From owner-svn-src-stable-9@FreeBSD.ORG Tue May 29 23:10:10 2012 Return-Path: Delivered-To: svn-src-stable-9@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id CE3AF106566B; Tue, 29 May 2012 23:10:10 +0000 (UTC) (envelope-from jimharris@FreeBSD.org) Received: from svn.freebsd.org (svn.freebsd.org [IPv6:2001:4f8:fff6::2c]) by mx1.freebsd.org (Postfix) with ESMTP id B78608FC0C; Tue, 29 May 2012 23:10:10 +0000 (UTC) Received: from svn.freebsd.org (localhost [127.0.0.1]) by svn.freebsd.org (8.14.4/8.14.4) with ESMTP id q4TNAAgn008698; Tue, 29 May 2012 23:10:10 GMT (envelope-from jimharris@svn.freebsd.org) Received: (from jimharris@localhost) by svn.freebsd.org (8.14.4/8.14.4/Submit) id q4TNAAgF008691; Tue, 29 May 2012 23:10:10 GMT (envelope-from jimharris@svn.freebsd.org) Message-Id: <201205292310.q4TNAAgF008691@svn.freebsd.org> From: Jim Harris Date: Tue, 29 May 2012 23:10:10 +0000 (UTC) To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-9@freebsd.org X-SVN-Group: stable-9 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Cc: Subject: svn commit: r236263 - stable/9/sys/dev/isci X-BeenThere: svn-src-stable-9@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: SVN commit messages for only the 9-stable src tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 29 May 2012 23:10:11 -0000 Author: jimharris Date: Tue May 29 23:10:10 2012 New Revision: 236263 URL: http://svn.freebsd.org/changeset/base/236263 Log: MFC r234106, r235751: Queue CCBs internally instead of using CAM_REQUEUE_REQ status. This fixes problem where userspace apps such as smartctl fail due to CAM_REQUEUE_REQ status getting returned when tagged commands are outstanding when smartctl sends its I/O using the pass(4) interface. Sponsored by: Intel Modified: stable/9/sys/dev/isci/isci.h stable/9/sys/dev/isci/isci_controller.c stable/9/sys/dev/isci/isci_interrupt.c stable/9/sys/dev/isci/isci_io_request.c stable/9/sys/dev/isci/isci_remote_device.c Directory Properties: stable/9/sys/ (props changed) stable/9/sys/dev/ (props changed) Modified: stable/9/sys/dev/isci/isci.h ============================================================================== --- stable/9/sys/dev/isci/isci.h Tue May 29 22:28:46 2012 (r236262) +++ stable/9/sys/dev/isci/isci.h Tue May 29 23:10:10 2012 (r236263) @@ -30,6 +30,9 @@ * $FreeBSD$ */ +#ifndef _ISCI_H +#define _ISCI_H + #include #include #include @@ -86,6 +89,31 @@ struct ISCI_REMOTE_DEVICE { BOOL is_resetting; uint32_t frozen_lun_mask; SCI_FAST_LIST_ELEMENT_T pending_device_reset_element; + + /* + * This queue maintains CCBs that have been returned with + * SCI_IO_FAILURE_INVALID_STATE from the SCI layer. These CCBs + * need to be retried, but we cannot return CAM_REQUEUE_REQ because + * this status gets passed all the way back up to users of the pass(4) + * interface and breaks things like smartctl. So instead, we queue + * these CCBs internally. + */ + TAILQ_HEAD(,ccb_hdr) queued_ccbs; + + /* + * Marker denoting this remote device needs its first queued ccb to + * be retried. + */ + BOOL release_queued_ccb; + + /* + * Points to a CCB in the queue that is currently being processed by + * SCIL. This allows us to keep in flight CCBs in the queue so as to + * maintain ordering (i.e. in case we retry an I/O and then find out + * it needs to be retried again - it just keeps its same place in the + * queue. + */ + union ccb * queued_ccb_in_progress; }; struct ISCI_DOMAIN { @@ -125,6 +153,7 @@ struct ISCI_CONTROLLER BOOL has_been_scanned; uint32_t initial_discovery_mask; BOOL is_frozen; + BOOL release_queued_ccbs; uint8_t *remote_device_memory; struct ISCI_MEMORY cached_controller_memory; struct ISCI_MEMORY uncached_controller_memory; @@ -290,6 +319,8 @@ int isci_controller_attach_to_cam(struct void isci_controller_start(void *controller); +void isci_controller_release_queued_ccbs(struct ISCI_CONTROLLER *controller); + void isci_domain_construct(struct ISCI_DOMAIN *domain, uint32_t domain_index, struct ISCI_CONTROLLER *controller); @@ -300,3 +331,5 @@ void isci_log_message(uint32_t verbosity char *log_message, ...); extern uint32_t g_isci_debug_level; + +#endif /* #ifndef _ISCI_H */ Modified: stable/9/sys/dev/isci/isci_controller.c ============================================================================== --- stable/9/sys/dev/isci/isci_controller.c Tue May 29 22:28:46 2012 (r236262) +++ stable/9/sys/dev/isci/isci_controller.c Tue May 29 23:10:10 2012 (r236263) @@ -201,6 +201,7 @@ void isci_controller_construct(struct IS controller->is_started = FALSE; controller->is_frozen = FALSE; + controller->release_queued_ccbs = FALSE; controller->sim = NULL; controller->initial_discovery_mask = 0; @@ -430,6 +431,9 @@ int isci_controller_allocate_memory(stru remote_device->frozen_lun_mask = 0; sci_fast_list_element_init(remote_device, &remote_device->pending_device_reset_element); + TAILQ_INIT(&remote_device->queued_ccbs); + remote_device->release_queued_ccb = FALSE; + remote_device->queued_ccb_in_progress = NULL; /* * For the first SCI_MAX_DOMAINS device objects, do not put @@ -693,3 +697,47 @@ void isci_action(struct cam_sim *sim, un } } +/* + * Unfortunately, SCIL doesn't cleanly handle retry conditions. + * CAM_REQUEUE_REQ works only when no one is using the pass(4) interface. So + * when SCIL denotes an I/O needs to be retried (typically because of mixing + * tagged/non-tagged ATA commands, or running out of NCQ slots), we queue + * these I/O internally. Once SCIL completes an I/O to this device, or we get + * a ready notification, we will retry the first I/O on the queue. + * Unfortunately, SCIL also doesn't cleanly handle starting the new I/O within + * the context of the completion handler, so we need to retry these I/O after + * the completion handler is done executing. + */ +void +isci_controller_release_queued_ccbs(struct ISCI_CONTROLLER *controller) +{ + struct ISCI_REMOTE_DEVICE *dev; + struct ccb_hdr *ccb_h; + int dev_idx; + + KASSERT(mtx_owned(&controller->lock), ("controller lock not owned")); + + controller->release_queued_ccbs = FALSE; + for (dev_idx = 0; + dev_idx < SCI_MAX_REMOTE_DEVICES; + dev_idx++) { + + dev = controller->remote_device[dev_idx]; + if (dev != NULL && + dev->release_queued_ccb == TRUE && + dev->queued_ccb_in_progress == NULL) { + dev->release_queued_ccb = FALSE; + ccb_h = TAILQ_FIRST(&dev->queued_ccbs); + + if (ccb_h == NULL) + continue; + + isci_log_message(1, "ISCI", "release %p %x\n", ccb_h, + ((union ccb *)ccb_h)->csio.cdb_io.cdb_bytes[0]); + + dev->queued_ccb_in_progress = (union ccb *)ccb_h; + isci_io_request_execute_scsi_io( + (union ccb *)ccb_h, controller); + } + } +} Modified: stable/9/sys/dev/isci/isci_interrupt.c ============================================================================== --- stable/9/sys/dev/isci/isci_interrupt.c