From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id 604B8A0524; Sun, 2 May 2021 12:45:30 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id F3EEA40687; Sun, 2 May 2021 12:45:29 +0200 (CEST) Received: from NAM10-DM6-obe.outbound.protection.outlook.com (mail-dm6nam10on2067.outbound.protection.outlook.com [40.107.93.67]) by mails.dpdk.org (Postfix) with ESMTP id 92FA340685 for ; Sun, 2 May 2021 12:45:28 +0200 (CEST) ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=JDO09jM26wFpRBwjqL5BtIQRipZw2ABoukPebnDQ3MYrW6PuBSo91rFXH/HrFHdSs3o4mit8f9Or1R+e9pK2+fGvVMBa7qmqq7keJril+NJ2F/9In6HXOuNcgd9eo8KWh/RJLVMkRdFpVkbnrbGAAe/+X1fyv90AWVSsRJXqOmJyMUi6U2dIdWvo/o2pgqnVa0HD3KfGnbuOXmW7Tz/KNK25TOPZSF4sJEfCT2r3lBKStIDA15WtYYBbzR66M5Y6EtfiYPUMgU6LSd8jHyGdK4WeepmP4AabNQZQZwJu1X6TVpxvNEcZ+8CoHBhCEK4kanzNeLvQ53mkOMlGi+OojA== 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-SenderADCheck; bh=WxHrD52cVkV+SZrvrsolbI7VZVYOXfLZWoMPFtr0pAM=; b=PoB1QhKYKBaDtZVpR+rq6JMikgQP3j3+FCWyy6Gi5WMLIDeq3O7Kpoonr+X+cQPMy6DTTnviH3zHUIdpjI7M0boVqrl9yiILekgipYxxkApU5izXsvOF6GHoQ15Lyk9hHr8Bu0XC34PV2tRsBwXJQPUMpYgKmIt4EGNgc/o8heuxY3+SrT5bx7lZWcV/HCoBkn8ho5Bz32VeMW6aKEdfNO96qQnoRlNOIRtFvsnL21ioI2WhymHLvzVtNERQQ++xl+F3nYsw5pdf3U9q1FCTFuE98jW4GyUOl+I7sQYh7uom7fXSQuAWGron93/zO6Qk3xaZT2XNF2MAhEWRSuR4iw== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass (sender ip is 216.228.112.34) smtp.rcpttodomain=intel.com smtp.mailfrom=nvidia.com; dmarc=pass (p=none sp=none pct=100) action=none header.from=nvidia.com; dkim=none (message not signed); arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=Nvidia.com; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=WxHrD52cVkV+SZrvrsolbI7VZVYOXfLZWoMPFtr0pAM=; b=f2TaGoJb8Sv2VLs2swH4kSWXYdWtiv+lgMwI7oim4oLJ+4zTmiRai0T5hxSqQl4vVpRDnIvbvsYT9iAQRnZ+ctixE0Z92QMbQt2d5t++VhRJCCrF+E6v5IMhPTs50AG4vZGdK/Bq0TUmqGorr6XSkJFMGISQugSpuCNAmp2SU/I4vOMytGS1XeoLDhtkUodUkQQfR3ZRUeG8DPBKipoMR5wsAK7sHdLBD7b2iH0/8uvPQnbYdGp/xtUurdvm4UH7jiiyFkW+xI0ouC9Tzwey5Ni+ENyDtPcQCBUb4tlVDzhkJ+6U3Sfj78wCSgH7tSwl+29ovOuntt07rKtpcP+cuA== Received: from DS7PR03CA0064.namprd03.prod.outlook.com (2603:10b6:5:3bb::9) by MN2PR12MB4567.namprd12.prod.outlook.com (2603:10b6:208:263::12) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.4065.23; Sun, 2 May 2021 10:45:27 +0000 Received: from DM6NAM11FT048.eop-nam11.prod.protection.outlook.com (2603:10b6:5:3bb:cafe::a7) by DS7PR03CA0064.outlook.office365.com (2603:10b6:5:3bb::9) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.4087.25 via Frontend Transport; Sun, 2 May 2021 10:45:27 +0000 X-MS-Exchange-Authentication-Results: spf=pass (sender IP is 216.228.112.34) smtp.mailfrom=nvidia.com; intel.com; dkim=none (message not signed) header.d=none;intel.com; dmarc=pass action=none header.from=nvidia.com; Received-SPF: Pass (protection.outlook.com: domain of nvidia.com designates 216.228.112.34 as permitted sender) receiver=protection.outlook.com; client-ip=216.228.112.34; helo=mail.nvidia.com; Received: