From patchwork Wed Aug 11 15:11:28 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Gal Pressman X-Patchwork-Id: 12431399 X-Patchwork-Delegate: jgg@ziepe.ca Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-19.5 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER,INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id D50B8C4320A for ; Wed, 11 Aug 2021 15:12:50 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id B87BD60FC3 for ; Wed, 11 Aug 2021 15:12:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232748AbhHKPNN (ORCPT ); Wed, 11 Aug 2021 11:13:13 -0400 Received: from smtp-fw-2101.amazon.com ([72.21.196.25]:4605 "EHLO smtp-fw-2101.amazon.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232878AbhHKPNK (ORCPT ); Wed, 11 Aug 2021 11:13:10 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1628694768; x=1660230768; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=0IxpAeJmDs7VAznGjulWQvvFCJXpZa8y4P5kq9CSImI=; b=wIG0lcBLkR0gjknQbvN5Nx9hAz0aPrBDn2rqQuGjjvI8sh3UQGZAGWF+ 2xbJFtGFA4oBp+XaPOvas6e0J0JpZULfCKtqFnxYoJGrqebGuGCgaZ71x bxd2St4J6b94jGY4fubUqgk7VaYUnoif5ak2Cywk6p+gKUehvG+JdIs/y A=; X-IronPort-AV: E=Sophos;i="5.84,313,1620691200"; d="scan'208";a="128755113" Received: from iad12-co-svc-p1-lb1-vlan2.amazon.com (HELO email-inbound-relay-1d-74cf8b49.us-east-1.amazon.com) ([10.43.8.2]) by smtp-border-fw-2101.iad2.amazon.com with ESMTP; 11 Aug 2021 15:12:38 +0000 Received: from EX13D13EUB002.ant.amazon.com (iad55-ws-svc-p15-lb9-vlan3.iad.amazon.com [10.40.159.166]) by email-inbound-relay-1d-74cf8b49.us-east-1.amazon.com (Postfix) with ESMTPS id A942CC01B1; Wed, 11 Aug 2021 15:12:35 +0000 (UTC) Received: from EX13MTAUEA002.ant.amazon.com (10.43.61.77) by EX13D13EUB002.ant.amazon.com (10.43.166.205) with Microsoft SMTP Server (TLS) id 15.0.1497.23; Wed, 11 Aug 2021 15:12:34 +0000 Received: from 8c85908914bf.ant.amazon.com.com (10.1.212.21) by mail-relay.amazon.com (10.43.61.169) with Microsoft SMTP Server id 15.0.1497.23 via Frontend Transport; Wed, 11 Aug 2021 15:12:31 +0000 From: Gal Pressman To: Jason Gunthorpe , Doug Ledford CC: , Alexander Matushevsky , Gal Pressman , Firas JahJah , Yossi Leybovich , Steve Wise , Jason Gunthorpe , Shiraz Saleem Subject: [PATCH for-next 1/4] RDMA/efa: Free IRQ vectors on error flow Date: Wed, 11 Aug 2021 18:11:28 +0300 Message-ID: <20210811151131.39138-2-galpress@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20210811151131.39138-1-galpress@amazon.com> References: <20210811151131.39138-1-galpress@amazon.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org Make sure to free the IRQ vectors in case the allocation doesn't return the expected number of IRQs. Fixes: b7f5e880f377 ("RDMA/efa: Add the efa module") Reviewed-by: Firas JahJah Reviewed-by: Yossi Leybovich Signed-off-by: Gal Pressman --- drivers/infiniband/hw/efa/efa_main.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/infiniband/hw/efa/efa_main.c b/drivers/infiniband/hw/efa/efa_main.c index 997947d77de6..42f9ac3f586e 100644 --- a/drivers/infiniband/hw/efa/efa_main.c +++ b/drivers/infiniband/hw/efa/efa_main.c @@ -358,6 +358,7 @@ static int efa_enable_msix(struct efa_dev *dev) } if (irq_num != msix_vecs) { + efa_disable_msix(dev); dev_err(&dev->pdev->dev, "Allocated %d MSI-X (out of %d requested)\n", irq_num, msix_vecs); From patchwork Wed Aug 11 15:11:29 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Gal Pressman X-Patchwork-Id: 12431397 X-Patchwork-Delegate: jgg@ziepe.ca Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-19.5 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER,INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 8B62EC432BE for ; Wed, 11 Aug 2021 15:12:50 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 6716460FC3 for ; Wed, 11 Aug 2021 15:12:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232894AbhHKPNN (ORCPT ); Wed, 11 Aug 2021 11:13:13 -0400 Received: from smtp-fw-6001.amazon.com ([52.95.48.154]:23824 "EHLO smtp-fw-6001.amazon.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232906AbhHKPNL (ORCPT ); Wed, 11 Aug 2021 11:13:11 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1628694768; x=1660230768; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=PUilMmo0sQUNdAun7MkBkFo03LtjLFGU6kVPIhBZT44=; b=VTj7KG+NVBvdzewM3USzUohHUK/PhDmss7Lr5CxuILX831ovFDVeWMYA xtmjIm3q57KNxLsES6jvrSZUcB3jhYIXJ0FtlbvRU7RbsRdL8HlxO5G2X ltAD5H9BHqBxLtZlO3EG/wF9O/H82Hs1jy9YWuW1RYEJjyOiXtsVqUXe6 c=; X-IronPort-AV: E=Sophos;i="5.84,313,1620691200"; d="scan'208";a="133210799" Received: from iad12-co-svc-p1-lb1-vlan3.amazon.com (HELO email-inbound-relay-1a-807d4a99.us-east-1.amazon.com) ([10.43.8.6]) by smtp-border-fw-6001.iad6.amazon.com with ESMTP; 11 Aug 2021 15:12:40 +0000 Received: from EX13D02EUC004.ant.amazon.com (iad55-ws-svc-p15-lb9-vlan3.iad.amazon.com [10.40.159.166]) by email-inbound-relay-1a-807d4a99.us-east-1.amazon.com (Postfix) with ESMTPS id 84640A07DC; Wed, 11 Aug 2021 15:12:38 +0000 (UTC) Received: from EX13MTAUEA002.ant.amazon.com (10.43.61.77) by EX13D02EUC004.ant.amazon.com (10.43.164.117) with Microsoft SMTP Server (TLS) id 15.0.1497.23; Wed, 11 Aug 2021 15:12:37 +0000 Received: from 8c85908914bf.ant.amazon.com.com (10.1.212.21) by mail-relay.amazon.com (10.43.61.169) with Microsoft SMTP Server id 15.0.1497.23 via Frontend Transport; Wed, 11 Aug 2021 15:12:35 +0000 From: Gal Pressman To: Jason Gunthorpe , Doug Ledford CC: , Alexander Matushevsky , Gal Pressman , Firas JahJah , Yossi Leybovich Subject: [PATCH for-next 2/4] RDMA/efa: Remove unused cpu field from irq struct Date: Wed, 11 Aug 2021 18:11:29 +0300 Message-ID: <20210811151131.39138-3-galpress@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20210811151131.39138-1-galpress@amazon.com> References: <20210811151131.39138-1-galpress@amazon.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org The cpu field in efa_irq struct is unused, remove it. Reviewed-by: Firas JahJah Reviewed-by: Yossi Leybovich Signed-off-by: Gal Pressman --- drivers/infiniband/hw/efa/efa.h | 3 +-- drivers/infiniband/hw/efa/efa_main.c | 1 - 2 files changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/efa/efa.h b/drivers/infiniband/hw/efa/efa.h index 1a1e60eee1dc..1148b308d3f6 100644 --- a/drivers/infiniband/hw/efa/efa.h +++ b/drivers/infiniband/hw/efa/efa.