From patchwork Wed Nov 15 14:12:55 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Vitaly Kuznetsov X-Patchwork-Id: 838189 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 3ycRDW0PMmz9sDB for ; Thu, 16 Nov 2017 01:13:14 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932368AbdKOONC (ORCPT ); Wed, 15 Nov 2017 09:13:02 -0500 Received: from mx1.redhat.com ([209.132.183.28]:36204 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753041AbdKOOM7 (ORCPT ); Wed, 15 Nov 2017 09:12:59 -0500 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 1583DC02C734; Wed, 15 Nov 2017 14:12:59 +0000 (UTC) Received: from vitty.brq.redhat.com (unknown [10.43.2.155]) by smtp.corp.redhat.com (Postfix) with ESMTP id 69D9C11E7D5; Wed, 15 Nov 2017 14:12:57 +0000 (UTC) From: Vitaly Kuznetsov To: netdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org, devel@linuxdriverproject.org, "K. Y. Srinivasan" , Haiyang Zhang , Stephen Hemminger , Mohammed Gamal Subject: [PATCH net v2] hv_netvsc: preserve hw_features on mtu/channels/ringparam changes Date: Wed, 15 Nov 2017 15:12:55 +0100 Message-Id: <20171115141255.24103-1-vkuznets@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.31]); Wed, 15 Nov 2017 14:12:59 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org rndis_filter_device_add() is called both from netvsc_probe() when we initially create the device and from set channels/mtu/ringparam routines where we basically remove the device and add it back. hw_features is reset in rndis_filter_device_add() and filled with host data. However, we lose all additional flags which are set outside of the driver, e.g. register_netdevice() adds NETIF_F_SOFT_FEATURES and many others. Unfortunately, calls to rndis_{query_hwcaps(), _set_offload_params()} calls cannot be avoided on every RNDIS reset: host expects us to set required features explicitly. Moreover, in theory hardware capabilities can change and we need to reflect the change in hw_features. Reset net->hw_features bits according to host data in rndis_netdev_set_hwcaps(), clear corresponding feature bits from net->features in case some features went missing (will never happen in real life I guess but let's be consistent). Signed-off-by: Vitaly Kuznetsov Reviewed-by: Haiyang Zhang --- Changes since v1: call rndis_netdev_set_hwcaps() unconditionally; clear net->hw_features flags we set and set them again based on host's data. Clear missing features from net->features. Unfortunately, not calling rndis_netdev_set_hwcaps() on mtu/channels/ ringparam changes is not an option: we still need to call rndis_{query_hwcaps(),set_offload_params()} or we will lose features [Haiyang Zhang] --- drivers/net/hyperv/hyperv_net.h | 4 ++ drivers/net/hyperv/netvsc_drv.c | 2 +- drivers/net/hyperv/rndis_filter.c | 136 ++++++++++++++++++++++---------------- 3 files changed, 83 insertions(+), 59 deletions(-) diff --git a/drivers/net/hyperv/hyperv_net.h b/drivers/net/hyperv/hyperv_net.h index 5176be76ca7d..b6cbda2963ad 100644 --- a/drivers/net/hyperv/hyperv_net.h +++ b/drivers/net/hyperv/hyperv_net.h @@ -646,6 +646,10 @@ struct nvsp_message { #define NETVSC_RECEIVE_BUFFER_ID 0xcafe #define NETVSC_SEND_BUFFER_ID 0 +#define NETVSC_SUPPORTED_HW_FEATURES (NETIF_F_RXCSUM | NETIF_F_IP_CSUM | \ + NETIF_F_TSO | NETIF_F_IPV6_CSUM | \ + NETIF_F_TSO6) + #define VRSS_SEND_TAB_SIZE 16 /* must be power of 2 */ #define VRSS_CHANNEL_MAX 64 #define VRSS_CHANNEL_DEFAULT 8 diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c index a32ae02e1b6c..fc1d0b4bca22 100644 --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@ -1953,7 +1953,7 @@ static int netvsc_probe(struct hv_device *dev, memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN); - /* hw_features computed in rndis_filter_device_add */ + /* hw_features computed in rndis_netdev_set_hwcaps() */ net->features = net->hw_features | NETIF_F_HIGHDMA | NETIF_F_SG | NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX; diff --git a/drivers/net/hyperv/rndis_filter.c b/drivers/net/hyperv/rndis_filter.c index 065b204d8e17..cf1243347a0f 100644 --- a/drivers/net/hyperv/rndis_filter.c +++ b/drivers/net/hyperv/rndis_filter.c @@ -1128,69 +1128,20 @@ void rndis_set_subchannel(struct work_struct *w) rtnl_unlock(); } -struct netvsc_device *rndis_filter_device_add(struct hv_device *dev, - struct netvsc_device_info *device_info) +static int rndis_netdev_set_hwcaps(struct rndis_device *rndis_device, + struct netvsc_device *nvdev) { - struct net_device *net = hv_get_drvdata(dev); + struct net_device *net = rndis_device->ndev; struct net_device_context *net_device_ctx = netdev_priv(net); - struct netvsc_device *net_device; - struct rndis_device *rndis_device; struct ndis_offload