From patchwork Thu Apr 6 06:56:51 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Gautam Dawar X-Patchwork-Id: 13202924 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 54E69C7618D for ; Thu, 6 Apr 2023 07:03:48 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236005AbjDFHDq (ORCPT ); Thu, 6 Apr 2023 03:03:46 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40556 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S235835AbjDFHCw (ORCPT ); Thu, 6 Apr 2023 03:02:52 -0400 Received: from NAM04-DM6-obe.outbound.protection.outlook.com (mail-dm6nam04on20614.outbound.protection.outlook.com [IPv6:2a01:111:f400:7e8b::614]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 5FB21B469; Thu, 6 Apr 2023 00:01:46 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=Jl8qN1eFKyFGHZDE7FTbEO+Wg18iEaC0jOGZLQKkqVr7u3zJPQyYcqBzMizSVlE7nfRjL++SpwuyxD91Vt+K/STZVtTVksWzuOUcLMlOLYiK3OUkCdDIynHNCblQq5cVrP0Mj5jxpaYqV3kBoxBGogg9tX5y2niQN4kmGRsuTw3qUaG+pR1rLr2fAz4RgLynZ8cGvASMPakIiBPc3exHpu0Nm9vh5R0nfDTWHKh86/cO5tltDhTIlaOGuqXHJlPhcQLiiWjrD9N9EoSq3I1sMiOGSMP0Qr7sdSCoUSpdccxWT5xQ4mlN5TOk1ui5Q1VwPRYW1pGP/2Dh6XdjGX/N5g== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=ISWQp6iwJ7hFkG2HbzSXaHXg2EwbrFjsjAZU6ewMwS0=; b=Enl0MsF2l9gXAOkh2VG1RYdEKvUhChs+Z/qAOnEhDDJnMBQRb7/5RoZGMznNgN7IaevDcEszv1iXVH0q8ewSMAS3ntNPyB3cRQdoK18rCNgJdEefn9Cm2nXNkCkmDDXoCYlZnJ4yMcmpUc/kpwjEFKV0ydYFQuf1hionSUyxMCP0CgW1B2arIxdNpiTzfcRRUBDpV4E9gQrsibLCDqR/hcg5OsZF927nLK95+eRZwUBOjTmXjAOZu3BQgdzKQlw22raKtLBBsGUnAj/8+cO5YIaldAgmzA1nSt5uCZkohjb/BS7O/ATnLJzeVaW0NTMjyP0zeiMVoBFoQu9IL3ZM2Q== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass (sender ip is 165.204.84.17) smtp.rcpttodomain=redhat.com smtp.mailfrom=amd.com; dmarc=pass (p=quarantine sp=quarantine pct=100) action=none header.from=amd.com; dkim=none (message not signed); arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amd.com; s=selector1; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=ISWQp6iwJ7hFkG2HbzSXaHXg2EwbrFjsjAZU6ewMwS0=; b=HZ9qqQDlpWDA1c1Mfc4mN0U6pTe7yhYy0VZviMhkUOmOJwY7azub93TfVzXmBjJtV1+AeJknxhlgWZJ0TqE+m+gMZqU4P+abMTS/0usRNypXRAUB6/7+fHngP4KB1xD3lipB8Vx8tmBuFf1I8AYN8uH3/4bKtpebqppPKCgj3aU= Received: from MW4PR04CA0221.namprd04.prod.outlook.com (2603:10b6:303:87::16) by BL1PR12MB5828.namprd12.prod.outlook.com (2603:10b6:208:397::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6254.35; Thu, 6 Apr 2023 07:01:22 +0000 Received: from CO1NAM11FT109.eop-nam11.prod.protection.outlook.com (2603:10b6:303:87:cafe::f7) by MW4PR04CA0221.outlook.office365.com (2603:10b6:303:87::16) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6277.31 via Frontend Transport; Thu, 6 Apr 2023 07:01:22 +0000 X-MS-Exchange-Authentication-Results: spf=pass (sender IP is 165.204.84.17) smtp.mailfrom=amd.com; dkim=none (message not signed) header.d=none;dmarc=pass action=none