From patchwork Thu Oct 1 15:55:26 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Matthew R. Ochs" X-Patchwork-Id: 7309611 Return-Path: X-Original-To: patchwork-linux-scsi@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 95E83BEEA4 for ; Thu, 1 Oct 2015 15:56:22 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 70C8220610 for ; Thu, 1 Oct 2015 15:56:21 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 04E0A205D6 for ; Thu, 1 Oct 2015 15:56:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756616AbbJAP4T (ORCPT ); Thu, 1 Oct 2015 11:56:19 -0400 Received: from e17.ny.us.ibm.com ([129.33.205.207]:55225 "EHLO e17.ny.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752818AbbJAP4S (ORCPT ); Thu, 1 Oct 2015 11:56:18 -0400 Received: from localhost by e17.ny.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 1 Oct 2015 11:56:17 -0400 Received: from d01dlp02.pok.ibm.com (9.56.250.167) by e17.ny.us.ibm.com (146.89.104.204) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Thu, 1 Oct 2015 11:56:16 -0400 X-IBM-Helo: d01dlp02.pok.ibm.com X-IBM-MailFrom: mrochs@linux.vnet.ibm.com X-IBM-RcptTo: linux-scsi@vger.kernel.org Received: from b01cxnp22036.gho.pok.ibm.com (b01cxnp22036.gho.pok.ibm.com [9.57.198.26]) by d01dlp02.pok.ibm.com (Postfix) with ESMTP id 1F6916E804B for ; Thu, 1 Oct 2015 11:44:28 -0400 (EDT) Received: from d01av03.pok.ibm.com (d01av03.pok.ibm.com [9.56.224.217]) by b01cxnp22036.gho.pok.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id t91FuFUQ54591568 for ; Thu, 1 Oct 2015 15:56:15 GMT Received: from d01av03.pok.ibm.com (localhost [127.0.0.1]) by d01av03.pok.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id t91FuCfS000463 for ; Thu, 1 Oct 2015 11:56:15 -0400 Received: from p8tul1-build.aus.stglabs.ibm.com (als141206.austin.ibm.com [9.3.141.206]) by d01av03.pok.ibm.com (8.14.4/8.14.4/NCO v10.0 AVin) with ESMTP id t91FuBdU032752; Thu, 1 Oct 2015 11:56:11 -0400 From: "Matthew R. Ochs" To: linux-scsi@vger.kernel.org, James Bottomley , "Nicholas A. Bellinger" , Brian King , Ian Munsie , Daniel Axtens , Andrew Donnellan , Tomas Henzl , David Laight Cc: Michael Neuling , linuxppc-dev@lists.ozlabs.org, "Manoj N. Kumar" Subject: [PATCH v5 08/34] cxlflash: Fix to avoid CXL services during EEH Date: Thu, 1 Oct 2015 10:55:26 -0500 Message-Id: <1443714926-15545-1-git-send-email-mrochs@linux.vnet.ibm.com> X-Mailer: git-send-email 2.1.0 In-Reply-To: <1443714773-9176-1-git-send-email-mrochs@linux.vnet.ibm.com> References: <1443714773-9176-1-git-send-email-mrochs@linux.vnet.ibm.com> X-TM-AS-MML: disable X-Content-Scanned: Fidelis XPS MAILER x-cbid: 15100115-0041-0000-0000-000001B80D16 Sender: linux-scsi-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-scsi@vger.kernel.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, T_RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP During an EEH freeze event, certain CXL services should not be called until after the hardware reset has taken place. Doing so can result in unnecessary failures and possibly cause other ill effects by triggering hardware accesses. This translates to a requirement to quiesce all threads that may potentially use CXL runtime service during this window. In particular, multiple ioctls make use of the CXL services when acting on contexts on behalf of the user. Thus, it is essential to 'drain' running ioctls _before_ proceeding with handling the EEH freeze event. Create the ability to drain ioctls by wrapping the ioctl handler call in a read semaphore and then