from mail.nvidia.com (216.228.112.34) by DM6NAM11FT048.mail.protection.outlook.com (10.13.173.114) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384) id 15.20.4087.32 via Frontend Transport; Sun, 2 May 2021 10:45:26 +0000 Received: from nvidia.com (172.20.145.6) by HQMAIL107.nvidia.com (172.20.187.13) with Microsoft SMTP Server (TLS) id 15.0.1497.2; Sun, 2 May 2021 10:45:24 +0000 From: Matan Azrad To: CC: , , , , Xueming Li Date: Sun, 2 May 2021 13:45:10 +0300 Message-ID: <20210502104510.3102095-1-matan@nvidia.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <1616058798-303292-1-git-send-email-matan@nvidia.com> References: <1616058798-303292-1-git-send-email-matan@nvidia.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Type: text/plain X-Originating-IP: [172.20.145.6] X-ClientProxiedBy: HQMAIL101.nvidia.com (172.20.187.10) To HQMAIL107.nvidia.com (172.20.187.13) X-EOPAttributedMessage: 0 X-MS-PublicTrafficType: Email X-MS-Office365-Filtering-Correlation-Id: 094736f7-8d87-4049-1dad-08d90d576592 X-MS-TrafficTypeDiagnostic: MN2PR12MB4567: X-LD-Processed: 43083d15-7273-40c1-b7db-39efd9ccc17a,ExtAddr X-Microsoft-Antispam-PRVS: X-MS-Oob-TLC-OOBClassifiers: OLM:305; X-MS-Exchange-SenderADCheck: 1 X-Microsoft-Antispam: BCL:0; X-Microsoft-Antispam-Message-Info: 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 X-Forefront-Antispam-Report: CIP:216.228.112.34; CTRY:US; LANG:en; SCL:1; SRV:; IPV:NLI; SFV:NSPM; H:mail.nvidia.com; PTR:schybrid03.nvidia.com; CAT:NONE; SFS:(4636009)(39860400002)(376002)(136003)(396003)(346002)(46966006)(36840700001)(478600001)(82310400003)(316002)(356005)(8936002)(36756003)(8676002)(36906005)(7636003)(6666004)(54906003)(7696005)(82740400003)(30864003)(70586007)(2906002)(36860700001)(5660300002)(6286002)(55016002)(47076005)(4326008)(26005)(2616005)(86362001)(426003)(107886003)(186003)(16526019)(1076003)(83380400001)(336012)(6916009)(70206006); DIR:OUT; SFP:1101; X-OriginatorOrg: Nvidia.com X-MS-Exchange-CrossTenant-OriginalArrivalTime: 02 May 2021 10:45:26.5916 (UTC) X-MS-Exchange-CrossTenant-Network-Message-Id: 094736f7-8d87-4049-1dad-08d90d576592 X-MS-Exchange-CrossTenant-Id: 43083d15-7273-40c1-b7db-39efd9ccc17a X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=43083d15-7273-40c1-b7db-39efd9ccc17a; Ip=[216.228.112.34]; Helo=[mail.nvidia.com] X-MS-Exchange-CrossTenant-AuthSource: DM6NAM11FT048.eop-nam11.prod.protection.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Anonymous X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem X-MS-Exchange-Transport-CrossTenantHeadersStamped: MN2PR12MB4567 Subject: [dpdk-dev] [PATCH v2] vdpa/mlx5: improve interrupt management X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" The driver should notify the guest for each traffic burst detected by CQ polling. The CQ polling trigger is defined by `event_mode` device argument, either by busy polling on all the CQs or by blocked call to HW completion event using DevX channel. Also, the polling