hwcaps; struct ndis_offload_params offloads; - struct ndis_recv_scale_cap rsscap; - u32 rsscap_size = sizeof(struct ndis_recv_scale_cap); unsigned int gso_max_size = GSO_MAX_SIZE; - u32 mtu, size; - const struct cpumask *node_cpu_mask; - u32 num_possible_rss_qs; - int i, ret; - - rndis_device = get_rndis_device(); - if (!rndis_device) - return ERR_PTR(-ENODEV); - - /* - * Let the inner driver handle this first to create the netvsc channel - * NOTE! Once the channel is created, we may get a receive callback - * (RndisFilterOnReceive()) before this call is completed - */ - net_device = netvsc_device_add(dev, device_info); - if (IS_ERR(net_device)) { - kfree(rndis_device); - return net_device; - } - - /* Initialize the rndis device */ - net_device->max_chn = 1; - net_device->num_chn = 1; - - net_device->extension = rndis_device; - rndis_device->ndev = net; - - /* Send the rndis initialization message */ - ret = rndis_filter_init_device(rndis_device, net_device); - if (ret != 0) - goto err_dev_remv; - - /* Get the MTU from the host */ - size = sizeof(u32); - ret = rndis_filter_query_device(rndis_device, net_device, - RNDIS_OID_GEN_MAXIMUM_FRAME_SIZE, - &mtu, &size); - if (ret == 0 && size == sizeof(u32) && mtu < net->mtu) - net->mtu = mtu; - - /* Get the mac address */ - ret = rndis_filter_query_device_mac(rndis_device, net_device); - if (ret != 0) - goto err_dev_remv; - - memcpy(device_info->mac_adr, rndis_device->hw_mac_adr, ETH_ALEN); + int ret; /* Find HW offload capabilities */ - ret = rndis_query_hwcaps(rndis_device, net_device, &hwcaps); + ret = rndis_query_hwcaps(rndis_device, nvdev, &hwcaps); if (ret != 0) - goto err_dev_remv; + return ret; /* A value of zero means "no change"; now turn on what we want. */ memset(&offloads, 0, sizeof(struct ndis_offload_params)); @@ -1198,8 +1149,12 @@ struct netvsc_device *rndis_filter_device_add(struct hv_device *dev, /* Linux does not care about IP checksum, always does in kernel */ offloads.ip_v4_csum = NDIS_OFFLOAD_PARAMETERS_TX_RX_DISABLED; + /* Reset previously set hw_features flags */ + net->hw_features &= ~NETVSC_SUPPORTED_HW_FEATURES; + net_device_ctx->tx_checksum_mask = 0; + /* Compute tx offload settings based on hw capabilities */ - net->hw_features = NETIF_F_RXCSUM; + net->hw_features |= NETIF_F_RXCSUM; if ((hwcaps.csum.ip4_txcsum & NDIS_TXCSUM_ALL_TCP4) == NDIS_TXCSUM_ALL_TCP4) { /* Can checksum TCP */ @@ -1243,10 +1198,75 @@ struct netvsc_device *rndis_filter_device_add(struct hv_device *dev, } } + /* In case some hw_features disappeared we need to remove them from + * net->features list as they're no longer supported. + */ + net->features &= ~NETVSC_SUPPORTED_HW_FEATURES | net->hw_features; + netif_set_gso_max_size(net, gso_max_size); - ret = rndis_filter_set_offload_params(net, net_device, &offloads); - if (ret) + ret = rndis_filter_set_offload_params(net, nvdev, &offloads); + + return ret; +} + +struct netvsc_device *rndis_filter_device_add(struct hv_device *dev, + struct netvsc_device_info *device_info) +{ + struct net_device *net = hv_get_drvdata(dev); + struct netvsc_device *net_device; + struct rndis_device *rndis_device; + struct ndis_recv_scale_cap rsscap; + u32 rsscap_size = sizeof(struct ndis_recv_scale_cap); + u32 mtu, size; + const struct cpumask *node_cpu_mask; + u32 num_possible_rss_qs; + int i, ret; + + rndis_device = get_rndis_device(); + if (!rndis_device) + return ERR_PTR(-ENODEV); + + /* Let the inner driver handle this first to create the netvsc channel + * NOTE! Once the channel is created, we may get a receive callback + * (RndisFilterOnReceive()) before this call is completed + */ + net_device = netvsc_device_add(dev, device_info); + if (IS_ERR(net_device)) { + kfree(rndis_device); + return net_device; + } + + /* Initialize the rndis device */ + net_device->max_chn = 1; + net_device->num_chn = 1; + + net_device->extension = rndis_device; + rndis_device->ndev = net; + + /* Send the rndis initialization message */ + ret = rndis_filter_init_device(rndis_device, net_device); + if (ret != 0) + goto err_dev_remv; + + /* Get the MTU from the host */ + size = sizeof(u32); + ret = rndis_filter_query_device(rndis_device, net_device, + RNDIS_OID_GEN_MAXIMUM_FRAME_SIZE, + &mtu, &size); + if (ret == 0 && size == sizeof(u32) && mtu < net->mtu) + net->mtu = mtu; + + /* Get the mac address */ + ret = rndis_filter_query_device_mac(rndis_device, net_device); + if (ret != 0) + goto err_dev_remv; + + memcpy(device_info->mac_adr, rndis_device->hw_mac_adr, ETH_ALEN); + + /* Query and set hardware capabilities */ + ret = rndis_netdev_set_hwcaps(rndis_device, net_device); + if (ret != 0) goto err_dev_remv; rndis_filter_query_device_link_status(rndis_device, net_device);