Tue May 29 22:28:46 2012 (r236262) +++ stable/9/sys/dev/isci/isci_interrupt.c Tue May 29 23:10:10 2012 (r236263) @@ -177,6 +177,9 @@ isci_interrupt_legacy_handler(void *arg) if (interrupt_handler(scic_controller_handle)) { mtx_lock(&controller->lock); completion_handler(scic_controller_handle); + if (controller->release_queued_ccbs == TRUE) + isci_controller_release_queued_ccbs( + controller); mtx_unlock(&controller->lock); } } @@ -204,6 +207,13 @@ isci_interrupt_msix_handler(void *arg) if (interrupt_handler(scic_controller_handle)) { mtx_lock(&controller->lock); completion_handler(scic_controller_handle); + /* + * isci_controller_release_queued_ccb() is a relatively + * expensive routine, so we don't call it until the controller + * level flag is set to TRUE. + */ + if (controller->release_queued_ccbs == TRUE) + isci_controller_release_queued_ccbs(controller); mtx_unlock(&controller->lock); } } Modified: stable/9/sys/dev/isci/isci_io_request.c ============================================================================== --- stable/9/sys/dev/isci/isci_io_request.c Tue May 29 22:28:46 2012 (r236262) +++ stable/9/sys/dev/isci/isci_io_request.c Tue May 29 23:10:10 2012 (r236263) @@ -85,7 +85,9 @@ isci_io_request_complete(SCI_CONTROLLER_ struct ISCI_CONTROLLER *isci_controller; struct ISCI_REMOTE_DEVICE *isci_remote_device; union ccb *ccb; + BOOL complete_ccb; + complete_ccb = TRUE; isci_controller = (struct ISCI_CONTROLLER *) sci_object_get_association(scif_controller); isci_remote_device = (struct ISCI_REMOTE_DEVICE *) sci_object_get_association(remote_device); @@ -163,9 +165,7 @@ isci_io_request_complete(SCI_CONTROLLER_ case SCI_IO_FAILURE_INVALID_STATE: case SCI_IO_FAILURE_INSUFFICIENT_RESOURCES: - ccb->ccb_h.status |= CAM_REQUEUE_REQ; - isci_remote_device_freeze_lun_queue(isci_remote_device, - ccb->ccb_h.target_lun); + complete_ccb = FALSE; break; case SCI_IO_FAILURE_INVALID_REMOTE_DEVICE: @@ -189,7 +189,7 @@ isci_io_request_complete(SCI_CONTROLLER_ scif_remote_device_get_max_queue_depth(remote_device); xpt_action((union ccb *)&ccb_relsim); xpt_free_path(path); - ccb->ccb_h.status |= CAM_REQUEUE_REQ; + complete_ccb = FALSE; } break; @@ -209,17 +209,6 @@ isci_io_request_complete(SCI_CONTROLLER_ break; } - if (ccb->ccb_h.status != CAM_REQ_CMP) { - /* ccb will be completed with some type of non-success - * status. So temporarily freeze the queue until the - * upper layers can act on the status. The CAM_DEV_QFRZN - * flag will then release the queue after the status is - * acted upon. - */ - ccb->ccb_h.status |= CAM_DEV_QFRZN; - xpt_freeze_devq(ccb->ccb_h.path, 1); - } - callout_stop(&isci_request->parent.timer); bus_dmamap_sync(isci_request->parent.dma_tag, isci_request->parent.dma_map, @@ -228,20 +217,82 @@ isci_io_request_complete(SCI_CONTROLLER_ bus_dmamap_unload(isci_request->parent.dma_tag, isci_request->parent.dma_map); - if (isci_remote_device->frozen_lun_mask != 0 && - !