event modes can move to blocked call when the traffic rate is low. The current blocked call uses the EAL interrupt API suffering a lot of overhead in the API management and serve all the drivers and libraries using only single thread. Use blocking FD of the DevX channel in order to do blocked call directly by the DevX channel FD mechanism. Signed-off-by: Matan Azrad Acked-by: Xueming Li --- doc/guides/vdpadevs/mlx5.rst | 8 +- drivers/vdpa/mlx5/mlx5_vdpa.c | 8 +- drivers/vdpa/mlx5/mlx5_vdpa.h | 8 +- drivers/vdpa/mlx5/mlx5_vdpa_event.c | 304 +++++++++++----------------- 4 files changed, 130 insertions(+), 198 deletions(-) v2: Rebased on top of latest updates by Thomas. diff --git a/doc/guides/vdpadevs/mlx5.rst b/doc/guides/vdpadevs/mlx5.rst index 1f2ae6fb92..9b2f9f12c7 100644 --- a/doc/guides/vdpadevs/mlx5.rst +++ b/doc/guides/vdpadevs/mlx5.rst @@ -129,10 +129,10 @@ Driver options - ``no_traffic_time`` parameter [int] - A nonzero value defines the traffic off time, in seconds, that moves the - driver to no-traffic mode. In this mode the timer events are stopped and - interrupts are configured to the device in order to notify traffic for the - driver. Default value is 2s. + A nonzero value defines the traffic off time, in polling cycle time units, + that moves the driver to no-traffic mode. In this mode the polling is stopped + and interrupts are configured to the device in order to notify traffic for the + driver. Default value is 16. - ``event_core`` parameter [int] diff --git a/drivers/vdpa/mlx5/mlx5_vdpa.c b/drivers/vdpa/mlx5/mlx5_vdpa.c index 898e50f807..619e31d61c 100644 --- a/drivers/vdpa/mlx5/mlx5_vdpa.c +++ b/drivers/vdpa/mlx5/mlx5_vdpa.c @@ -44,7 +44,7 @@ #define MLX5_VDPA_MAX_RETRIES 20 #define MLX5_VDPA_USEC 1000 -#define MLX5_VDPA_DEFAULT_NO_TRAFFIC_TIME_S 2LLU +#define MLX5_VDPA_DEFAULT_NO_TRAFFIC_MAX 16LLU TAILQ_HEAD(mlx5_vdpa_privs, mlx5_vdpa_priv) priv_list = TAILQ_HEAD_INITIALIZER(priv_list); @@ -632,7 +632,7 @@ mlx5_vdpa_args_check_handler(const char *key, const char *val, void *opaque) } else if (strcmp(key, "event_us") == 0) { priv->event_us = (uint32_t)tmp; } else if (strcmp(key, "no_traffic_time") == 0) { - priv->no_traffic_time_s = (uint32_t)tmp; + priv->no_traffic_max = (uint32_t)tmp; } else if (strcmp(key, "event_core") == 0) { if (tmp >= (unsigned long)n_cores) DRV_LOG(WARNING, "Invalid event_core %s.", val); @@ -658,7 +658,7 @@ mlx5_vdpa_config_get(struct rte_devargs *devargs, struct mlx5_vdpa_priv *priv) priv->event_mode = MLX5_VDPA_EVENT_MODE_FIXED_TIMER; priv->event_us = 0; priv->event_core = -1; - priv->no_traffic_time_s = MLX5_VDPA_DEFAULT_NO_TRAFFIC_TIME_S; + priv->no_traffic_max = MLX5_VDPA_DEFAULT_NO_TRAFFIC_MAX; if (devargs == NULL) return; kvlist = rte_kvargs_parse(devargs->args, NULL); @@ -671,7 +671,7 @@ mlx5_vdpa_config_get(struct rte_devargs *devargs, struct mlx5_vdpa_priv *priv) priv->event_us = MLX5_VDPA_DEFAULT_TIMER_STEP_US; DRV_LOG(DEBUG, "event mode is %d.", priv->event_mode); DRV_LOG(DEBUG, "event_us