header.from=amd.com; Received-SPF: Pass (protection.outlook.com: domain of amd.com designates 165.204.84.17 as permitted sender) receiver=protection.outlook.com; client-ip=165.204.84.17; helo=SATLEXMB04.amd.com; pr=C Received: from SATLEXMB04.amd.com (165.204.84.17) by CO1NAM11FT109.mail.protection.outlook.com (10.13.174.176) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.20.6277.30 via Frontend Transport; Thu, 6 Apr 2023 07:01:21 +0000 Received: from SATLEXMB04.amd.com (10.181.40.145) by SATLEXMB04.amd.com (10.181.40.145) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2375.34; Thu, 6 Apr 2023 02:01:06 -0500 Received: from xndengvm004102.xilinx.com (10.180.168.240) by SATLEXMB04.amd.com (10.181.40.145) with Microsoft SMTP Server id 15.1.2375.34 via Frontend Transport; Thu, 6 Apr 2023 02:01:02 -0500 From: Gautam Dawar To: , , Edward Cree , Martin Habets , "David S. Miller" , Eric Dumazet , "Jakub Kicinski" , Paolo Abeni , Richard Cochran , , CC: , , , , Gautam Dawar Subject: [PATCH net-next v3 06/14] sfc: implement vDPA management device operations Date: Thu, 6 Apr 2023 12:26:51 +0530 Message-ID: <20230406065706.59664-7-gautam.dawar@amd.com> X-Mailer: git-send-email 2.30.1 In-Reply-To: <20230406065706.59664-1-gautam.dawar@amd.com> References: <20230406065706.59664-1-gautam.dawar@amd.com> MIME-Version: 1.0 X-EOPAttributedMessage: 0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: CO1NAM11FT109:EE_|BL1PR12MB5828:EE_ X-MS-Office365-Filtering-Correlation-Id: 1615bcc7-7c67-4ac3-e3de-08db366cbabd X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; X-Microsoft-Antispam-Message-Info: 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 X-Forefront-Antispam-Report: CIP:165.204.84.17;CTRY:US;LANG:en;SCL:1;SRV:;IPV:CAL;SFV:NSPM;H:SATLEXMB04.amd.com;PTR:InfoDomainNonexistent;CAT:NONE;SFS:(13230028)(4636009)(376002)(39860400002)(136003)(346002)(396003)(451199021)(36840700001)(46966006)(40470700004)(6666004)(478600001)(86362001)(40480700001)(36756003)(83380400001)(40460700003)(47076005)(356005)(82740400003)(81166007)(921005)(2616005)(36860700001)(426003)(336012)(7416002)(44832011)(30864003)(2906002)(54906003)(110136005)(316002)(186003)(1076003)(26005)(8936002)(5660300002)(4326008)(8676002)(82310400005)(41300700001)(70586007)(70206006)(2004002)(36900700001);DIR:OUT;SFP:1101; X-OriginatorOrg: amd.com X-MS-Exchange-CrossTenant-OriginalArrivalTime: 06 Apr 2023 07:01:21.9025 (UTC) X-MS-Exchange-CrossTenant-Network-Message-Id: 1615bcc7-7c67-4ac3-e3de-08db366cbabd X-MS-Exchange-CrossTenant-Id: 3dd8961f-e488-4e60-8e11-a82d994e183d X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=3dd8961f-e488-4e60-8e11-a82d994e183d;Ip=[165.204.84.17];Helo=[SATLEXMB04.amd.com] X-MS-Exchange-CrossTenant-AuthSource: CO1NAM11FT109.eop-nam11.prod.protection.