implementing a small routine that obtains the write semaphore, effectively creating a wait point for all currently executing ioctls. Signed-off-by: Matthew R. Ochs Signed-off-by: Manoj N. Kumar Reviewed-by: Brian King Reviewed-by: Daniel Axtens Reviewed-by: Tomas Henzl --- drivers/scsi/cxlflash/common.h | 2 + drivers/scsi/cxlflash/main.c | 18 +++++-- drivers/scsi/cxlflash/superpipe.c | 98 ++++++++++++++++++++++++--------------- 3 files changed, 77 insertions(+), 41 deletions(-) diff --git a/drivers/scsi/cxlflash/common.h b/drivers/scsi/cxlflash/common.h index 1c56037..1abe4e0 100644 --- a/drivers/scsi/cxlflash/common.h +++ b/drivers/scsi/cxlflash/common.h @@ -16,6 +16,7 @@ #define _CXLFLASH_COMMON_H #include +#include #include #include #include @@ -110,6 +111,7 @@ struct cxlflash_cfg { atomic_t recovery_threads; struct mutex ctx_recovery_mutex; struct mutex ctx_tbl_list_mutex; + struct rw_semaphore ioctl_rwsem; struct ctx_info *ctx_tbl[MAX_CONTEXT]; struct list_head ctx_err_recovery; /* contexts w/ recovery pending */ struct file_operations cxl_fops; diff --git a/drivers/scsi/cxlflash/main.c b/drivers/scsi/cxlflash/main.c index 3e3ccf1..6e85c77 100644 --- a/drivers/scsi/cxlflash/main.c +++ b/drivers/scsi/cxlflash/main.c @@ -2311,6 +2311,7 @@ static int cxlflash_probe(struct pci_dev *pdev, cfg->lr_port = -1; mutex_init(&cfg->ctx_tbl_list_mutex); mutex_init(&cfg->ctx_recovery_mutex); + init_rwsem(&cfg->ioctl_rwsem); INIT_LIST_HEAD(&cfg->ctx_err_recovery); INIT_LIST_HEAD(&cfg->lluns); @@ -2365,6 +2366,19 @@ out_remove: } /** + * drain_ioctls() - wait until all currently executing ioctls have completed + * @cfg: Internal structure associated with the host. + * + * Obtain write access to read/write semaphore that wraps ioctl + * handling to 'drain' ioctls currently executing. + */ +static void drain_ioctls(struct cxlflash_cfg *cfg) +{ + down_write(&cfg->ioctl_rwsem); + up_write(&cfg->ioctl_rwsem); +} + +/** * cxlflash_pci_error_detected() - called when a PCI error is detected * @pdev: PCI device struct. * @state: PCI channel state. @@ -2383,16 +2397,14 @@ static pci_ers_result_t cxlflash_pci_error_detected(struct pci_dev *pdev, switch (state) { case pci_channel_io_frozen: cfg->state = STATE_LIMBO; - - /* Turn off legacy I/O */ scsi_block_requests(cfg->host); + drain_ioctls(cfg); rc = cxlflash_mark_contexts_error(cfg); if (unlikely(rc)) dev_err(dev, "%s: Failed to mark user contexts!(%d)\n", __func__, rc); term_mc(cfg, UNDO_START); stop_afu(cfg); - return PCI_ERS_RESULT_NEED_RESET; case pci_channel_io_perm_failure: cfg->state = STATE_FAILTERM; diff --git a/drivers/scsi/cxlflash/superpipe.c b/drivers/scsi/cxlflash/superpipe.c index 28aa9d9..655cbf1 100644 --- a/drivers/scsi/cxlflash/superpipe.c +++ b/drivers/scsi/cxlflash/superpipe.c @@ -1214,6 +1214,46 @@ static const struct file_operations null_fops = { }; /** + * check_state() - checks and responds to the current adapter state + * @cfg: Internal structure associated with the host. + * + * This routine can block and should only be used on process context. + * It assumes that the caller is an ioctl thread and holding the ioctl + * read semaphore. This is temporarily let up across the wait to allow + * for draining actively running ioctls. Also note that when waking up + * from waiting in reset, the state is unknown and must be checked again + * before proceeding. + * + * Return: 0 on success, -errno on failure + */ +static int check_state(struct cxlflash_cfg *cfg) +{ + struct device *dev = &cfg->dev->dev; + int rc = 0; + +retry: + switch (cfg->state) { + case STATE_LIMBO: + dev_dbg(dev, "%s: Limbo state, going to wait...