(ccb->ccb_h.status & CAM_REQUEUE_REQ)) - isci_remote_device_release_device_queue(isci_remote_device); - - xpt_done(ccb); isci_request->ccb = NULL; - if (isci_controller->is_frozen == TRUE) { - isci_controller->is_frozen = FALSE; - xpt_release_simq(isci_controller->sim, TRUE); - } - sci_pool_put(isci_controller->request_pool, (struct ISCI_REQUEST *)isci_request); + + if (complete_ccb) { + if ((ccb->ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP) { + /* ccb will be completed with some type of non-success + * status. So temporarily freeze the queue until the + * upper layers can act on the status. The + * CAM_DEV_QFRZN flag will then release the queue + * after the status is acted upon. + */ + ccb->ccb_h.status |= CAM_DEV_QFRZN; + xpt_freeze_devq(ccb->ccb_h.path, 1); + } + + if (ccb->ccb_h.status & CAM_SIM_QUEUED) { + + KASSERT(ccb == isci_remote_device->queued_ccb_in_progress, + ("multiple internally queued ccbs in flight")); + + TAILQ_REMOVE(&isci_remote_device->queued_ccbs, + &ccb->ccb_h, sim_links.tqe); + ccb->ccb_h.status &= ~CAM_SIM_QUEUED; + + /* + * This CCB that was in the queue was completed, so + * set the in_progress pointer to NULL denoting that + * we can retry another CCB from the queue. We only + * allow one CCB at a time from the queue to be + * in progress so that we can effectively maintain + * ordering. + */ + isci_remote_device->queued_ccb_in_progress = NULL; + } + + if (isci_remote_device->frozen_lun_mask != 0) { + isci_remote_device_release_device_queue(isci_remote_device); + } + + xpt_done(ccb); + + if (isci_controller->is_frozen == TRUE) { + isci_controller->is_frozen = FALSE; + xpt_release_simq(isci_controller->sim, TRUE); + } + } else { + isci_remote_device_freeze_lun_queue(isci_remote_device, + ccb->ccb_h.target_lun); + + if (ccb->ccb_h.status & CAM_SIM_QUEUED) { + + KASSERT(ccb == isci_remote_device->queued_ccb_in_progress, + ("multiple internally queued ccbs in flight")); + + /* + * Do nothing, CCB is already on the device's queue. + * We leave it on the queue, to be retried again + * next time a CCB on this device completes, or we + * get a ready notification for this device. + */ + isci_log_message(1, "ISCI", "already queued %p %x\n", + ccb, ccb->csio.cdb_io.cdb_bytes[0]); + + isci_remote_device->queued_ccb_in_progress = NULL; + + } else { + isci_log_message(1, "ISCI", "queue %p %x\n", ccb, + ccb->csio.cdb_io.cdb_bytes[0]); + ccb->ccb_h.status |= CAM_SIM_QUEUED; + + TAILQ_INSERT_TAIL(&isci_remote_device->queued_ccbs, + &ccb->ccb_h, sim_links.tqe); + } + } } /** Modified: stable/9/sys/dev/isci/isci_remote_device.c ============================================================================== --- stable/9/sys/dev/isci/isci_remote_device.c Tue May 29 22:28:46 2012 (r236262) +++ stable/9/sys/dev/isci/isci_remote_device.c Tue May 29 23:10:10 2012 (r236263) @@ -289,9 +289,26 @@ isci_remote_device_release_lun_queue(str void isci_remote_device_release_device_queue( - struct ISCI_REMOTE_DEVICE *remote_device) + struct ISCI_REMOTE_DEVICE *device) { - lun_id_t lun; - for (lun = 0; lun < ISCI_MAX_LUN; lun++) - isci_remote_device_release_lun_queue(remote_device, lun); + if (TAILQ_EMPTY(&device->queued_ccbs)) { + lun_id_t lun; + + for (lun = 0; lun < ISCI_MAX_LUN; lun++) + isci_remote_device_release_lun_queue(device, lun); + } else { + /* + * We cannot unfreeze the devq, because there are still + * CCBs in our internal queue that need to be processed + * first. Mark this device, and the controller, so that + * the first CCB in this device's internal queue will be + * resubmitted after the current completion context + * unwinds. + */ + device->release_queued_ccb = TRUE; + device->domain->controller->release_queued_ccbs = TRUE; + + isci_log_message(1, "ISCI", "schedule %p for release\n", + device); + } }