is %u us.", priv->event_us); - DRV_LOG(DEBUG, "no traffic time is %u s.", priv->no_traffic_time_s); + DRV_LOG(DEBUG, "no traffic max is %u.", priv->no_traffic_max); } /** diff --git a/drivers/vdpa/mlx5/mlx5_vdpa.h b/drivers/vdpa/mlx5/mlx5_vdpa.h index d93b430c97..722c72b65e 100644 --- a/drivers/vdpa/mlx5/mlx5_vdpa.h +++ b/drivers/vdpa/mlx5/mlx5_vdpa.h @@ -120,16 +120,13 @@ struct mlx5_vdpa_priv { TAILQ_ENTRY(mlx5_vdpa_priv) next; uint8_t configured; pthread_mutex_t vq_config_lock; - uint64_t last_traffic_tic; + uint64_t no_traffic_counter; pthread_t timer_tid; - pthread_mutex_t timer_lock; - pthread_cond_t timer_cond; - volatile uint8_t timer_on; int event_mode; int event_core; /* Event thread cpu affinity core. */ uint32_t event_us; uint32_t timer_delay_us; - uint32_t no_traffic_time_s; + uint32_t no_traffic_max; uint8_t hw_latency_mode; /* Hardware CQ moderation mode. */ uint16_t hw_max_latency_us; /* Hardware CQ moderation period in usec. */ uint16_t hw_max_pending_comp; /* Hardware CQ moderation counter. */ @@ -146,7 +143,6 @@ struct mlx5_vdpa_priv { struct mlx5dv_devx_event_channel *eventc; struct mlx5dv_devx_event_channel *err_chnl; struct mlx5dv_devx_uar *uar; - struct rte_intr_handle intr_handle; struct rte_intr_handle err_intr_handle; struct mlx5_devx_obj *td; struct mlx5_devx_obj *tiss[16]; /* TIS list for each LAG port. */ diff --git a/drivers/vdpa/mlx5/mlx5_vdpa_event.c b/drivers/vdpa/mlx5/mlx5_vdpa_event.c index ca6e985336..88f6a4256d 100644 --- a/drivers/vdpa/mlx5/mlx5_vdpa_event.c +++ b/drivers/vdpa/mlx5/mlx5_vdpa_event.c @@ -36,17 +36,6 @@ mlx5_vdpa_event_qp_global_release(struct mlx5_vdpa_priv *priv) } #ifdef HAVE_IBV_DEVX_EVENT if (priv->eventc) { - union { - struct mlx5dv_devx_async_event_hdr event_resp; - uint8_t buf[sizeof(struct mlx5dv_devx_async_event_hdr) - + 128]; - } out; - - /* Clean all pending events. */ - while (mlx5_glue->devx_get_event(priv->eventc, &out.event_resp, - sizeof(out.buf)) >= - (ssize_t)sizeof(out.event_resp.cookie)) - ; mlx5_os_devx_destroy_event_channel(priv->eventc); priv->eventc = NULL; } @@ -57,8 +46,6 @@ mlx5_vdpa_event_qp_global_release(struct mlx5_vdpa_priv *priv) static int mlx5_vdpa_event_qp_global_prepare(struct mlx5_vdpa_priv *priv) { - int flags, ret; - if (priv->eventc) return 0; priv->eventc = mlx5_os_devx_create_event_channel(priv->ctx, @@ -69,12 +56,6 @@ mlx5_vdpa_event_qp_global_prepare(struct mlx5_vdpa_priv *priv) rte_errno); goto error; } - flags = fcntl(priv->eventc->fd, F_GETFL); - ret = fcntl(priv->eventc->fd, F_SETFL, flags | O_NONBLOCK); - if (ret) { - DRV_LOG(ERR, "Failed to change event channel FD."); - goto error; - } /* * This PMD always claims the write memory barrier on UAR * registers writings, it is safe to allocate UAR with any @@ -238,122 +219,112 @@ mlx5_vdpa_timer_sleep(struct mlx5_vdpa_priv *priv, uint32_t max) sched_yield(); } -static void * -mlx5_vdpa_poll_handle(void *arg) +/* Notify virtio device for specific virtq