From patchwork Mon Oct 30 03:44:27 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Mike Christie X-Patchwork-Id: 10031969 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 69EAC603B4 for ; Mon, 30 Oct 2017 03:44:54 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 5B189286BD for ; Mon, 30 Oct 2017 03:44:54 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 5028028763; Mon, 30 Oct 2017 03:44:54 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=unavailable version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id EFC23286BD for ; Mon, 30 Oct 2017 03:44:53 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752203AbdJ3Dow (ORCPT ); Sun, 29 Oct 2017 23:44:52 -0400 Received: from mx1.redhat.com ([209.132.183.28]:38000 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752200AbdJ3Dou (ORCPT ); Sun, 29 Oct 2017 23:44:50 -0400 Received: from smtp.corp.redhat.com (int-mx04.intmail.prod.int.phx2.redhat.com [10.5.11.14]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id B1F7E49012; Mon, 30 Oct 2017 03:44:50 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com B1F7E49012 Authentication-Results: ext-mx09.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com Authentication-Results: ext-mx09.extmail.prod.ext.phx2.redhat.com; spf=fail smtp.mailfrom=mchristi@redhat.com Received: from rh2.redhat.com (ovpn-120-93.rdu2.redhat.com [10.10.120.93]) by smtp.corp.redhat.com (Postfix) with ESMTP id BBC8A5D9C9; Mon, 30 Oct 2017 03:44:49 +0000 (UTC) From: Mike Christie To: martin.petersen@oracle.com, jejb@linux.vnet.ibm.com, linux-scsi@vger.kernel.org, target-devel@vger.kernel.org, nab@linux-iscsi.org Cc: Mike Christie Subject: [PATCH 07/19] tcmu: fix unmap thread race Date: Sun, 29 Oct 2017 22:44:27 -0500 Message-Id: <1509335079-5276-8-git-send-email-mchristi@redhat.com> In-Reply-To: <1509335079-5276-1-git-send-email-mchristi@redhat.com> References: <1509335079-5276-1-git-send-email-mchristi@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.14 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.38]); Mon, 30 Oct 2017 03:44:50 +0000 (UTC) Sender: target-devel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: target-devel@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP If the unmap thread has already run find_free_blocks but not yet run prepare_to_wait when a wake_up(&unmap_wait) call is done, the unmap thread is going to miss the wake call. Instead of adding checks for if new waiters were added this just has use a work queue which will run us again in this type of case. Signed-off-by: Mike Christie --- drivers/target/target_core_user.c | 43 +++++++++------------------------------ 1 file changed, 10 insertions(+), 33 deletions(-) diff --git a/drivers/target/target_core_user.c b/drivers/target/target_core_user.c index 15b54fd..14d9b79 100644 --- a/drivers/target/target_core_user.c +++ b/drivers/target/target_core_user.c @@ -32,7 +32,7 @@ #include #include #include -#include +#include #include #include #include @@ -176,12 +176,11 @@ struct tcmu_cmd { unsigned long flags; }; -static struct task_struct *unmap_thread; -static wait_queue_head_t unmap_wait; static DEFINE_MUTEX(root_udev_mutex); static LIST_HEAD(root_udev); static atomic_t global_db_count = ATOMIC_INIT(0); +static struct work_struct tcmu_unmap_work; static struct kmem_cache *tcmu_cmd_cache; @@ -389,8 +388,7 @@ static bool tcmu_get_empty_blocks(struct tcmu_dev *udev, err: udev->waiting_global = true; - /* Try to wake up the unmap thread */ - wake_up(&unmap_wait); + schedule_work(&tcmu_unmap_work); return false; } @@ -1063,8 +1061,7 @@ static void tcmu_device_timedout(unsigned long data) idr_for_each(&udev->commands, tcmu_check_expired_cmd, NULL); spin_unlock_irqrestore(&udev->commands_lock, flags); - /* Try to wake up the ummap thread */ - wake_up(&unmap_wait); + schedule_work(&tcmu_unmap_work); /* * We don't need to wakeup threads on wait_cmdr since they have their @@ -2042,23 +2039,10 @@ static void run_cmdr_queues(void) mutex_unlock(&root_udev_mutex); } -static int unmap_thread_fn(void *data) +static void tcmu_unmap_work_fn(struct work_struct *work) { - while (!kthread_should_stop()) { - DEFINE_WAIT(__wait); - - prepare_to_wait(&unmap_wait, &__wait, TASK_INTERRUPTIBLE); - schedule(); - finish_wait(&unmap_wait, &__wait); - - if (kthread_should_stop()) - break; - - find_free_blocks(); - run_cmdr_queues(); - } - - return 0; + find_free_blocks(); + run_cmdr_queues(); } static int __init tcmu_module_init(void) @@ -2067,6 +2051,8 @@ static int __init tcmu_module_init(void) BUILD_BUG_ON((sizeof(struct tcmu_cmd_entry) % TCMU_OP_ALIGN_SIZE) != 0); + INIT_WORK(&tcmu_unmap_work, tcmu_unmap_work_fn); + tcmu_cmd_cache = kmem_cache_create("tcmu_cmd_cache", sizeof(struct tcmu_cmd), __alignof__(struct tcmu_cmd), @@ -2112,17 +2098,8 @@ static int __init tcmu_module_init(void) if (ret) goto out_attrs; - init_waitqueue_head(&unmap_wait); - unmap_thread = kthread_run(unmap_thread_fn, NULL, "tcmu_unmap"); - if (IS_ERR(unmap_thread)) { - ret = PTR_ERR(unmap_thread); - goto out_unreg_transport; - } - return 0; -out_unreg_transport: - target_backend_unregister(&tcmu_ops); out_attrs: kfree(tcmu_attrs); out_unreg_genl: @@ -2137,7 +2114,7 @@ static int __init tcmu_module_init(void) static void __exit tcmu_module_exit(void) { - kthread_stop(unmap_thread); + cancel_work_sync(&tcmu_unmap_work); target_backend_unregister(&tcmu_ops); kfree(tcmu_attrs); genl_unregister_family(&tcmu_genl_family);