\n", __func__); + up_read(&cfg->ioctl_rwsem); + rc = wait_event_interruptible(cfg->limbo_waitq, + cfg->state != STATE_LIMBO); + down_read(&cfg->ioctl_rwsem); + if (unlikely(rc)) + break; + goto retry; + case STATE_FAILTERM: + dev_dbg(dev, "%s: Failed/Terminating!\n", __func__); + rc = -ENODEV; + break; + default: + break; + } + + return rc; +} + +/** * cxlflash_disk_attach() - attach a LUN to a context * @sdev: SCSI device associated with LUN. * @attach: Attach ioctl data structure. @@ -1523,41 +1563,6 @@ err1: } /** - * check_state() - checks and responds to the current adapter state - * @cfg: Internal structure associated with the host. - * - * This routine can block and should only be used on process context. - * Note that when waking up from waiting in limbo, the state is unknown - * and must be checked again before proceeding. - * - * Return: 0 on success, -errno on failure - */ -static int check_state(struct cxlflash_cfg *cfg) -{ - struct device *dev = &cfg->dev->dev; - int rc = 0; - -retry: - switch (cfg->state) { - case STATE_LIMBO: - dev_dbg(dev, "%s: Limbo, going to wait...\n", __func__); - rc = wait_event_interruptible(cfg->limbo_waitq, - cfg->state != STATE_LIMBO); - if (unlikely(rc)) - break; - goto retry; - case STATE_FAILTERM: - dev_dbg(dev, "%s: Failed/Terminating!\n", __func__); - rc = -ENODEV; - break; - default: - break; - } - - return rc; -} - -/** * cxlflash_afu_recover() - initiates AFU recovery * @sdev: SCSI device associated with LUN. * @recover: Recover ioctl data structure. @@ -1646,9 +1651,14 @@ retry_recover: /* Test if in error state */ reg = readq_be(&afu->ctrl_map->mbox_r); if (reg == -1) { - dev_dbg(dev, "%s: MMIO read fail! Wait for recovery...\n", - __func__); - mutex_unlock(&ctxi->mutex); + dev_dbg(dev, "%s: MMIO fail, wait for recovery.\n", __func__); + + /* + * Before checking the state, put back the context obtained with + * get_context() as it is no longer needed and sleep for a short + * period of time (see prolog notes). + */ + put_context(ctxi); ctxi = NULL; ssleep(1); rc = check_state(cfg); @@ -1967,6 +1977,14 @@ out: * @cmd: IOCTL command. * @arg: Userspace ioctl data structure. * + * A read/write semaphore is used to implement a 'drain' of currently + * running ioctls. The read semaphore is taken at the beginning of each + * ioctl thread and released upon concluding execution. Additionally the + * semaphore should be released and then reacquired in any ioctl execution + * path which will wait for an event to occur that is outside the scope of + * the ioctl (i.e. an adapter reset). To drain the ioctls currently running, + * a thread simply needs to acquire the write semaphore. + * * Return: 0 on success, -errno on failure */ int cxlflash_ioctl(struct scsi_device *sdev, int cmd, void __user *arg) @@ -2001,6 +2019,9 @@ int cxlflash_ioctl(struct scsi_device *sdev, int cmd, void __user *arg) {sizeof(struct dk_cxlflash_clone), (sioctl)cxlflash_disk_clone}, }; + /* Hold read semaphore so we can drain if needed */ + down_read(&cfg->ioctl_rwsem); + /* Restrict command set to physical support only for internal LUN */ if (afu->internal_lun) switch (cmd) { @@ -2082,6 +2103,7 @@ int cxlflash_ioctl(struct scsi_device *sdev, int cmd, void __user *arg) /* fall through to exit */ cxlflash_ioctl_exit: + up_read(&cfg->ioctl_rwsem); if (unlikely(rc && known_ioctl)) dev_err(dev, "%s: ioctl %s (%08X) on dev(%d/%d/%d/%llu) " "returned rc %d\n", __func__,