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause */ /* - * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved. + * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved. */ #ifndef _EFA_H_ @@ -27,7 +27,6 @@ struct efa_irq { irq_handler_t handler; void *data; - int cpu; u32 vector; cpumask_t affinity_hint_mask; char name[EFA_IRQNAME_SIZE]; diff --git a/drivers/infiniband/hw/efa/efa_main.c b/drivers/infiniband/hw/efa/efa_main.c index 42f9ac3f586e..5b2fcfe9be20 100644 --- a/drivers/infiniband/hw/efa/efa_main.c +++ b/drivers/infiniband/hw/efa/efa_main.c @@ -109,7 +109,6 @@ static void efa_setup_mgmnt_irq(struct efa_dev *dev) dev->admin_irq.vector = pci_irq_vector(dev->pdev, dev->admin_msix_vector_idx); cpu = cpumask_first(cpu_online_mask); - dev->admin_irq.cpu = cpu; cpumask_set_cpu(cpu, &dev->admin_irq.affinity_hint_mask); dev_info(&dev->pdev->dev, "Setup irq:0x%p vector:%d name:%s\n", From patchwork Wed Aug 11 15:11:30 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Gal Pressman X-Patchwork-Id: 12431401 X-Patchwork-Delegate: jgg@ziepe.ca Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-19.5 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER,INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 14765C4338F for ; Wed, 11 Aug 2021 15:12:52 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id EC16860FC3 for ; Wed, 11 Aug 2021 15:12:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232878AbhHKPNN (ORCPT ); Wed, 11 Aug 2021 11:13:13 -0400 Received: from smtp-fw-80007.amazon.com ([99.78.197.218]:25160 "EHLO smtp-fw-80007.amazon.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232814AbhHKPNN (ORCPT ); Wed, 11 Aug 2021 11:13:13 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1628694769; x=1660230769; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=HiWdvnOuicCk9FNTzMD9lpRsyTXUJRgnFPcs4QdDbaQ=; b=Z/6g52yQDtzddFdcJfNDoahoUlOUdSxXXzP92t48gt4ClFiYivacDTiE LbRVTvlWVw1CpiQ8SeY+bPNvpv0wTTv9HWhea7Lngnol5lqNv68EsANAB tjOUknfCJgGqrLfxXsFRgdUfsu4kLoO4EOHOB5rlyj/uLhRNV6dRpCemc 4=; X-IronPort-AV: E=Sophos;i="5.84,313,1620691200"; d="scan'208";a="18574463" Received: from pdx4-co-svc-p1-lb2-vlan2.amazon.com (HELO email-inbound-relay-2a-f14f4a47.us-west-2.amazon.com) ([10.25.36.210]) by smtp-border-fw-80007.pdx80.corp.amazon.com with ESMTP; 11 Aug 2021 15:12:42 +0000 Received: from EX13D19EUB001.ant.amazon.com (pdx1-ws-svc-p6-lb9-vlan2.pdx.amazon.com [10.236.137.194]) by email-inbound-relay-2a-f14f4a47.us-west-2.amazon.com (Postfix) with ESMTPS id 71095A256C; Wed, 11 Aug 2021 15:12:42 +0000 (UTC) Received: from EX13MTAUEA002.ant.amazon.com (10.43.61.77) by EX13D19EUB001.ant.amazon.com (10.43.166.229) with Microsoft SMTP Server (TLS) id 15.0.1497.23; Wed, 11 Aug 2021 15:12:41 +0000 Received: from 8c85908914bf.ant.amazon.com.com (10.1.212.21) by mail-relay.amazon.com (10.43.61.169) with Microsoft SMTP Server id 15.0.1497.23 via Frontend Transport; Wed, 11 Aug 2021 15:12:38 +0000 From: Gal Pressman To: Jason Gunthorpe , Doug Ledford CC: , Alexander Matushevsky , Gal Pressman , Firas JahJah , Yossi Leybovich Subject: [PATCH for-next 3/4] RDMA/efa: Rename vector field in efa_irq struct to irqn Date: Wed, 11 Aug 2021 18:11:30 +0300 Message-ID: <20210811151131.39138-4-galpress@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20210811151131.39138-1-galpress@amazon.com> References: <20210811151131.39138-1-galpress@amazon.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org The vector field naming is quite confusing, it is better referred to as irqn. Reviewed-by: Firas JahJah Reviewed-by: Yossi Leybovich Signed-off-by: Gal Pressman --- drivers/infiniband/hw/efa/efa.h | 2 +- drivers/infiniband/hw/efa/efa_main.c | 18 ++++++++---------- 2 files changed, 9 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/hw/efa/efa.h b/drivers/infiniband/hw/efa/efa.h index 1148b308d3f6..87b1dadeb7fe 100644 --- a/drivers/infiniband/hw/efa/efa.h +++ b/drivers/infiniband/hw/efa/efa.h @@ -27,7 +27,7 @@ struct efa_irq { irq_handler_t handler; void *data; - u32 vector; + u32 irqn; cpumask_t affinity_hint_mask; char name[EFA_IRQNAME_SIZE]; }; diff --git a/drivers/infiniband/hw/efa/efa_main.c b/drivers/infiniband/hw/efa/efa_main.c index 5b2fcfe9be20..417dea5f90cf 100644 --- a/drivers/infiniband/hw/efa/efa_main.c +++ b/drivers/infiniband/hw/efa/efa_main.c @@ -83,8 +83,7 @@ static int efa_request_mgmnt_irq(struct efa_dev *dev) int err; irq = &dev->admin_irq; - err = request_irq(irq->vector, irq->handler, 0, irq->name, - irq->data); + err = request_irq(irq->irqn, irq->handler, 0, irq->name, irq->data); if (err) { dev_err(&dev->pdev->dev, "Failed to request admin irq (%d)\n", err); @@ -92,8 +91,8 @@ static int efa_request_mgmnt_irq(struct efa_dev *dev) } dev_dbg(&dev->pdev->dev, "Set affinity hint of mgmnt irq to %*pbl (irq vector: %d)\n", - nr_cpumask_bits, &irq->affinity_hint_mask, irq->vector); - irq_set_affinity_hint(irq->vector, &irq->affinity_hint_mask); + nr_cpumask_bits, &irq->affinity_hint_mask, irq->irqn); + irq_set_affinity_hint(irq->irqn, &irq->affinity_hint_mask); return 0; } @@ -106,14 +105,13 @@ static void efa_setup_mgmnt_irq(struct efa_dev *dev) "efa-mgmnt@pci:%s", pci_name(dev->pdev)); dev->admin_irq.handler = efa_intr_msix_mgmnt; dev->admin_irq.data = dev; - dev->admin_irq.vector = + dev->admin_irq.irqn = pci_irq_vector(dev->pdev, dev->admin_msix_vector_idx); cpu = cpumask_first(cpu_online_mask); cpumask_set_cpu(cpu, &dev->admin_irq.affinity_hint_mask); - dev_info(&dev->pdev->dev, "Setup irq:0x%p vector:%d name:%s\n", - &dev->admin_irq, - dev->admin_irq.vector, + dev_info(&dev->pdev->dev, "Setup irq:%d name:%s\n", + dev->admin_irq.irqn, dev->admin_irq.name); } @@ -122,8 +120,8 @@ static void efa_free_mgmnt_irq(struct efa_dev *dev) struct efa_irq *irq; irq = &dev->admin_irq; - irq_set_affinity_hint(irq->vector, NULL); - free_irq(irq->vector, irq->data); + irq_set_affinity_hint(irq->irqn, NULL); + free_irq(irq->irqn, irq->data); } static int efa_set_mgmnt_irq(struct efa_dev *dev) From patchwork Wed Aug 11 15:11:31 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Gal Pressman X-Patchwork-Id: 12431403 X-Patchwork-Delegate: jgg@ziepe.ca Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-19.5 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER,INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 7439DC4320E for ; Wed, 11 Aug 2021 15:13:01 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 5AEE960F21 for ; Wed, 11 Aug 2021 15:13:01 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232814AbhHKPNW (ORCPT ); Wed, 11 Aug 2021 11:13:22 -0400 Received: from smtp-fw-80006.amazon.com ([99.78.197.217]:8505 "EHLO smtp-fw-80006.amazon.