new traffic. */ +static uint32_t +mlx5_vdpa_queue_complete(struct mlx5_vdpa_cq *cq) { - struct mlx5_vdpa_priv *priv = arg; - int i; - struct mlx5_vdpa_cq *cq; - uint32_t max; - uint64_t current_tic; - - pthread_mutex_lock(&priv->timer_lock); - while (!priv->timer_on) - pthread_cond_wait(&priv->timer_cond, &priv->timer_lock); - pthread_mutex_unlock(&priv->timer_lock); - priv->timer_delay_us = priv->event_mode == - MLX5_VDPA_EVENT_MODE_DYNAMIC_TIMER ? - MLX5_VDPA_DEFAULT_TIMER_DELAY_US : - priv->event_us; - while (1) { - max = 0; - pthread_mutex_lock(&priv->vq_config_lock); - for (i = 0; i < priv->nr_virtqs; i++) { - cq = &priv->virtqs[i].eqp.cq; - if (cq->cq_obj.cq && !cq->armed) { - uint32_t comp = mlx5_vdpa_cq_poll(cq); - - if (comp) { - /* Notify guest for descs consuming. */ - if (cq->callfd != -1) - eventfd_write(cq->callfd, - (eventfd_t)1); - if (comp > max) - max = comp; - } - } + uint32_t comp = 0; + + if (cq->cq_obj.cq) { + comp = mlx5_vdpa_cq_poll(cq); + if (comp) { + if (cq->callfd != -1) + eventfd_write(cq->callfd, (eventfd_t)1); + cq->armed = 0; } - current_tic = rte_rdtsc(); - if (!max) { - /* No traffic ? stop timer and load interrupts. */ - if (current_tic - priv->last_traffic_tic >= - rte_get_timer_hz() * priv->no_traffic_time_s) { - DRV_LOG(DEBUG, "Device %s traffic was stopped.", - priv->vdev->device->name); - mlx5_vdpa_arm_all_cqs(priv); - pthread_mutex_unlock(&priv->vq_config_lock); - pthread_mutex_lock(&priv->timer_lock); - priv->timer_on = 0; - while (!priv->timer_on) - pthread_cond_wait(&priv->timer_cond, - &priv->timer_lock); - pthread_mutex_unlock(&priv->timer_lock); - priv->timer_delay_us = priv->event_mode == - MLX5_VDPA_EVENT_MODE_DYNAMIC_TIMER ? - MLX5_VDPA_DEFAULT_TIMER_DELAY_US : - priv->event_us; - continue; - } - } else { - priv->last_traffic_tic = current_tic; - } - pthread_mutex_unlock(&priv->vq_config_lock); - mlx5_vdpa_timer_sleep(priv, max); } - return NULL; + return comp; } -static void -mlx5_vdpa_interrupt_handler(void *cb_arg) +/* Notify virtio device for any virtq new traffic. */ +static uint32_t +mlx5_vdpa_queues_complete(struct mlx5_vdpa_priv *priv) +{ + int i; + uint32_t max = 0; + + for (i = 0; i < priv->nr_virtqs; i++) { + struct mlx5_vdpa_cq *cq = &priv->virtqs[i].eqp.cq; + uint32_t comp = mlx5_vdpa_queue_complete(cq); + + if (comp > max) + max = comp; + } + return max; +} + +/* Wait on all CQs channel for completion event. */ +static struct mlx5_vdpa_cq * +mlx5_vdpa_event_wait(struct mlx5_vdpa_priv *priv __rte_unused) { - struct mlx5_vdpa_priv *priv = cb_arg; #ifdef HAVE_IBV_DEVX_EVENT union { struct mlx5dv_devx_async_event_hdr event_resp; uint8_t buf[sizeof(struct mlx5dv_devx_async_event_hdr) + 128]; } out; + int ret = mlx5_glue->devx_get_event(priv->eventc, &out.event_resp, + sizeof(out.buf)); - pthread_mutex_lock(&priv->vq_config_lock); - while (mlx5_glue->devx_get_event(priv->eventc, &out.event_resp, - sizeof(out.buf)) >= - (ssize_t)sizeof(out.event_resp.cookie)) { - struct