contains the pending_idx, which is * also an index in pending_tx_info array. It is initialized in diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c index 61b97c3..b42f260 100644 --- a/drivers/net/xen-netback/netback.c +++ b/drivers/net/xen-netback/netback.c @@ -95,6 +95,7 @@ static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, static void make_tx_response(struct xenvif_queue *queue, struct xen_netif_tx_request *txp, + unsigned int extra_count, s8 st); static void push_tx_responses(struct xenvif_queue *queue); @@ -696,14 +697,15 @@ void xenvif_tx_credit_callback(unsigned long data) } static void xenvif_tx_err(struct xenvif_queue *queue, - struct xen_netif_tx_request *txp, RING_IDX end) + struct xen_netif_tx_request *txp, + unsigned int extra_count, RING_IDX end) { RING_IDX cons = queue->tx.req_cons; unsigned long flags; do { spin_lock_irqsave(&queue->response_lock, flags); - make_tx_response(queue, txp, XEN_NETIF_RSP_ERROR); + make_tx_response(queue, txp, extra_count, XEN_NETIF_RSP_ERROR); push_tx_responses(queue); spin_unlock_irqrestore(&queue->response_lock, flags); if (cons == end) @@ -724,6 +726,7 @@ static void xenvif_fatal_tx_err(struct xenvif *vif) static int xenvif_count_requests(struct xenvif_queue *queue, struct xen_netif_tx_request *first, + unsigned int extra_count, struct xen_netif_tx_request *txp, int work_to_do) { @@ -812,7 +815,7 @@ static int xenvif_count_requests(struct xenvif_queue *queue, } while (more_data); if (drop_err) { - xenvif_tx_err(queue, first, cons + slots); + xenvif_tx_err(queue, first, extra_count, cons + slots); return drop_err; } @@ -827,9 +830,10 @@ struct xenvif_tx_cb { #define XENVIF_TX_CB(skb) ((struct xenvif_tx_cb *)(skb)->cb) static inline void xenvif_tx_create_map_op(struct xenvif_queue *queue, - u16 pending_idx, - struct xen_netif_tx_request *txp, - struct gnttab_map_grant_ref *mop) + u16 pending_idx, + struct xen_netif_tx_request *txp, + unsigned int extra_count, + struct gnttab_map_grant_ref *mop) { queue->pages_to_map[mop-queue->tx_map_ops] = queue->mmap_pages[pending_idx]; gnttab_set_map_op(mop, idx_to_kaddr(queue, pending_idx), @@ -838,6 +842,7 @@ static inline void xenvif_tx_create_map_op(struct xenvif_queue *queue, memcpy(&queue->pending_tx_info[pending_idx].req, txp, sizeof(*txp)); + queue->pending_tx_info[pending_idx].extra_count = extra_count; } static inline struct sk_buff *xenvif_alloc_skb(unsigned int size) @@ -880,7 +885,7 @@ static struct gnttab_map_grant_ref *xenvif_get_requests(struct xenvif_queue *que shinfo->nr_frags++, txp++, gop++) { index = pending_index(queue->pending_cons++); pending_idx = queue->pending_ring[index]; - xenvif_tx_create_map_op(queue, pending_idx, txp, gop); + xenvif_tx_create_map_op(queue, pending_idx, txp, 0, gop); frag_set_pending_idx(&frags[shinfo->nr_frags], pending_idx); } @@ -893,7 +898,8 @@ static struct gnttab_map_grant_ref *xenvif_get_requests(struct xenvif_queue *que shinfo->nr_frags++, txp++, gop++) { index = pending_index(queue->pending_cons++); pending_idx = queue->pending_ring[index]; - xenvif_tx_create_map_op(queue, pending_idx, txp, gop); + xenvif_tx_create_map_op(queue, pending_idx, txp, 0, + gop); frag_set_pending_idx(&frags[shinfo->nr_frags], pending_idx); } @@ -1095,8 +1101,9 @@ static void xenvif_fill_frags(struct xenvif_queue *queue, struct sk_buff *skb) } static int xenvif_get_extras(struct xenvif_queue *queue, - struct xen_netif_extra_info *extras, - int work_to_do) + struct xen_netif_extra_info *extras, + unsigned int *extra_count, + int work_to_do) { struct xen_netif_extra_info extra; RING_IDX cons = queue->tx.req_cons; @@ -1109,9 +1116,12 @@ static int xenvif_get_extras(struct xenvif_queue *queue, } RING_COPY_REQUEST(&queue->tx, cons, &extra); + + queue->tx.req_cons = ++cons; + (*extra_count)++; + if (unlikely(!extra.type || extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) { - queue->tx.req_cons = ++cons; netdev_err(queue->vif->dev, "Invalid extra type: %d\n", extra.type); xenvif_fatal_tx_err(queue->vif); @@ -1119,7 +1129,6 @@ static int xenvif_get_extras(struct xenvif_queue *queue, } memcpy(&extras[extra.type - 1], &extra, sizeof(extra)); - queue->tx.req_cons = ++cons; } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE); return work_to_do; @@ -1294,6 +1303,7 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, struct xen_netif_tx_request txreq; struct xen_netif_tx_request txfrags[XEN_NETBK_LEGACY_SLOTS_MAX]; struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX-1]; + unsigned int extra_count; u16 pending_idx; RING_IDX idx; int work_to_do; @@ -1330,8 +1340,10 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, queue->tx.req_cons = ++idx; memset(extras, 0, sizeof(extras)); + extra_count = 0; if (txreq.flags & XEN_NETTXF_extra_info) { work_to_do = xenvif_get_extras(queue, extras, + &extra_count, work_to_do); idx = queue->tx.req_cons; if (unlikely(work_to_do < 0)) @@ -1344,7 +1356,7 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, extra = &extras[XEN_NETIF_EXTRA_TYPE_MCAST_ADD - 1]; ret = xenvif_mcast_add(queue->vif, extra->u.mcast.addr); - make_tx_response(queue, &txreq, + make_tx_response(queue, &txreq, extra_count, (ret == 0) ? XEN_NETIF_RSP_OKAY : XEN_NETIF_RSP_ERROR); @@ -1358,12 +1370,14 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, extra = &extras[XEN_NETIF_EXTRA_TYPE_MCAST_DEL - 1]; xenvif_mcast_del(queue->vif, extra->u.mcast.addr); - make_tx_response(queue, &txreq, XEN_NETIF_RSP_OKAY); + make_tx_response(queue, &txreq, extra_count, + XEN_NETIF_RSP_OKAY); push_tx_responses(queue); continue; } - ret = xenvif_count_requests(queue, &txreq, txfrags, work_to_do); + ret = xenvif_count_requests(queue, &txreq, extra_count, + txfrags, work_to_do); if (unlikely(ret < 0)) break; @@ -1372,7 +1386,7 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, if (unlikely(txreq.size < ETH_HLEN)) { netdev_dbg(queue->vif->dev, "Bad packet size: %d\n", txreq.size); - xenvif_tx_err(queue, &txreq, idx); + xenvif_tx_err(queue, &txreq, extra_count, idx); break; } @@ -1397,7 +1411,7 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, if (unlikely(skb == NULL)) { netdev_dbg(queue->vif->dev, "Can't allocate a skb in start_xmit.\n"); - xenvif_tx_err(queue, &txreq, idx); + xenvif_tx_err(queue, &txreq, extra_count, idx); break; } @@ -1416,7 +1430,7 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, nskb = xenvif_alloc_skb(0); if (unlikely(nskb == NULL)) { kfree_skb(skb); - xenvif_tx_err(queue, &txreq, idx); + xenvif_tx_err(queue, &txreq, extra_count, idx); if (net_ratelimit()) netdev_err(queue->vif->dev, "Can't allocate the frag_list skb.\n"); @@ -1457,13 +1471,16 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue, if (data_len < txreq.size) { frag_set_pending_idx(&skb_shinfo(skb)->frags[0], pending_idx); - xenvif_tx_create_map_op(queue, pending_idx, &txreq, gop); + xenvif_tx_create_map_op(queue, pending_idx, &txreq, + extra_count, gop); gop++; } else { frag_set_pending_idx(&skb_shinfo(skb)->frags[0], INVALID_PENDING_IDX); - memcpy(&queue->pending_tx_info[pending_idx].req, &txreq, - sizeof(txreq)); + memcpy(&queue->pending_tx_info[pending_idx].req, + &txreq, sizeof(txreq)); + queue->pending_tx_info[pending_idx].extra_count = + extra_count; } queue->pending_cons++; @@ -1804,7 +1821,8 @@ static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, spin_lock_irqsave(&queue->response_lock, flags); - make_tx_response(queue, &pending_tx_info->req, status); + make_tx_response(queue, &pending_tx_info->req, + pending_tx_info->extra_count, status); /* Release the pending index before pusing the Tx response so * its available before a new Tx request is pushed by the @@ -1821,6 +1839,7 @@ static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, static void make_tx_response(struct xenvif_queue *queue, struct xen_netif_tx_request *txp, + unsigned int extra_count, s8 st) { RING_IDX i = queue->tx.rsp_prod_pvt; @@ -1830,7 +1849,7 @@ static void make_tx_response(struct xenvif_queue *queue, resp->id = txp->id; resp->status = st; - if (txp->flags & XEN_NETTXF_extra_info) + while (extra_count-- != 0) RING_GET_RESPONSE(&queue->tx, ++i)->status = XEN_NETIF_RSP_NULL; queue->tx.rsp_prod_pvt = ++i;