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Anonymous X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem X-MS-Exchange-Transport-CrossTenantHeadersStamped: BL1PR12MB5828 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org To allow vDPA device creation and deletion, add a vDPA management device per function. Currently, the vDPA devices can be created only on a VF. Also, for now only network class of vDPA devices are supported. Signed-off-by: Gautam Dawar --- drivers/net/ethernet/sfc/Makefile | 2 +- drivers/net/ethernet/sfc/ef10.c | 2 +- drivers/net/ethernet/sfc/ef100_nic.c | 26 ++- drivers/net/ethernet/sfc/ef100_nic.h | 9 + drivers/net/ethernet/sfc/ef100_vdpa.c | 228 ++++++++++++++++++++++ drivers/net/ethernet/sfc/ef100_vdpa.h | 84 ++++++++ drivers/net/ethernet/sfc/ef100_vdpa_ops.c | 30 +++ drivers/net/ethernet/sfc/mcdi_functions.c | 9 +- drivers/net/ethernet/sfc/mcdi_functions.h | 3 +- drivers/net/ethernet/sfc/net_driver.h | 6 + 10 files changed, 392 insertions(+), 7 deletions(-) create mode 100644 drivers/net/ethernet/sfc/ef100_vdpa_ops.c diff --git a/drivers/net/ethernet/sfc/Makefile b/drivers/net/ethernet/sfc/Makefile index 3a2bb98d1c3f..bd8ba588b968 100644 --- a/drivers/net/ethernet/sfc/Makefile +++ b/drivers/net/ethernet/sfc/Makefile @@ -12,7 +12,7 @@ sfc-$(CONFIG_SFC_MTD) += mtd.o sfc-$(CONFIG_SFC_SRIOV) += sriov.o ef10_sriov.o ef100_sriov.o ef100_rep.o \ mae.o tc.o tc_bindings.o tc_counters.o -sfc-$(CONFIG_SFC_VDPA) += mcdi_vdpa.o ef100_vdpa.o +sfc-$(CONFIG_SFC_VDPA) += mcdi_vdpa.o ef100_vdpa.o ef100_vdpa_ops.o obj-$(CONFIG_SFC) += sfc.o obj-$(CONFIG_SFC_FALCON) += falcon/ diff --git a/drivers/net/ethernet/sfc/ef10.c b/drivers/net/ethernet/sfc/ef10.c index 7022fb2005a2..366ecd3c80b1 100644 --- a/drivers/net/ethernet/sfc/ef10.c +++ b/drivers/net/ethernet/sfc/ef10.c @@ -589,7 +589,7 @@ static int efx_ef10_probe(struct efx_nic *efx) if (rc) goto fail4; - rc = efx_get_pf_index(efx, &nic_data->pf_index); + rc = efx_get_fn_info(efx, &nic_data->pf_index, NULL); if (rc) goto fail5; diff --git a/drivers/net/ethernet/sfc/ef100_nic.c b/drivers/net/ethernet/sfc/ef100_nic.c index e139cd634ff9..dd26f06665af 100644 --- a/drivers/net/ethernet/sfc/ef100_nic.c +++ b/drivers/net/ethernet/sfc/ef100_nic.c @@ -1171,7 +1171,7 @@ static int ef100_probe_main(struct efx_nic *efx) if (rc) goto fail; - rc = efx_get_pf_index(efx, &nic_data->pf_index); + rc = efx_get_fn_info(efx, &nic_data->pf_index, &nic_data->vf_index); if (rc) goto fail; @@ -1287,13 +1287,35 @@ int ef100_probe_netdev_pf(struct efx_nic *efx) int ef100_probe_vf(struct efx_nic *efx) { - return ef100_probe_main(efx); + struct ef100_nic_data *nic_data __maybe_unused; + int err __maybe_unused; + int rc; + + rc = ef100_probe_main(efx); + if (rc) + return rc; + + if (IS_ENABLED(CONFIG_SFC_VDPA)) { + nic_data = efx->nic_data; + if (nic_data->vdpa_supported) { + err = ef100_vdpa_register_mgmtdev(efx); + if (err) + pci_warn(efx->pci_dev, + "register_mgmtdev failed, rc: %d\n", + err); + } + } + + return 0; } void ef100_remove(struct efx_nic *efx) { struct ef100_nic_data *nic_data = efx->nic_data; + if (IS_ENABLED(CONFIG_SFC_VDPA) && nic_data->vdpa_supported) + ef100_vdpa_unregister_mgmtdev(efx); + if (IS_ENABLED(CONFIG_SFC_SRIOV) && efx->mae) { efx_ef100_fini_reps(efx); efx_fini_mae(efx); diff --git a/drivers/net/ethernet/sfc/ef100_nic.h b/drivers/net/ethernet/sfc/ef100_nic.h index a01e9d643ccd..e63ea555116c 100644 --- a/drivers/net/ethernet/sfc/ef100_nic.h +++ b/drivers/net/ethernet/sfc/ef100_nic.h @@ -69,6 +69,13 @@ enum ef100_bar_config { EF100_BAR_CONFIG_VDPA, }; +#ifdef CONFIG_SFC_VDPA +enum ef100_vdpa_class { + EF100_VDPA_CLASS_NONE, + EF100_VDPA_CLASS_NET, +}; +#endif + struct ef100_nic_data { struct efx_nic *efx; struct efx_buffer mcdi_buf; @@ -76,9 +83,11 @@ struct ef100_nic_data { u32 datapath_caps2; u32 datapath_caps3; unsigned int pf_index; + unsigned int vf_index; u16 warm_boot_count; #ifdef CONFIG_SFC_VDPA bool vdpa_supported; /* true if vdpa is supported on this PCIe FN */ + enum ef100_vdpa_class vdpa_class; #endif u8 port_id[ETH_ALEN]; DECLARE_BITMAP(evq_phases, EFX_MAX_CHANNELS); diff --git a/drivers/net/ethernet/sfc/ef100_vdpa.c b/drivers/net/ethernet/sfc/ef100_vdpa.c index 268c973f7376..4c5a98c9d6c3 100644 --- a/drivers/net/ethernet/sfc/ef100_vdpa.c +++ b/drivers/net/ethernet/sfc/ef100_vdpa.c @@ -10,11 +10,17 @@ #include #include #include +#include #include "ef100_vdpa.h" #include "mcdi_vdpa.h" #include "mcdi_filters.h" #include "ef100_netdev.h" +static struct virtio_device_id ef100_vdpa_id_table[] = { + { .device = VIRTIO_ID_NET, .vendor = PCI_VENDOR_ID_REDHAT_QUMRANET }, + { 0 }, +}; + int ef100_vdpa_init(struct efx_probe_data *probe_data) { struct efx_nic *efx = &probe_data->efx; @@ -41,17 +47,239 @@ int ef100_vdpa_init(struct efx_probe_data *probe_data) return rc; } +static void ef100_vdpa_delete(struct efx_nic *efx) +{ + if (efx->vdpa_nic) { + /* replace with _vdpa_unregister_device later */ + put_device(&efx->vdpa_nic->vdpa_dev.dev); + } +} + void ef100_vdpa_fini(struct efx_probe_data *probe_data) { struct efx_nic *efx = &probe_data->efx; + struct ef100_nic_data *nic_data; if (efx->state != STATE_VDPA && efx->state != STATE_DISABLED) { pci_err(efx->pci_dev, "Invalid efx state %u", efx->state); return; } + /* Handle vdpa device deletion, if not done explicitly */ + ef100_vdpa_delete(efx); + nic_data = efx->nic_data; + nic_data->vdpa_class = EF100_VDPA_CLASS_NONE; efx->state = STATE_PROBED; down_write(&efx->filter_sem); efx_mcdi_filter_table_remove(efx); up_write(&efx->filter_sem); } + +static int get_net_config(struct ef100_vdpa_nic *vdpa_nic) +{ + struct efx_nic *efx = vdpa_nic->efx; + u16 mtu; + int rc; + + vdpa_nic->net_config.max_virtqueue_pairs = + cpu_to_efx_vdpa16(vdpa_nic, vdpa_nic->max_queue_pairs); + + rc = efx_vdpa_get_mtu(efx, &mtu); + if (rc) { + dev_err(&vdpa_nic->vdpa_dev.dev, + "%s: Get MTU for vf:%u failed:%d\n", __func__, + vdpa_nic->vf_index, rc); + return rc; + } + vdpa_nic->net_config.mtu = cpu_to_efx_vdpa16(vdpa_nic, mtu); + vdpa_nic->net_config.status = cpu_to_efx_vdpa16(vdpa_nic, + VIRTIO_NET_S_LINK_UP); + return 0; +} + +static struct ef100_vdpa_nic *ef100_vdpa_create(struct efx_nic *efx, + const char *dev_name, + enum ef100_vdpa_class dev_type, + const u8 *mac) +{ + struct ef100_nic_data *nic_data = efx->nic_data; + struct ef100_vdpa_nic *vdpa_nic; + int rc; + + nic_data->vdpa_class = dev_type; + vdpa_nic = vdpa_alloc_device(struct ef100_vdpa_nic, + vdpa_dev, &efx->pci_dev->dev, + &ef100_vdpa_config_ops, + 1, 1, + dev_name, false); + if (!vdpa_nic) { + pci_err(efx->pci_dev, + "vDPA device allocation failed for vf: %u\n", + nic_data->vf_index); + nic_data->vdpa_class = EF100_VDPA_CLASS_NONE; + return ERR_PTR(-ENOMEM); + } + + mutex_init(&vdpa_nic->lock); + efx->vdpa_nic = vdpa_nic; + vdpa_nic->vdpa_dev.dma_dev = &efx->pci_dev->dev; + vdpa_nic->vdpa_dev.mdev = efx->mgmt_dev; + vdpa_nic->efx = efx; + vdpa_nic->pf_index = nic_data->pf_index; + vdpa_nic->vf_index = nic_data->vf_index; + vdpa_nic->vdpa_state = EF100_VDPA_STATE_INITIALIZED; + vdpa_nic->mac_address = (u8 *)&vdpa_nic->net_config.mac; + + rc = get_net_config(vdpa_nic); + if (rc) + goto err_put_device; + + if (mac) { + ether_addr_copy(vdpa_nic->mac_address, mac); + vdpa_nic->mac_configured = true; + } + + /* _vdpa_register_device when its ready */ + + return vdpa_nic; + +err_put_device: + /* put_device invokes ef100_vdpa_free */ + put_device(&vdpa_nic->vdpa_dev.dev); + return ERR_PTR(rc); +} + +static void ef100_vdpa_net_dev_del(struct vdpa_mgmt_dev *mgmt_dev, + struct vdpa_device *vdev) +{ + struct ef100_nic_data *nic_data; + struct efx_nic *efx; + int rc; + + efx = pci_get_drvdata(to_pci_dev(mgmt_dev->device)); + nic_data = efx->nic_data; + + rc = efx_ef100_set_bar_config(efx, EF100_BAR_CONFIG_EF100); + if (rc) + pci_err(efx->pci_dev, + "set_bar_config EF100 failed, err: %d\n", rc); + else + pci_dbg(efx->pci_dev, + "vdpa net device deleted, vf: %u\n", + nic_data->vf_index); +} + +static int ef100_vdpa_net_dev_add(struct vdpa_mgmt_dev *mgmt_dev, + const char *name, + const struct vdpa_dev_set_config *config) +{ + struct ef100_vdpa_nic *vdpa_nic; + struct ef100_nic_data *nic_data; + const u8 *mac = NULL; + struct efx_nic *efx; + int rc, err; + + if (config->mask & BIT_ULL(VDPA_ATTR_DEV_NET_CFG_MACADDR)) { + if (!is_valid_ether_addr(config->net.mac)) { + pci_err(efx->pci_dev, "Invalid MAC address %pM\n", + config->net.mac); + return -EINVAL; + } + mac = (const u8 *)config->net.mac; + } + + efx = pci_get_drvdata(to_pci_dev(mgmt_dev->device)); + if (efx->vdpa_nic) { + pci_warn(efx->pci_dev, + "vDPA device already exists on this VF\n"); + return -EEXIST; + } + + nic_data = efx->nic_data; + + rc = efx_ef100_set_bar_config(efx, EF100_BAR_CONFIG_VDPA); + if (rc) { + pci_err(efx->pci_dev, + "set_bar_config vDPA failed, err: %d\n", rc); + goto err_set_bar_config; + } + + vdpa_nic = ef100_vdpa_create(efx, name, EF100_VDPA_CLASS_NET, mac); + if (IS_ERR(vdpa_nic)) { + pci_err(efx->pci_dev, + "vDPA device creation failed, vf: %u, err: %ld\n", + nic_data->vf_index, PTR_ERR(vdpa_nic)); + rc = PTR_ERR(vdpa_nic); + goto err_set_bar_config; + } else { + pci_dbg(efx->pci_dev, + "vdpa net device created, vf: %u\n", + nic_data->vf_index); + } + + return 0; + +err_set_bar_config: + err = efx_ef100_set_bar_config(efx, EF100_BAR_CONFIG_EF100); + if (err) + pci_err(efx->pci_dev, + "set_bar_config EF100 failed, err: %d\n", err); + + return rc; +} + +static const struct vdpa_mgmtdev_ops ef100_vdpa_net_mgmtdev_ops = { + .dev_add = ef100_vdpa_net_dev_add, + .dev_del = ef100_vdpa_net_dev_del +}; + +int ef100_vdpa_register_mgmtdev(struct efx_nic *efx) +{ + struct vdpa_mgmt_dev *mgmt_dev; + u64 features; + int rc; + + mgmt_dev = kzalloc(sizeof(*mgmt_dev), GFP_KERNEL); + if (!mgmt_dev) + return -ENOMEM; + + rc = efx_vdpa_get_features(efx, EF100_VDPA_DEVICE_TYPE_NET, &features); + if (rc) { + pci_err(efx->pci_dev, "%s: MCDI get features error:%d\n", + __func__, rc); + goto err_get_features; + } + + efx->mgmt_dev = mgmt_dev; + mgmt_dev->device = &efx->pci_dev->dev; + mgmt_dev->id_table = ef100_vdpa_id_table; + mgmt_dev->ops = &ef100_vdpa_net_mgmtdev_ops; + mgmt_dev->supported_features = features; + mgmt_dev->max_supported_vqs = EF100_VDPA_MAX_QUEUES_PAIRS * 2; + mgmt_dev->config_attr_mask = BIT_ULL(VDPA_ATTR_DEV_NET_CFG_MACADDR); + + rc = vdpa_mgmtdev_register(mgmt_dev); + if (rc) { + pci_err(efx->pci_dev, + "vdpa_mgmtdev_register failed, err: %d\n", rc); + goto err_mgmtdev_register; + } + + return 0; + +err_mgmtdev_register: +err_get_features: + kfree(mgmt_dev); + efx->mgmt_dev = NULL; + + return rc; +} + +void ef100_vdpa_unregister_mgmtdev(struct efx_nic *efx) +{ + if (efx->mgmt_dev) { + vdpa_mgmtdev_unregister(efx->mgmt_dev); + kfree(efx->mgmt_dev); + efx->mgmt_dev = NULL; + } +} diff --git a/drivers/net/ethernet/sfc/ef100_vdpa.h b/drivers/net/ethernet/sfc/ef100_vdpa.h index ccc5eb0a2a84..1101b30f56e7 100644 --- a/drivers/net/ethernet/sfc/ef100_vdpa.h +++ b/drivers/net/ethernet/sfc/ef100_vdpa.h @@ -17,6 +17,24 @@ #if defined(CONFIG_SFC_VDPA) +/* Max queue pairs currently supported */ +#define EF100_VDPA_MAX_QUEUES_PAIRS 1 + +/** + * enum ef100_vdpa_nic_state - possible states for a vDPA NIC + * + * @EF100_VDPA_STATE_INITIALIZED: State after vDPA NIC created + * @EF100_VDPA_STATE_NEGOTIATED: State after feature negotiation + * @EF100_VDPA_STATE_STARTED: State after driver ok + * @EF100_VDPA_STATE_NSTATES: Number of VDPA states + */ +enum ef100_vdpa_nic_state { + EF100_VDPA_STATE_INITIALIZED, + EF100_VDPA_STATE_NEGOTIATED, + EF100_VDPA_STATE_STARTED, + EF100_VDPA_STATE_NSTATES +}; + enum ef100_vdpa_device_type { EF100_VDPA_DEVICE_TYPE_NET, }; @@ -27,7 +45,73 @@ enum ef100_vdpa_vq_type { EF100_VDPA_VQ_NTYPES }; +/** + * struct ef100_vdpa_nic - vDPA NIC data structure + * + * @vdpa_dev: vdpa_device object which registers on the vDPA bus. + * @vdpa_state: NIC state machine governed by ef100_vdpa_nic_state + * @efx: pointer to the VF's efx_nic object + * @lock: Managing access to vdpa config operations + * @pf_index: PF index of the vDPA VF + * @vf_index: VF index of the vDPA VF + * @status: device status as per VIRTIO spec + * @features: negotiated feature bits + * @max_queue_pairs: maximum number of queue pairs supported + * @net_config: virtio_net_config data + * @mac_address: mac address of interface associated with this vdpa device + * @mac_configured: true after MAC address is configured + */ +struct ef100_vdpa_nic { + struct vdpa_device vdpa_dev; + enum ef100_vdpa_nic_state vdpa_state; + struct efx_nic *efx; + /* for synchronizing access to vdpa config operations */ + struct mutex lock; + u32 pf_index; + u32 vf_index; + u8 status; + u64 features; + u32 max_queue_pairs; + struct virtio_net_config net_config; + u8 *mac_address; + bool mac_configured; +}; + int ef100_vdpa_init(struct efx_probe_data *probe_data); void ef100_vdpa_fini(struct efx_probe_data *probe_data); +int ef100_vdpa_register_mgmtdev(struct efx_nic *efx); +void ef100_vdpa_unregister_mgmtdev(struct efx_nic *efx); + +static inline bool efx_vdpa_is_little_endian(struct ef100_vdpa_nic *vdpa_nic) +{ + return virtio_legacy_is_little_endian() || + (vdpa_nic->features & (1ULL << VIRTIO_F_VERSION_1)); +} + +static inline u16 efx_vdpa16_to_cpu(struct ef100_vdpa_nic *vdpa_nic, + __virtio16 val) +{ + return __virtio16_to_cpu(efx_vdpa_is_little_endian(vdpa_nic), val); +} + +static inline __virtio16 cpu_to_efx_vdpa16(struct ef100_vdpa_nic *vdpa_nic, + u16 val) +{ + return __cpu_to_virtio16(efx_vdpa_is_little_endian(vdpa_nic), val); +} + +static inline u32 efx_vdpa32_to_cpu(struct ef100_vdpa_nic *vdpa_nic, + __virtio32 val) +{ + return __virtio32_to_cpu(efx_vdpa_is_little_endian(vdpa_nic), val); +} + +static inline __virtio32 cpu_to_efx_vdpa32(struct ef100_vdpa_nic *vdpa_nic, + u32 val) +{ + return __cpu_to_virtio32(efx_vdpa_is_little_endian(vdpa_nic), val); +} + +extern const struct vdpa_config_ops ef100_vdpa_config_ops; #endif /* CONFIG_SFC_VDPA */ #endif /* __EF100_VDPA_H__ */ diff --git a/drivers/net/ethernet/sfc/ef100_vdpa_ops.c b/drivers/net/ethernet/sfc/ef100_vdpa_ops.c new file mode 100644 index 000000000000..f1ce011adc43 --- /dev/null +++ b/drivers/net/ethernet/sfc/ef100_vdpa_ops.c @@ -0,0 +1,30 