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232965AbhHKPNT (ORCPT ); Wed, 11 Aug 2021 11:13:19 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1628694776; x=1660230776; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=68mr5wZOwMhmCcTHBA0eYDs/Nan4Mesc22DKsd7T+Y8=; b=dtRKh2UtZUy7oqMiNNN0LlGx72r5BBo91STOZVEiIxnok2XhPPR+p7LF vla09RKIystgBiDRqP7nquXll2nkkdK9nkoZUgQXQ3iQedSePUfWrgCsi Zu/ske6P0r0LmID20jB5JuRNKwZLJfhtHqRg2bXaJIanlqqPnHS88EDOG o=; X-IronPort-AV: E=Sophos;i="5.84,313,1620691200"; d="scan'208";a="18583024" Received: from pdx4-co-svc-p1-lb2-vlan3.amazon.com (HELO email-inbound-relay-2a-6e2fc477.us-west-2.amazon.com) ([10.25.36.214]) by smtp-border-fw-80006.pdx80.corp.amazon.com with ESMTP; 11 Aug 2021 15:12:48 +0000 Received: from EX13D19EUB001.ant.amazon.com (pdx1-ws-svc-p6-lb9-vlan2.pdx.amazon.com [10.236.137.194]) by email-inbound-relay-2a-6e2fc477.us-west-2.amazon.com (Postfix) with ESMTPS id CC09DA05D3; Wed, 11 Aug 2021 15:12:47 +0000 (UTC) Received: from EX13MTAUEA002.ant.amazon.com (10.43.61.77) by EX13D19EUB001.ant.amazon.com (10.43.166.229) with Microsoft SMTP Server (TLS) id 15.0.1497.23; Wed, 11 Aug 2021 15:12:46 +0000 Received: from 8c85908914bf.ant.amazon.com.com (10.1.212.21) by mail-relay.amazon.com (10.43.61.169) with Microsoft SMTP Server id 15.0.1497.23 via Frontend Transport; Wed, 11 Aug 2021 15:12:43 +0000 From: Gal Pressman To: Jason Gunthorpe , Doug Ledford CC: , Alexander Matushevsky , Gal Pressman , Firas JahJah , Yossi Leybovich , Jason Gunthorpe Subject: [PATCH for-next 4/4] RDMA/efa: CQ notifications Date: Wed, 11 Aug 2021 18:11:31 +0300 Message-ID: <20210811151131.39138-5-galpress@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20210811151131.39138-1-galpress@amazon.com> References: <20210811151131.39138-1-galpress@amazon.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org This patch adds support for CQ notifications through the standard verbs api. In order to achieve that, a new event queue (EQ) object is introduced, which is in charge of reporting completion events to the driver. On driver load, EQs are allocated and their affinity is set to a single cpu. When a user app creates a CQ with a completion channel, the completion vector number is converted to a EQ number, which is in charge of reporting the CQ events. In addition, the CQ creation admin command now returns an offset for the CQ doorbell, which is mapped to the userspace provider and is used to arm the CQ when requested by the user. The EQs use a single doorbell (located on the registers BAR), which encodes the EQ number and arm as part of the doorbell value. The EQs are polled by the driver on each new EQE, and arm it when the poll is completed. Reviewed-by: Firas JahJah Reviewed-by: Yossi Leybovich Signed-off-by: Gal Pressman --- drivers/infiniband/hw/efa/efa.h | 16 +- .../infiniband/hw/efa/efa_admin_cmds_defs.h | 100 +++++++++- drivers/infiniband/hw/efa/efa_admin_defs.h | 41 ++++ drivers/infiniband/hw/efa/efa_com.c | 171 ++++++++++++++++ drivers/infiniband/hw/efa/efa_com.h | 38 +++- drivers/infiniband/hw/efa/efa_com_cmd.c | 35 +++- drivers/infiniband/hw/efa/efa_com_cmd.h | 10 +- drivers/infiniband/hw/efa/efa_main.c | 184 +++++++++++++++--- drivers/infiniband/hw/efa/efa_regs_defs.h | 7 +- drivers/infiniband/hw/efa/efa_verbs.c | 60 +++++- include/uapi/rdma/efa-abi.h | 18 +- 11 files changed, 625 insertions(+), 55 deletions(-) diff --git a/drivers/infiniband/hw/efa/efa.h b/drivers/infiniband/hw/efa/efa.h index 87b1dadeb7fe..e72bc4741ca1 100644 --- a/drivers/infiniband/hw/efa/efa.h +++ b/drivers/infiniband/hw/efa/efa.h @@ -20,14 +20,14 @@ #define EFA_IRQNAME_SIZE 40 -/* 1 for AENQ + ADMIN */ -#define EFA_NUM_MSIX_VEC 1 #define EFA_MGMNT_MSIX_VEC_IDX 0 +#define EFA_COMP_EQS_VEC_BASE 1 struct efa_irq { irq_handler_t handler; void *data; u32 irqn; + u32 vector; cpumask_t affinity_hint_mask; char name[EFA_IRQNAME_SIZE]; }; @@ -61,6 +61,12 @@ struct efa_dev { struct efa_irq admin_irq; struct efa_stats stats; + + /* Array of completion EQs */ + struct efa_eq *eqs; + unsigned int neqs; + + struct xarray cqs_xa; }; struct efa_ucontext { @@ -84,6 +90,7 @@ struct efa_cq { dma_addr_t dma_addr; void *cpu_addr; struct rdma_user_mmap_entry *mmap_entry; + struct rdma_user_mmap_entry *db_mmap_entry; size_t size; u16 cq_idx; }; @@ -116,6 +123,11 @@ struct efa_ah { u8 id[EFA_GID_SIZE]; }; +struct efa_eq { + struct efa_com_eq eeq; + struct efa_irq irq; +}; + int efa_query_device(struct ib_device *ibdev, struct ib_device_attr *props, struct ib_udata *udata); diff --git a/drivers/infiniband/hw/efa/efa_admin_cmds_defs.h b/drivers/infiniband/hw/efa/efa_admin_cmds_defs.h index fa38b34eddb8..0b0b93b529f3 100644 --- a/drivers/infiniband/hw/efa/efa_admin_cmds_defs.h +++ b/drivers/infiniband/hw/efa/efa_admin_cmds_defs.h @@ -28,7 +28,9 @@ enum efa_admin_aq_opcode { EFA_ADMIN_DEALLOC_PD = 15, EFA_ADMIN_ALLOC_UAR = 16, EFA_ADMIN_DEALLOC_UAR = 17, - EFA_ADMIN_MAX_OPCODE = 17, + EFA_ADMIN_CREATE_EQ = 18, + EFA_ADMIN_DESTROY_EQ = 19, + EFA_ADMIN_MAX_OPCODE = 19, }; enum efa_admin_aq_feature_id { @@ -38,6 +40,7 @@ enum efa_admin_aq_feature_id { EFA_ADMIN_QUEUE_ATTR = 4, EFA_ADMIN_HW_HINTS = 5, EFA_ADMIN_HOST_INFO = 6, + EFA_ADMIN_EVENT_QUEUE_ATTR = 7, }; /* QP transport type */ @@ -430,8 +433,8 @@ struct efa_admin_create_cq_cmd { /* * 4:0 : reserved5 - MBZ * 5 : interrupt_mode_enabled - if set, cq operates - * in interrupt mode (i.e. CQ events and MSI-X are - * generated), otherwise - polling + * in interrupt mode (i.e. CQ events and EQ elements + * are generated), otherwise - polling * 6 : virt - If set, ring base address is virtual * (IOVA returned by MR registration) * 7 : reserved6 - MBZ @@ -448,8 +451,11 @@ struct efa_admin_create_cq_cmd { /* completion queue depth in # of entries. must be power of 2 */ u16 cq_depth; - /* msix vector assigned to this cq */ - u32 msix_vector_idx; + /* EQ number assigned to this cq */ + u16 eqn; + + /* MBZ */ + u16 reserved; /* * CQ ring base address, virtual or physical depending on 'virt' @@ -480,6 +486,15 @@ struct efa_admin_create_cq_resp { /* actual cq depth in number of entries */ u16 cq_actual_depth; + + /* CQ doorbell address, as offset to PCIe DB BAR */ + u32 db_offset; + + /* + * 0 : db_valid - If set, doorbell offset is valid. + * Always set when interrupts are requested. + */ + u32 flags; }; struct efa_admin_destroy_cq_cmd { @@ -669,6 +684,17 @@ struct efa_admin_feature_queue_attr_desc { u16 max_tx_batch; }; +struct efa_admin_event_queue_attr_desc { + /* The maximum number of event queues supported */ + u32 max_eq; + + /* Maximum number of EQEs per Event Queue */ + u32 max_eq_depth; + + /* Supported events bitmask */ + u32 event_bitmask; +}; + struct efa_admin_feature_aenq_desc { /* bitmask for AENQ groups the device can report */ u32 supported_groups; @@ -727,6 +753,8 @@ struct efa_admin_get_feature_resp { struct efa_admin_feature_queue_attr_desc queue_attr; + struct efa_admin_event_queue_attr_desc event_queue_attr; + struct efa_admin_hw_hints hw_hints; } u; }; @@ -810,6 +838,60 @@ struct efa_admin_dealloc_uar_resp { struct efa_admin_acq_common_desc acq_common_desc; }; +struct efa_admin_create_eq_cmd { + struct efa_admin_aq_common_desc aq_common_descriptor; + + /* Size of the EQ in entries, must be power of 2 */ + u16 depth; + + /* MSI-X table entry index */ + u8 msix_vec; + + /* + * 4:0 : entry_size_words - size of EQ entry in + * 32-bit words + * 7:5 : reserved - MBZ + */ + u8 caps; + + /* EQ ring base address */ + struct efa_common_mem_addr ba; + + /* + * Enabled events on this EQ + * 0 : completion_events - Enable completion events + * 31:1 : reserved - MBZ + */ + u32 event_bitmask; + + /* MBZ */ + u32 reserved; +}; + +struct efa_admin_create_eq_resp { + struct efa_admin_acq_common_desc acq_common_desc; + + /* EQ number */ + u16 eqn; + + /* MBZ */ + u16 reserved; +}; + +struct efa_admin_destroy_eq_cmd { + struct efa_admin_aq_common_desc aq_common_descriptor; + + /* EQ number */ + u16 eqn; + + /* MBZ */ + u16 reserved; +}; + +struct efa_admin_destroy_eq_resp { + struct efa_admin_acq_common_desc acq_common_desc; +}; + /* asynchronous event notification groups */ enum efa_admin_aenq_group { EFA_ADMIN_FATAL_ERROR = 1, @@ -899,10 +981,18 @@ struct efa_admin_host_info { #define EFA_ADMIN_CREATE_CQ_CMD_VIRT_MASK BIT(6) #define EFA_ADMIN_CREATE_CQ_CMD_CQ_ENTRY_SIZE_WORDS_MASK GENMASK(4, 0) +/* create_cq_resp */ +#define EFA_ADMIN_CREATE_CQ_RESP_DB_VALID_MASK BIT(0) + /* feature_device_attr_desc */ #define EFA_ADMIN_FEATURE_DEVICE_ATTR_DESC_RDMA_READ_MASK BIT(0) #define EFA_ADMIN_FEATURE_DEVICE_ATTR_DESC_RNR_RETRY_MASK BIT(1) +/* create_eq_cmd */ +#define EFA_ADMIN_CREATE_EQ_CMD_ENTRY_SIZE_WORDS_MASK GENMASK(4, 0) +#define EFA_ADMIN_CREATE_EQ_CMD_VIRT_MASK BIT(6) +#define EFA_ADMIN_CREATE_EQ_CMD_COMPLETION_EVENTS_MASK BIT(0) + /* host_info */ #define EFA_ADMIN_HOST_INFO_DRIVER_MODULE_TYPE_MASK GENMASK(7, 0) #define EFA_ADMIN_HOST_INFO_DRIVER_SUB_MINOR_MASK GENMASK(15, 8) diff --git a/drivers/infiniband/hw/efa/efa_admin_defs.h b/drivers/infiniband/hw/efa/efa_admin_defs.h index 78ff9389ae25..83f20c38a840 100644 --- a/drivers/infiniband/hw/efa/efa_admin_defs.h +++ b/drivers/infiniband/hw/efa/efa_admin_defs.h @@ -118,6 +118,43 @@ struct efa_admin_aenq_entry { u32 inline_data_w4[12]; }; +enum efa_admin_eqe_event_type { + EFA_ADMIN_EQE_EVENT_TYPE_COMPLETION = 0, +}; + +/* Completion event */ +struct efa_admin_comp_event { + /* CQ number */ + u16 cqn; + + /* MBZ */ + u16 reserved; + + /* MBZ */ + u32 reserved2; +}; + +/* Event Queue Element */ +struct efa_admin_eqe { + /* + * 0 : phase + * 8:1 : event_type - Event type + * 31:9 : reserved - MBZ + */ + u32 common; + + /* MBZ */ + u32 reserved; + + union { + /* Event data */ + u32 event_data[2]; + + /* Completion Event */ + struct efa_admin_comp_event comp_event; + } u; +}; + /* aq_common_desc */ #define EFA_ADMIN_AQ_COMMON_DESC_COMMAND_ID_MASK GENMASK(11, 0) #define EFA_ADMIN_AQ_COMMON_DESC_PHASE_MASK BIT(0) @@ -131,4 +168,8 @@ struct efa_admin_aenq_entry { /* aenq_common_desc */ #define EFA_ADMIN_AENQ_COMMON_DESC_PHASE_MASK BIT(0) +/* eqe */ +#define EFA_ADMIN_EQE_PHASE_MASK BIT(0) +#define EFA_ADMIN_EQE_EVENT_TYPE_MASK GENMASK(8, 1) + #endif /* _EFA_ADMIN_H_ */ diff --git a/drivers/infiniband/hw/efa/efa_com.c b/drivers/infiniband/hw/efa/efa_com.c index 0d523ad736c7..c00c7f526067 100644 --- a/drivers/infiniband/hw/efa/efa_com.c +++ b/drivers/infiniband/hw/efa/efa_com.c @@ -56,11 +56,19 @@ static const char *efa_com_cmd_str(u8 cmd) EFA_CMD_STR_CASE(DEALLOC_PD); EFA_CMD_STR_CASE(ALLOC_UAR); EFA_CMD_STR_CASE(DEALLOC_UAR); + EFA_CMD_STR_CASE(CREATE_EQ); + EFA_CMD_STR_CASE(DESTROY_EQ); default: return "unknown command opcode"; } #undef EFA_CMD_STR_CASE } +void efa_com_set_dma_addr(dma_addr_t addr, u32 *addr_high, u32 *addr_low) +{ + *addr_low = lower_32_bits(addr); + *addr_high = upper_32_bits(addr); +} + static u32 efa_com_reg_read32(struct efa_com_dev *edev, u16 offset) { struct efa_com_mmio_read *mmio_read = &edev->mmio_read; @@ -1081,3 +1089,166 @@ int efa_com_dev_reset(struct efa_com_dev *edev, return 0; } + +static int efa_com_create_eq(struct efa_com_dev *edev, + struct efa_com_create_eq_params *params, + struct efa_com_create_eq_result *result) +{ + struct efa_com_admin_queue *aq = &edev->aq; + struct efa_admin_create_eq_resp resp = {}; + struct efa_admin_create_eq_cmd cmd = {}; + int