From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Williams Subject: [PATCH v3 14/14] isci: remove IDEV_EH hack to disable "discovery-time" ata resets Date: Thu, 05 Jan 2012 17:00:12 -0800 Message-ID: <20120106010012.11464.84924.stgit@localhost6.localdomain6> References: <20120106005634.11464.64030.stgit@localhost6.localdomain6> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: Received: from mga14.intel.com ([143.182.124.37]:35545 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1030239Ab2AFBAO (ORCPT ); Thu, 5 Jan 2012 20:00:14 -0500 In-Reply-To: <20120106005634.11464.64030.stgit@localhost6.localdomain6> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: linux-scsi@vger.kernel.org Cc: linux-ide@vger.kernel.org, Xiangliang Yu Prior to commit 61aaff49 "isci: filter broadcast change notifications during SMP phy resets" we borrowed the MVS_DEV_EH approach from the mvsas driver for preventing ->lldd_I_T_nexus_reset() events during ata discovery. This hack was protecting against the old ->phy_reset() in ata_bus_probe(), but since the conversion to the new error handling this hack is preventing resets from reaching ata devices. Cc: Xiangliang Yu Signed-off-by: Dan Williams --- drivers/scsi/isci/remote_device.c | 1 - drivers/scsi/isci/remote_device.h | 7 +++---- drivers/scsi/isci/task.c | 9 ++++----- 3 files changed, 7 insertions(+), 10 deletions(-) diff --git a/drivers/scsi/isci/remote_device.c b/drivers/scsi/isci/remote_device.c index 967394d..f83be32 100644 --- a/drivers/scsi/isci/remote_device.c +++ b/drivers/scsi/isci/remote_device.c @@ -1304,7 +1304,6 @@ void isci_remote_device_release(struct kref *kref) clear_bit(IDEV_STOP_PENDING, &idev->flags); clear_bit(IDEV_IO_READY, &idev->flags); clear_bit(IDEV_GONE, &idev->flags); - clear_bit(IDEV_EH, &idev->flags); smp_mb__before_clear_bit(); clear_bit(IDEV_ALLOCATED, &idev->flags); wake_up(&ihost->eventq); diff --git a/drivers/scsi/isci/remote_device.h b/drivers/scsi/isci/remote_device.h index 483ee50..98c2801 100644 --- a/drivers/scsi/isci/remote_device.h +++ b/drivers/scsi/isci/remote_device.h @@ -82,10 +82,9 @@ struct isci_remote_device { #define IDEV_START_PENDING 0 #define IDEV_STOP_PENDING 1 #define IDEV_ALLOCATED 2 - #define IDEV_EH 3 - #define IDEV_GONE 4 - #define IDEV_IO_READY 5 - #define IDEV_IO_NCQERROR 6 + #define IDEV_GONE 3 + #define IDEV_IO_READY 4 + #define IDEV_IO_NCQERROR 5 unsigned long flags; struct kref kref; struct isci_port *isci_port; diff --git a/drivers/scsi/isci/task.c b/drivers/scsi/isci/task.c index edaba78..fd9f67c 100644 --- a/drivers/scsi/isci/task.c +++ b/drivers/scsi/isci/task.c @@ -846,7 +846,6 @@ int isci_task_lu_reset(struct domain_device *dev, u8 *lun) ret = TMF_RESP_FUNC_COMPLETE; goto out; } - set_bit(IDEV_EH, &isci_device->flags); /* Send the task management part of the reset. */ if (dev_is_sata(dev)) { @@ -981,9 +980,6 @@ int isci_task_abort_task(struct sas_task *task) "%s: dev = %p, task = %p, old_request == %p\n", __func__, isci_device, task, old_request); - if (isci_device) - set_bit(IDEV_EH, &isci_device->flags); - /* Device reset conditions signalled in task_state_flags are the * responsbility of libsas to observe at the start of the error * handler thread. @@ -1312,7 +1308,10 @@ int isci_task_I_T_nexus_reset(struct domain_device *dev) idev = isci_lookup_device(dev); spin_unlock_irqrestore(&ihost->scic_lock, flags); - if (!idev || !test_bit(IDEV_EH, &idev->flags)) { + if (!idev) { + /* XXX: need to cleanup any ireqs targeting this + * domain_device + */ ret = TMF_RESP_FUNC_COMPLETE; goto out; }