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Driver for AMD network controllers and boards + * Copyright(C) 2023, Advanced Micro Devices, Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include +#include "ef100_vdpa.h" + +static struct ef100_vdpa_nic *get_vdpa_nic(struct vdpa_device *vdev) +{ + return container_of(vdev, struct ef100_vdpa_nic, vdpa_dev); +} + +static void ef100_vdpa_free(struct vdpa_device *vdev) +{ + struct ef100_vdpa_nic *vdpa_nic = get_vdpa_nic(vdev); + + if (vdpa_nic) { + mutex_destroy(&vdpa_nic->lock); + vdpa_nic->efx->vdpa_nic = NULL; + } +} + +const struct vdpa_config_ops ef100_vdpa_config_ops = { + .free = ef100_vdpa_free, +}; diff --git a/drivers/net/ethernet/sfc/mcdi_functions.c b/drivers/net/ethernet/sfc/mcdi_functions.c index d3e6d8239f5c..4415f19cf68f 100644 --- a/drivers/net/ethernet/sfc/mcdi_functions.c +++ b/drivers/net/ethernet/sfc/mcdi_functions.c @@ -413,7 +413,8 @@ int efx_mcdi_window_mode_to_stride(struct efx_nic *efx, u8 vi_window_mode) return 0; } -int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index) +int efx_get_fn_info(struct efx_nic *efx, unsigned int *pf_index, + unsigned int *vf_index) { MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_FUNCTION_INFO_OUT_LEN); size_t outlen; @@ -426,6 +427,10 @@ int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index) if (outlen < sizeof(outbuf)) return -EIO; - *pf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_PF); + if (pf_index) + *pf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_PF); + + if (efx->type->is_vf && vf_index) + *vf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_VF); return 0; } diff --git a/drivers/net/ethernet/sfc/mcdi_functions.h b/drivers/net/ethernet/sfc/mcdi_functions.h index b0e2f53a0d9b..76dc0a13463e 100644 --- a/drivers/net/ethernet/sfc/mcdi_functions.h +++ b/drivers/net/ethernet/sfc/mcdi_functions.h @@ -28,6 +28,7 @@ void efx_mcdi_rx_remove(struct efx_rx_queue *rx_queue); void efx_mcdi_rx_fini(struct efx_rx_queue *rx_queue); int efx_fini_dmaq(struct efx_nic *efx); int efx_mcdi_window_mode_to_stride(struct efx_nic *efx, u8 vi_window_mode); -int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index); +int efx_get_fn_info(struct efx_nic *efx, unsigned int *pf_index, + unsigned int *vf_index); #endif diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h index 3dc9eae5a81d..1da71deac71c 100644 --- a/drivers/net/ethernet/sfc/net_driver.h +++ b/drivers/net/ethernet/sfc/net_driver.h @@ -1090,6 +1090,12 @@ struct efx_nic { int rx_packet_len_offset; int rx_packet_ts_offset; bool rx_scatter; +#ifdef CONFIG_SFC_VDPA + /** @mgmt_dev: vDPA Management device */ + struct vdpa_mgmt_dev *mgmt_dev; + /** @vdpa_nic: vDPA device structure (EF100) */ + struct ef100_vdpa_nic *vdpa_nic; +#endif struct efx_rss_context rss_context; struct mutex rss_lock; u32 vport_id;