err; + + cmd.aq_common_descriptor.opcode = EFA_ADMIN_CREATE_EQ; + EFA_SET(&cmd.caps, EFA_ADMIN_CREATE_EQ_CMD_ENTRY_SIZE_WORDS, + params->entry_size_in_bytes / 4); + cmd.depth = params->depth; + cmd.event_bitmask = params->event_bitmask; + cmd.msix_vec = params->msix_vec; + + efa_com_set_dma_addr(params->dma_addr, &cmd.ba.mem_addr_high, + &cmd.ba.mem_addr_low); + + err = efa_com_cmd_exec(aq, + (struct efa_admin_aq_entry *)&cmd, + sizeof(cmd), + (struct efa_admin_acq_entry *)&resp, + sizeof(resp)); + if (err) { + ibdev_err_ratelimited(edev->efa_dev, + "Failed to create eq[%d]\n", err); + return err; + } + + result->eqn = resp.eqn; + + return 0; +} + +static int efa_com_destroy_eq(struct efa_com_dev *edev, + struct efa_com_destroy_eq_params *params) +{ + struct efa_com_admin_queue *aq = &edev->aq; + struct efa_admin_destroy_eq_resp resp = {}; + struct efa_admin_destroy_eq_cmd cmd = {}; + int err; + + cmd.aq_common_descriptor.opcode = EFA_ADMIN_DESTROY_EQ; + cmd.eqn = params->eqn; + + err = efa_com_cmd_exec(aq, + (struct efa_admin_aq_entry *)&cmd, + sizeof(cmd), + (struct efa_admin_acq_entry *)&resp, + sizeof(resp)); + + if (err) { + ibdev_err_ratelimited(edev->efa_dev, + "Failed to destroy EQ-%u [%d]\n", cmd.eqn, + err); + return err; + } + + return 0; +} + +static void efa_com_arm_eq(struct efa_com_dev *edev, struct efa_com_eq *eeq) +{ + u32 val = 0; + + EFA_SET(&val, EFA_REGS_EQ_DB_EQN, eeq->eqn); + EFA_SET(&val, EFA_REGS_EQ_DB_ARM, 1); + + writel(val, edev->reg_bar + EFA_REGS_EQ_DB_OFF); +} + +void efa_com_eq_comp_intr_handler(struct efa_com_dev *edev, + struct efa_com_eq *eeq) +{ + struct efa_admin_eqe *eqe; + u32 processed = 0; + u8 phase; + u32 ci; + + ci = eeq->cc & (eeq->depth - 1); + phase = eeq->phase; + eqe = &eeq->eqes[ci]; + + /* Go over all the events */ + while ((READ_ONCE(eqe->common) & EFA_ADMIN_EQE_PHASE_MASK) == phase) { + /* + * Do not read the rest of the completion entry before the + * phase bit was validated + */ + dma_rmb(); + + eeq->cb(eeq, eqe); + + /* Get next event entry */ + ci++; + processed++; + + if (ci == eeq->depth) { + ci = 0; + phase = !phase; + } + + eqe = &eeq->eqes[ci]; + } + + eeq->cc += processed; + eeq->phase = phase; + efa_com_arm_eq(eeq->edev, eeq); +} + +int efa_com_eq_destroy(struct efa_com_dev *edev, struct efa_com_eq *eeq) +{ + struct efa_com_destroy_eq_params params = { + .eqn = eeq->eqn, + }; + + efa_com_destroy_eq(edev, ¶ms); + dma_free_coherent(edev->dmadev, eeq->depth * sizeof(*eeq->eqes), + eeq->eqes, eeq->dma_addr); + + return 0; +} + +int efa_com_eq_init(struct efa_com_dev *edev, struct efa_com_eq *eeq, + efa_eqe_handler cb, u16 depth, u8 msix_vec) +{ + struct efa_com_create_eq_params params = {}; + struct efa_com_create_eq_result result = {}; + int err; + + params.depth = depth; + params.entry_size_in_bytes = sizeof(*eeq->eqes); + EFA_SET(¶ms.event_bitmask, + EFA_ADMIN_CREATE_EQ_CMD_COMPLETION_EVENTS, 1); + params.msix_vec = msix_vec; + + eeq->eqes = dma_alloc_coherent(edev->dmadev, + params.depth * sizeof(*eeq->eqes), + ¶ms.dma_addr, GFP_KERNEL); + if (!eeq->eqes) + return -ENOMEM; + + err = efa_com_create_eq(edev, ¶ms, &result); + if (err) + goto err_free_coherent; + + eeq->eqn = result.eqn; + eeq->edev = edev; + eeq->dma_addr = params.dma_addr; + eeq->phase = 1; + eeq->depth = params.depth; + eeq->cb = cb; + efa_com_arm_eq(edev, eeq); + + return 0; + +err_free_coherent: + dma_free_coherent(edev->dmadev, params.depth * sizeof(*eeq->eqes), + eeq->eqes, params.dma_addr); + return err; +} diff --git a/drivers/infiniband/hw/efa/efa_com.h b/drivers/infiniband/hw/efa/efa_com.h index 5e4c88877ddb..0fe241cd48e7 100644 --- a/drivers/infiniband/hw/efa/efa_com.h +++ b/drivers/infiniband/hw/efa/efa_com.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause */ /* - * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved. + * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved. */ #ifndef _EFA_COM_H_ @@ -80,6 +80,9 @@ struct efa_com_admin_queue { }; struct efa_aenq_handlers; +struct efa_com_eq; +typedef void (*efa_eqe_handler)(struct efa_com_eq *eeq, + struct efa_admin_eqe *eqe); struct efa_com_aenq { struct efa_admin_aenq_entry *entries; @@ -112,6 +115,33 @@ struct efa_com_dev { struct efa_com_mmio_read mmio_read; }; +struct efa_com_eq { + struct efa_com_dev *edev; + struct efa_admin_eqe *eqes; + dma_addr_t dma_addr; + u32 cc; /* Consumer counter */ + u16 eqn; + u16 depth; + u8 phase; + efa_eqe_handler cb; +}; + +struct efa_com_create_eq_params { + dma_addr_t dma_addr; + u32 event_bitmask; + u16 depth; + u8 entry_size_in_bytes; + u8 msix_vec; +}; + +struct efa_com_create_eq_result { + u16 eqn; +}; + +struct efa_com_destroy_eq_params { + u16 eqn; +}; + typedef void (*efa_aenq_handler)(void *data, struct efa_admin_aenq_entry *aenq_e); @@ -121,9 +151,13 @@ struct efa_aenq_handlers { efa_aenq_handler unimplemented_handler; }; +void efa_com_set_dma_addr(dma_addr_t addr, u32 *addr_high, u32 *addr_low); int efa_com_admin_init(struct efa_com_dev *edev, struct efa_aenq_handlers *aenq_handlers); void efa_com_admin_destroy(struct efa_com_dev *edev); +int efa_com_eq_init(struct efa_com_dev *edev, struct efa_com_eq *eeq, + efa_eqe_handler cb, u16 depth, u8 msix_vec); +int efa_com_eq_destroy(struct efa_com_dev *edev, struct efa_com_eq *eeq); int efa_com_dev_reset(struct efa_com_dev *edev, enum efa_regs_reset_reason_types reset_reason); void efa_com_set_admin_polling_mode(struct efa_com_dev *edev, bool polling); @@ -140,5 +174,7 @@ int efa_com_cmd_exec(struct efa_com_admin_queue *aq, struct efa_admin_acq_entry *comp, size_t comp_size); void efa_com_aenq_intr_handler(struct efa_com_dev *edev, void *data); +void efa_com_eq_comp_intr_handler(struct efa_com_dev *edev, + struct efa_com_eq *eeq); #endif /* _EFA_COM_H_ */ diff --git a/drivers/infiniband/hw/efa/efa_com_cmd.c b/drivers/infiniband/hw/efa/efa_com_cmd.c index f752ef64159c..fb405da4e1db 100644 --- a/drivers/infiniband/hw/efa/efa_com_cmd.c +++ b/drivers/infiniband/hw/efa/efa_com_cmd.c @@ -1,17 +1,11 @@ // SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause /* - * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved. + * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved. */ #include "efa_com.h" #include "efa_com_cmd.h" -void efa_com_set_dma_addr(dma_addr_t addr, u32 *addr_high, u32 *addr_low) -{ - *addr_low = lower_32_bits(addr); - *addr_high = upper_32_bits(addr); -} - int efa_com_create_qp(struct efa_com_dev *edev, struct efa_com_create_qp_params *params, struct efa_com_create_qp_result *res) @@ -157,7 +151,7 @@ int efa_com_create_cq(struct efa_com_dev *edev, struct efa_com_create_cq_params *params, struct efa_com_create_cq_result *result) { - struct efa_admin_create_cq_resp cmd_completion; + struct efa_admin_create_cq_resp cmd_completion = {}; struct efa_admin_create_cq_cmd create_cmd = {}; struct efa_com_admin_queue *aq = &edev->aq; int err; @@ -169,6 +163,11 @@ int efa_com_create_cq(struct efa_com_dev *edev, create_cmd.cq_depth = params->cq_depth; create_cmd.num_sub_cqs = params->num_sub_cqs; create_cmd.uar = params->uarn; + if (params->interrupt_mode_enabled) { + EFA_SET(&create_cmd.cq_caps_1, + EFA_ADMIN_CREATE_CQ_CMD_INTERRUPT_MODE_ENABLED, 1); + create_cmd.eqn = params->eqn; + } efa_com_set_dma_addr(params->dma_addr, &create_cmd.cq_ba.mem_addr_high, @@ -187,6 +186,9 @@ int efa_com_create_cq(struct efa_com_dev *edev, result->cq_idx = cmd_completion.cq_idx; result->actual_depth = params->cq_depth; + result->db_off = cmd_completion.db_offset; + result->db_valid = EFA_GET(&cmd_completion.flags, + EFA_ADMIN_CREATE_CQ_RESP_DB_VALID); return 0; } @@ -497,6 +499,23 @@ int efa_com_get_device_attr(struct efa_com_dev *edev, sizeof(resp.u.network_attr.addr)); result->mtu = resp.u.network_attr.mtu; + if (efa_com_check_supported_feature_id(edev, + EFA_ADMIN_EVENT_QUEUE_ATTR)) { + err = efa_com_get_feature(edev, &resp, + EFA_ADMIN_EVENT_QUEUE_ATTR); + if (err) { + ibdev_err_ratelimited( + edev->efa_dev, + "Failed to get event queue attributes %d\n", + err); + return err; + } + + result->max_eq = resp.u.event_queue_attr.max_eq; + result->max_eq_depth = resp.u.event_queue_attr.max_eq_depth; + result->event_bitmask = resp.u.event_queue_attr.event_bitmask; + } + return 0; } diff --git a/drivers/infiniband/hw/efa/efa_com_cmd.h b/drivers/infiniband/hw/efa/efa_com_cmd.h index eea4ebfbe6ec..c33010bbf9e8 100644 --- a/drivers/infiniband/hw/efa/efa_com_cmd.h +++ b/drivers/infiniband/hw/efa/efa_com_cmd.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause */ /* - * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved. + * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved. */ #ifndef _EFA_COM_CMD_H_ @@ -73,7 +73,9 @@ struct efa_com_create_cq_params { u16 cq_depth; u16 num_sub_cqs; u16 uarn; + u16 eqn; u8 entry_size_in_bytes; + bool interrupt_mode_enabled; }; struct efa_com_create_cq_result { @@ -81,6 +83,8 @@ struct efa_com_create_cq_result { u16 cq_idx; /* actual cq depth in # of entries */ u16 actual_depth; + u32 db_off; + bool db_valid; }; struct efa_com_destroy_cq_params { @@ -125,6 +129,9 @@ struct efa_com_get_device_attr_result { u32 max_llq_size; u32 max_rdma_size; u32 device_caps; + u32 max_eq; + u32 max_eq_depth; + u32 event_bitmask; /* EQ events bitmask */ u16 sub_cqs_per_cq; u16 max_sq_sge; u16 max_rq_sge; @@ -260,7 +267,6 @@ union efa_com_get_stats_result { struct efa_com_rdma_read_stats rdma_read_stats; }; -void efa_com_set_dma_addr(dma_addr_t addr, u32 *addr_high, u32 *addr_low); int efa_com_create_qp(struct efa_com_dev *edev, struct efa_com_create_qp_params *params, struct efa_com_create_qp_result *res); diff --git a/drivers/infiniband/hw/efa/efa_main.c b/drivers/infiniband/hw/efa/efa_main.c index 417dea5f90cf..29db4dec02f0 100644 --- a/drivers/infiniband/hw/efa/efa_main.c +++ b/drivers/infiniband/hw/efa/efa_main.c @@ -67,6 +67,46 @@ static void efa_release_bars(struct efa_dev *dev, int bars_mask) pci_release_selected_regions(pdev, release_bars); } +static void efa_process_comp_eqe(struct efa_dev *dev, struct efa_admin_eqe *eqe) +{ + u16 cqn = eqe->u.comp_event.cqn; + struct efa_cq *cq; + + cq = xa_load(&dev->cqs_xa, cqn); + if (unlikely(!cq)) { + ibdev_err_ratelimited(&dev->ibdev, + "Completion event on non-existent CQ[%u]", + cqn); + return; + } + + cq->ibcq.comp_handler(&cq->ibcq, cq->ibcq.cq_context); +} + +static void efa_process_eqe(struct efa_com_eq *eeq, struct efa_admin_eqe *eqe) +{ + struct efa_dev *dev = container_of(eeq->edev, struct efa_dev, edev); + + if (likely(EFA_GET(&eqe->common, EFA_ADMIN_EQE_EVENT_TYPE) == + EFA_ADMIN_EQE_EVENT_TYPE_COMPLETION)) + efa_process_comp_eqe(dev, eqe); + else + ibdev_err_ratelimited(&dev->ibdev, + "Unknown event type received %lu", + EFA_GET(&eqe->common, + EFA_ADMIN_EQE_EVENT_TYPE)); +} + +static irqreturn_t efa_intr_msix_comp(int irq, void *data) +{ + struct efa_eq *eq = data; + struct efa_com_dev *edev = eq->eeq.edev; + + efa_com_eq_comp_intr_handler(edev, &eq->eeq); + + return IRQ_HANDLED; +} + static irqreturn_t efa_intr_msix_mgmnt(int irq, void *data) { struct efa_dev *dev = data; @@ -77,26 +117,43 @@ static irqreturn_t efa_intr_msix_mgmnt(int irq, void *data) return IRQ_HANDLED; } -static int efa_request_mgmnt_irq(struct efa_dev *dev) +static int efa_request_irq(struct efa_dev *dev, struct efa_irq *irq) { - struct efa_irq *irq; int err; - irq = &dev->admin_irq; err = request_irq(irq->irqn, irq->handler, 0, irq->name, irq->data); if (err) { - dev_err(&dev->pdev->dev, "Failed to request admin irq (%d)\n", - err); + dev_err(&dev->pdev->dev, "Failed to request irq %s (%d)\n", + irq->name, err); return err; } - dev_dbg(&dev->pdev->dev, "Set affinity hint of mgmnt irq to %*pbl (irq vector: %d)\n", - nr_cpumask_bits, &irq->affinity_hint_mask, irq->irqn); irq_set_affinity_hint(irq->irqn, &irq->affinity_hint_mask); return 0; } +static void efa_setup_comp_irq(struct efa_dev *dev, struct efa_eq *eq, + int vector) +{ + u32 cpu; + + cpu = vector - EFA_COMP_EQS_VEC_BASE; + snprintf(eq->irq.name, EFA_IRQNAME_SIZE, "efa-comp%d@pci:%s", cpu, + pci_name(dev->pdev)); + eq->irq.handler = efa_intr_msix_comp; + eq->irq.data = eq; + eq->irq.vector = vector; + eq->irq.irqn = pci_irq_vector(dev->pdev, vector); + cpumask_set_cpu(cpu, &eq->irq.affinity_hint_mask); +} + +static void efa_free_irq(struct efa_dev *dev, struct efa_irq *irq) +{ + irq_set_affinity_hint(irq->irqn, NULL); + free_irq(irq->irqn, irq->data); +} + static void efa_setup_mgmnt_irq(struct efa_dev *dev) { u32 cpu; @@ -105,8 +162,9 @@ static void efa_setup_mgmnt_irq(struct efa_dev *dev) "efa-mgmnt@pci:%s", pci_name(dev->pdev)); dev->admin_irq.handler = efa_intr_msix_mgmnt; dev->admin_irq.data = dev; - dev->admin_irq.irqn = - pci_irq_vector(dev->pdev, dev->admin_msix_vector_idx); + dev->admin_irq.vector = dev->admin_msix_vector_idx; + dev->admin_irq.irqn = pci_irq_vector(dev->pdev, + dev->admin_msix_vector_idx); cpu = cpumask_first(cpu_online_mask); cpumask_set_cpu(cpu, &dev->admin_irq.affinity_hint_mask); @@ -115,20 +173,11 @@ static void efa_setup_mgmnt_irq(struct efa_dev *dev) dev->admin_irq.name); } -static void efa_free_mgmnt_irq(struct efa_dev *dev) -{ - struct efa_irq *irq; - - irq = &dev->admin_irq; - irq_set_affinity_hint(irq->irqn, NULL); - free_irq(irq->irqn, irq->data); -} - static int efa_set_mgmnt_irq(struct efa_dev *dev) { efa_setup_mgmnt_irq(dev); - return efa_request_mgmnt_irq(dev); + return efa_request_irq(dev, &dev->admin_irq); } static int efa_request_doorbell_bar(struct efa_dev *dev) @@ -234,6 +283,76 @@ static void efa_set_host_info(struct efa_dev *dev) dma_free_coherent(&dev->pdev->dev, bufsz, hinf, hinf_dma); } +static int efa_destroy_eq(struct efa_dev *dev, struct efa_eq *eq) +{ + efa_com_eq_destroy(&dev->edev, &eq->eeq); + efa_free_irq(dev, &eq->irq); + + return 0; +} + +static int efa_create_eq(struct efa_dev *dev, struct efa_eq *eq, u8 msix_vec) +{ + int err; + + efa_setup_comp_irq(dev, eq, msix_vec); + err = efa_request_irq(dev, &eq->irq); + if (err) + return err; + + err = efa_com_eq_init(&dev->edev, &eq->eeq, efa_process_eqe, + dev->dev_attr.max_eq_depth, msix_vec); + if (err) + goto err_free_comp_irq; + + return 0; + +err_free_comp_irq: + efa_free_irq(dev, &eq->irq); + return err; +} + +static int efa_create_eqs(struct efa_dev *dev) +{ + unsigned int neqs = dev->dev_attr.max_eq; + int err; + int i; + + neqs = min_t(unsigned int, neqs, num_online_cpus()); + dev->neqs = neqs; + dev->eqs = kcalloc(neqs, sizeof(*dev->eqs), GFP_KERNEL); + if (!dev->eqs) + return -ENOMEM; + + for (i = 0; i < neqs; i++) { + err = efa_create_eq(dev, &dev->eqs[i], + i + EFA_COMP_EQS_VEC_BASE); + if (err) + goto err_destroy_eqs; + } + + return 0; + +err_destroy_eqs: + for (i--; i >= 0; i--) + efa_destroy_eq(dev, &dev->eqs[i]); + kfree(dev->eqs); + + return err; +} + +static int efa_destroy_eqs(struct efa_dev *dev) +{ + int i; + + for (i = 0; i < dev->neqs; i++) + efa_destroy_eq(dev, &dev->eqs[i]); + + kfree(dev->eqs); + + return 0; +} + static const struct ib_device_ops efa_dev_ops = { .owner = THIS_MODULE, .driver_id = RDMA_DRIVER_EFA, @@ -300,23 +419,29 @@ static int efa_ib_device_add(struct efa_dev *dev) if (err) goto err_release_doorbell_bar; + err = efa_create_eqs(dev); + if (err) + goto err_release_doorbell_bar; + efa_set_host_info(dev); dev->ibdev.node_type = RDMA_NODE_UNSPECIFIED; dev->ibdev.phys_port_cnt = 1; - dev->ibdev.num_comp_vectors = 1; + dev->ibdev.num_comp_vectors = dev->neqs ?: 1; dev->ibdev.dev.parent = &pdev->dev; ib_set_device_ops(&dev->ibdev, &efa_dev_ops); err = ib_register_device(&dev->ibdev, "efa_%d", &pdev->dev); if (err) - goto err_release_doorbell_bar; + goto err_destroy_eqs; ibdev_info(&dev->ibdev, "IB device registered\n"); return 0; +err_destroy_eqs: + efa_destroy_eqs(dev); err_release_doorbell_bar: efa_release_doorbell_bar(dev); return err; @@ -324,9 +449,10 @@ static int efa_ib_device_add(struct efa_dev *dev) static void efa_ib_device_remove(struct efa_dev *dev) { - efa_com_dev_reset(&dev->edev, EFA_REGS_RESET_NORMAL); ibdev_info(&dev->ibdev, "Unregister ib device\n"); ib_unregister_device(&dev->ibdev); + efa_destroy_eqs(dev); + efa_com_dev_reset(&dev->edev, EFA_REGS_RESET_NORMAL); efa_release_doorbell_bar(dev); } @@ -339,8 +465,12 @@ static int efa_enable_msix(struct efa_dev *dev) { int msix_vecs, irq_num; - /* Reserve the max msix vectors we might need */ - msix_vecs = EFA_NUM_MSIX_VEC; + /* + * Reserve the max msix vectors we might need, one vector is reserved + * for admin. + */ + msix_vecs = min_t(int, pci_msix_vec_count(dev->pdev), + num_online_cpus() + 1); dev_dbg(&dev->pdev->dev, "Trying to enable MSI-X, vectors %d\n", msix_vecs); @@ -421,6 +551,7 @@ static struct efa_dev *efa_probe_device(struct pci_dev *pdev) edev->efa_dev = dev; edev->dmadev = &pdev->dev; dev->pdev = pdev; + xa_init_flags(&dev->cqs_xa, XA_FLAGS_LOCK_IRQ); bars = pci_select_bars(pdev, IORESOURCE_MEM) & EFA_BASE_BAR_MASK; err = pci_request_selected_regions(pdev, bars, DRV_MODULE_NAME); @@ -476,7 +607,7 @@ static struct efa_dev *efa_probe_device(struct pci_dev *pdev) return dev; err_free_mgmnt_irq: - efa_free_mgmnt_irq(dev); + efa_free_irq(dev, &dev->admin_irq); err_disable_msix: efa_disable_msix(dev); err_reg_read_destroy: @@ -499,11 +630,12 @@ static void efa_remove_device(struct pci_dev *pdev) edev = &dev->edev; efa_com_admin_destroy(edev); - efa_free_mgmnt_irq(dev); + efa_free_irq(dev, &dev->admin_irq); efa_disable_msix(dev); efa_com_mmio_reg_read_destroy(edev); devm_iounmap(&pdev->dev, edev->reg_bar); efa_release_bars(dev, EFA_BASE_BAR_MASK); + xa_destroy(&dev->cqs_xa); ib_dealloc_device(&dev->ibdev); pci_disable_device(pdev); } diff --git a/drivers/infiniband/hw/efa/efa_regs_defs.h b/drivers/infiniband/hw/efa/efa_regs_defs.h index 4017982fe13b..714ae6258800 100644 --- a/drivers/infiniband/hw/efa/efa_regs_defs.h +++ b/drivers/infiniband/hw/efa/efa_regs_defs.