mlx5_vdpa_cq *cq = (struct mlx5_vdpa_cq *) - (uintptr_t)out.event_resp.cookie; - struct mlx5_vdpa_event_qp *eqp = container_of(cq, - struct mlx5_vdpa_event_qp, cq); - struct mlx5_vdpa_virtq *virtq = container_of(eqp, - struct mlx5_vdpa_virtq, eqp); - - if (!virtq->enable) - continue; - mlx5_vdpa_cq_poll(cq); - /* Notify guest for descs consuming. */ - if (cq->callfd != -1) - eventfd_write(cq->callfd, (eventfd_t)1); - if (priv->event_mode == MLX5_VDPA_EVENT_MODE_ONLY_INTERRUPT) { - mlx5_vdpa_cq_arm(priv, cq); - pthread_mutex_unlock(&priv->vq_config_lock); - return; - } - /* Don't arm again - timer will take control. */ - DRV_LOG(DEBUG, "Device %s virtq %d cq %d event was captured." - " Timer is %s, cq ci is %u.\n", - priv->vdev->device->name, - (int)virtq->index, cq->cq_obj.cq->id, - priv->timer_on ? "on" : "off", cq->cq_ci); - cq->armed = 0; - } + if (ret >= 0) + return (struct mlx5_vdpa_cq *)(uintptr_t)out.event_resp.cookie; + DRV_LOG(INFO, "Got error in devx_get_event, ret = %d, errno = %d.", + ret, errno); #endif + return NULL; +} - /* Traffic detected: make sure timer is on. */ - priv->last_traffic_tic = rte_rdtsc(); - pthread_mutex_lock(&priv->timer_lock); - if (!priv->timer_on) { - priv->timer_on = 1; - pthread_cond_signal(&priv->timer_cond); +static void * +mlx5_vdpa_event_handle(void *arg) +{ + struct mlx5_vdpa_priv *priv = arg; + struct mlx5_vdpa_cq *cq; + uint32_t max; + + switch (priv->event_mode) { + case MLX5_VDPA_EVENT_MODE_DYNAMIC_TIMER: + case MLX5_VDPA_EVENT_MODE_FIXED_TIMER: + priv->timer_delay_us = priv->event_us; + while (1) { + pthread_mutex_lock(&priv->vq_config_lock); + max = mlx5_vdpa_queues_complete(priv); + if (max == 0 && priv->no_traffic_counter++ >= + priv->no_traffic_max) { + DRV_LOG(DEBUG, "Device %s traffic was stopped.", + priv->vdev->device->name); + mlx5_vdpa_arm_all_cqs(priv); + do { + pthread_mutex_unlock + (&priv->vq_config_lock); + cq = mlx5_vdpa_event_wait(priv); + pthread_mutex_lock + (&priv->vq_config_lock); + if (cq == NULL || + mlx5_vdpa_queue_complete(cq) > 0) + break; + } while (1); + priv->timer_delay_us = priv->event_us; + priv->no_traffic_counter = 0; + } else if (max != 0) { + priv->no_traffic_counter = 0; + } + pthread_mutex_unlock(&priv->vq_config_lock); + mlx5_vdpa_timer_sleep(priv, max); + } + return NULL; + case MLX5_VDPA_EVENT_MODE_ONLY_INTERRUPT: + do { + cq = mlx5_vdpa_event_wait(priv); + if (cq != NULL) { + pthread_mutex_lock(&priv->vq_config_lock); + if (mlx5_vdpa_queue_complete(cq) > 0) + mlx5_vdpa_cq_arm(priv, cq); + pthread_mutex_unlock(&priv->vq_config_lock); + } + } while (1); + return NULL; + default: + return NULL; } - pthread_mutex_unlock(&priv->timer_lock); - pthread_mutex_unlock(&priv->vq_config_lock); } static void @@ -511,80 +482,45 @@ mlx5_vdpa_cqe_event_setup(struct mlx5_vdpa_priv *priv) if (!priv->eventc) /* All virtqs are in poll mode. */ return 0; - if (priv->event_mode != MLX5_VDPA_EVENT_MODE_ONLY_INTERRUPT) { - pthread_mutex_init(&priv->timer_lock, NULL); - pthread_cond_init(&priv->timer_cond, NULL); - priv->timer_on = 0; - pthread_attr_init(&attr); - ret = pthread_attr_setschedpolicy(&attr, SCHED_RR); - if (ret) { - DRV_LOG(ERR, "Failed to set thread sched policy = RR."); - return -1; - } - ret = pthread_attr_setschedparam(&attr, &sp); - if (ret) { - DRV_LOG(ERR, "Failed to set thread priority."); - return -1; - } - ret = pthread_create(&priv->timer_tid, &attr, - mlx5_vdpa_poll_handle, (void *)priv); - if (ret) { - DRV_LOG(ERR, "Failed to create timer thread."); - return -1; - } - CPU_ZERO(&cpuset); - if (priv->event_core != -1) - CPU_SET(priv->event_core, &cpuset); - else - cpuset = rte_lcore_cpuset(rte_get_main_lcore()); - ret = pthread_setaffinity_np(priv->timer_tid, - sizeof(cpuset), &cpuset); - if (ret) { - DRV_LOG(ERR, "Failed to set thread affinity."); - goto error; - } - snprintf(name, sizeof(name), "vDPA-mlx5-%d", priv->vid); - ret = rte_thread_setname(priv->timer_tid, name); - if (ret) { - DRV_LOG(ERR, "Failed to set timer thread name."); - return -1; - } + pthread_attr_init(&attr); + ret = pthread_attr_setschedpolicy(&attr, SCHED_RR); + if (ret) { + DRV_LOG(ERR, "Failed to set thread sched policy = RR."); + return -1; } - priv->intr_handle.fd = priv->eventc->fd; - priv->intr_handle.type = RTE_INTR_HANDLE_EXT; - if (rte_intr_callback_register(&priv->intr_handle, - mlx5_vdpa_interrupt_handler, priv)) { - priv->intr_handle.fd = 0; - DRV_LOG(ERR, "Failed to register CQE interrupt %d.", rte_errno); - goto error; + ret = pthread_attr_setschedparam(&attr, &sp); + if (ret) { + DRV_LOG(ERR, "Failed to set thread priority."); + return -1; } + ret = pthread_create(&priv->timer_tid, &attr, mlx5_vdpa_event_handle, + (void *)priv); + if (ret) { + DRV_LOG(ERR, "Failed to create timer thread."); + return -1; + } + CPU_ZERO(&cpuset); + if (priv->event_core != -1) + CPU_SET(priv->event_core, &cpuset); + else + cpuset = rte_lcore_cpuset(rte_get_main_lcore()); + ret = pthread_setaffinity_np(priv->timer_tid, sizeof(cpuset), &cpuset); + if (ret) { + DRV_LOG(ERR, "Failed to set thread affinity."); + return -1; + } + snprintf(name, sizeof(name), "vDPA-mlx5-%d", priv->vid); + ret = rte_thread_setname(priv->timer_tid, name); + if (ret) + DRV_LOG(DEBUG, "Cannot set timer thread name."); return 0; -error: - mlx5_vdpa_cqe_event_unset(priv); - return -1; } void mlx5_vdpa_cqe_event_unset(struct mlx5_vdpa_priv *priv) { - int retries = MLX5_VDPA_INTR_RETRIES; - int ret = -EAGAIN; void *status; - if (priv->intr_handle.fd) { - while (retries-- && ret == -EAGAIN) { - ret = rte_intr_callback_unregister(&priv->intr_handle, - mlx5_vdpa_interrupt_handler, - priv); - if (ret == -EAGAIN) { - DRV_LOG(DEBUG, "Try again to unregister fd %d " - "of CQ interrupt, retries = %d.", - priv->intr_handle.fd, retries); - rte_pause(); - } - } - memset(&priv->intr_handle, 0, sizeof(priv->intr_handle)); - } if (priv->timer_tid) { pthread_cancel(priv->timer_tid); pthread_join(priv->timer_tid, &status); -- 2.25.1