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause */ /* - * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved. + * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved. */ #ifndef _EFA_REGS_H_ @@ -42,6 +42,7 @@ enum efa_regs_reset_reason_types { #define EFA_REGS_MMIO_REG_READ_OFF 0x5c #define EFA_REGS_MMIO_RESP_LO_OFF 0x60 #define EFA_REGS_MMIO_RESP_HI_OFF 0x64 +#define EFA_REGS_EQ_DB_OFF 0x68 /* version register */ #define EFA_REGS_VERSION_MINOR_VERSION_MASK 0xff @@ -93,4 +94,8 @@ enum efa_regs_reset_reason_types { #define EFA_REGS_MMIO_REG_READ_REQ_ID_MASK 0xffff #define EFA_REGS_MMIO_REG_READ_REG_OFF_MASK 0xffff0000 +/* eq_db register */ +#define EFA_REGS_EQ_DB_EQN_MASK 0xffff +#define EFA_REGS_EQ_DB_ARM_MASK 0x80000000 + #endif /* _EFA_REGS_H_ */ diff --git a/drivers/infiniband/hw/efa/efa_verbs.c b/drivers/infiniband/hw/efa/efa_verbs.c index e5f9d90aad5e..a5d1ff63bc7c 100644 --- a/drivers/infiniband/hw/efa/efa_verbs.c +++ b/drivers/infiniband/hw/efa/efa_verbs.c @@ -245,6 +245,9 @@ int efa_query_device(struct ib_device *ibdev, if (EFA_DEV_CAP(dev, RNR_RETRY)) resp.device_caps |= EFA_QUERY_DEVICE_CAPS_RNR_RETRY; + if (dev->neqs) + resp.device_caps |= EFA_QUERY_DEVICE_CAPS_CQ_NOTIFICATIONS; + err = ib_copy_to_udata(udata, &resp, min(sizeof(resp), udata->outlen)); if (err) { @@ -984,6 +987,12 @@ static int efa_destroy_cq_idx(struct efa_dev *dev, int cq_idx) return efa_com_destroy_cq(&dev->edev, ¶ms); } +static void efa_cq_user_mmap_entries_remove(struct efa_cq *cq) +{ + rdma_user_mmap_entry_remove(cq->db_mmap_entry); + rdma_user_mmap_entry_remove(cq->mmap_entry); +} + int efa_destroy_cq(struct ib_cq *ibcq, struct ib_udata *udata) { struct efa_dev *dev = to_edev(ibcq->device); @@ -993,15 +1002,22 @@ int efa_destroy_cq(struct ib_cq *ibcq, struct ib_udata *udata) "Destroy cq[%d] virt[0x%p] freed: size[%lu], dma[%pad]\n", cq->cq_idx, cq->cpu_addr, cq->size, &cq->dma_addr); - rdma_user_mmap_entry_remove(cq->mmap_entry); + xa_erase(&dev->cqs_xa, cq->cq_idx); + efa_cq_user_mmap_entries_remove(cq); efa_destroy_cq_idx(dev, cq->cq_idx); efa_free_mapped(dev, cq->cpu_addr, cq->dma_addr, cq->size, DMA_FROM_DEVICE); return 0; } +static int efa_vec2eqn(struct efa_dev *dev, int vec) +{ + return dev->eqs[vec].eeq.eqn; +} + static int cq_mmap_entries_setup(struct efa_dev *dev, struct efa_cq *cq, - struct efa_ibv_create_cq_resp *resp) + struct efa_ibv_create_cq_resp *resp, + bool db_valid) { resp->q_mmap_size = cq->size; cq->mmap_entry = efa_user_mmap_entry_insert(&cq->ucontext->ibucontext, @@ -1011,6 +1027,21 @@ static int cq_mmap_entries_setup(struct efa_dev *dev, struct efa_cq *cq, if (!cq->mmap_entry) return -ENOMEM; + if (db_valid) { + cq->db_mmap_entry = + efa_user_mmap_entry_insert(&cq->ucontext->ibucontext, + dev->db_bar_addr + resp->db_off, + PAGE_SIZE, EFA_MMAP_IO_NC, + &resp->db_mmap_key); + if (!cq->db_mmap_entry) { + rdma_user_mmap_entry_remove(cq->mmap_entry); + return -ENOMEM; + } + + resp->db_off &= ~PAGE_MASK; + resp->comp_mask |= EFA_CREATE_CQ_RESP_DB_OFF; + } + return 0; } @@ -1019,8 +1050,8 @@ int efa_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, { struct efa_ucontext *ucontext = rdma_udata_to_drv_context( udata, struct efa_ucontext, ibucontext); + struct efa_com_create_cq_params params = {}; struct efa_ibv_create_cq_resp resp = {}; - struct efa_com_create_cq_params params; struct efa_com_create_cq_result result; struct ib_device *ibdev = ibcq->device; struct efa_dev *dev = to_edev(ibdev); @@ -1065,7 +1096,7 @@ int efa_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, goto err_out; } - if (cmd.comp_mask || !is_reserved_cleared(cmd.reserved_50)) { + if (cmd.comp_mask || !is_reserved_cleared(cmd.reserved_58)) { ibdev_dbg(ibdev, "Incompatible ABI params, unknown fields in udata\n"); err = -EINVAL; @@ -1101,29 +1132,42 @@ int efa_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, params.dma_addr = cq->dma_addr; params.entry_size_in_bytes = cmd.cq_entry_size; params.num_sub_cqs = cmd.num_sub_cqs; + if (cmd.flags & EFA_CREATE_CQ_WITH_COMPLETION_CHANNEL) { + params.eqn = efa_vec2eqn(dev, attr->comp_vector); + params.interrupt_mode_enabled = true; + } + err = efa_com_create_cq(&dev->edev, ¶ms, &result); if (err) goto err_free_mapped; + resp.db_off = result.db_off; resp.cq_idx = result.cq_idx; cq->cq_idx = result.cq_idx; cq->ibcq.cqe = result.actual_depth; WARN_ON_ONCE(entries != result.actual_depth); - err = cq_mmap_entries_setup(dev, cq, &resp); + err = cq_mmap_entries_setup(dev, cq, &resp, result.db_valid); if (err) { ibdev_dbg(ibdev, "Could not setup cq[%u] mmap entries\n", cq->cq_idx); goto err_destroy_cq; } + err = xa_err(xa_store(&dev->cqs_xa, cq->cq_idx, cq, GFP_KERNEL)); + if (err) { + ibdev_dbg(ibdev, "Failed to store cq[%u] in xarray\n", + cq->cq_idx); + goto err_remove_mmap; + } + if (udata->outlen) { err = ib_copy_to_udata(udata, &resp, min(sizeof(resp), udata->outlen)); if (err) { ibdev_dbg(ibdev, "Failed to copy udata for create_cq\n"); - goto err_remove_mmap; + goto err_xa_erase; } } @@ -1132,8 +1176,10 @@ int efa_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, return 0; +err_xa_erase: + xa_erase(&dev->cqs_xa, cq->cq_idx); err_remove_mmap: - rdma_user_mmap_entry_remove(cq->mmap_entry); + efa_cq_user_mmap_entries_remove(cq); err_destroy_cq: efa_destroy_cq_idx(dev, cq->cq_idx); err_free_mapped: diff --git a/include/uapi/rdma/efa-abi.h b/include/uapi/rdma/efa-abi.h index f89fbb5b1e8d..08035ccf1fff 100644 --- a/include/uapi/rdma/efa-abi.h +++ b/include/uapi/rdma/efa-abi.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */ /* - * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved. + * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved. */ #ifndef EFA_ABI_USER_H @@ -52,11 +52,20 @@ struct efa_ibv_alloc_pd_resp { __u8 reserved_30[2]; }; +enum { + EFA_CREATE_CQ_WITH_COMPLETION_CHANNEL = 1 << 0, +}; + struct efa_ibv_create_cq { __u32 comp_mask; __u32 cq_entry_size; __u16 num_sub_cqs; - __u8 reserved_50[6]; + __u8 flags; + __u8 reserved_58[5]; +}; + +enum { + EFA_CREATE_CQ_RESP_DB_OFF = 1 << 0, }; struct efa_ibv_create_cq_resp { @@ -65,7 +74,9 @@ struct efa_ibv_create_cq_resp { __aligned_u64 q_mmap_key; __aligned_u64 q_mmap_size; __u16 cq_idx; - __u8 reserved_d0[6]; + __u8 reserved_d0[2]; + __u32 db_off; + __aligned_u64 db_mmap_key; }; enum { @@ -106,6 +117,7 @@ struct efa_ibv_create_ah_resp { enum { EFA_QUERY_DEVICE_CAPS_RDMA_READ = 1 << 0, EFA_QUERY_DEVICE_CAPS_RNR_RETRY = 1 << 1, + EFA_QUERY_DEVICE_CAPS_CQ_NOTIFICATIONS = 1 << 2, }; struct efa_ibv_ex_query_device_resp {