diff options
Diffstat (limited to 'drivers/ata/libata-eh.c')
| -rw-r--r-- | drivers/ata/libata-eh.c | 1847 |
1 files changed, 972 insertions, 875 deletions
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index c69fcce505c0..2586e77ebf45 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -1,35 +1,14 @@ +// SPDX-License-Identifier: GPL-2.0-or-later /* * libata-eh.c - libata error handling * - * Maintained by: Tejun Heo <tj@kernel.org> - * Please ALWAYS copy linux-ide@vger.kernel.org - * on emails. - * * Copyright 2006 Tejun Heo <htejun@gmail.com> * - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation; either version 2, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; see the file COPYING. If not, write to - * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, - * USA. - * - * * libata documentation is available via 'make {ps|pdf}docs', - * as Documentation/DocBook/libata.* + * as Documentation/driver-api/libata.rst * * Hardware documentation available from http://www.t13.org/ and * http://www.sata-io.org/ - * */ #include <linux/kernel.h> @@ -46,6 +25,7 @@ #include <linux/libata.h> +#include <trace/events/libata.h> #include "libata.h" enum { @@ -95,40 +75,47 @@ enum { * represents timeout for that try. The first try can be soft or * hardreset. All others are hardreset if available. In most cases * the first reset w/ 10sec timeout should succeed. Following entries - * are mostly for error handling, hotplug and retarded devices. + * are mostly for error handling, hotplug and those outlier devices that + * take an exceptionally long time to recover from reset. */ -static const unsigned long ata_eh_reset_timeouts[] = { +static const unsigned int ata_eh_reset_timeouts[] = { 10000, /* most drives spin up by 10sec */ 10000, /* > 99% working drives spin up before 20sec */ - 35000, /* give > 30 secs of idleness for retarded devices */ + 35000, /* give > 30 secs of idleness for outlier devices */ 5000, /* and sweet one last chance */ - ULONG_MAX, /* > 1 min has elapsed, give up */ + UINT_MAX, /* > 1 min has elapsed, give up */ }; -static const unsigned long ata_eh_identify_timeouts[] = { +static const unsigned int ata_eh_identify_timeouts[] = { 5000, /* covers > 99% of successes and not too boring on failures */ 10000, /* combined time till here is enough even for media access */ 30000, /* for true idiots */ - ULONG_MAX, + UINT_MAX, +}; + +static const unsigned int ata_eh_revalidate_timeouts[] = { + 15000, /* Some drives are slow to read log pages when waking-up */ + 15000, /* combined time till here is enough even for media access */ + UINT_MAX, }; -static const unsigned long ata_eh_flush_timeouts[] = { +static const unsigned int ata_eh_flush_timeouts[] = { 15000, /* be generous with flush */ 15000, /* ditto */ 30000, /* and even more generous */ - ULONG_MAX, + UINT_MAX, }; -static const unsigned long ata_eh_other_timeouts[] = { +static const unsigned int ata_eh_other_timeouts[] = { 5000, /* same rationale as identify timeout */ 10000, /* ditto */ /* but no merciful 30sec for other commands, it just isn't worth it */ - ULONG_MAX, + UINT_MAX, }; struct ata_eh_cmd_timeout_ent { const u8 *commands; - const unsigned long *timeouts; + const unsigned int *timeouts; }; /* The following table determines timeouts to use for EH internal @@ -148,6 +135,8 @@ static const struct ata_eh_cmd_timeout_ent ata_eh_cmd_timeout_table[ATA_EH_CMD_TIMEOUT_TABLE_SIZE] = { { .commands = CMDS(ATA_CMD_ID_ATA, ATA_CMD_ID_ATAPI), .timeouts = ata_eh_identify_timeouts, }, + { .commands = CMDS(ATA_CMD_READ_LOG_EXT, ATA_CMD_READ_LOG_DMA_EXT), + .timeouts = ata_eh_revalidate_timeouts, }, { .commands = CMDS(ATA_CMD_READ_NATIVE_MAX, ATA_CMD_READ_NATIVE_MAX_EXT), .timeouts = ata_eh_other_timeouts, }, { .commands = CMDS(ATA_CMD_SET_MAX, ATA_CMD_SET_MAX_EXT), @@ -158,6 +147,8 @@ ata_eh_cmd_timeout_table[ATA_EH_CMD_TIMEOUT_TABLE_SIZE] = { .timeouts = ata_eh_other_timeouts, }, { .commands = CMDS(ATA_CMD_FLUSH, ATA_CMD_FLUSH_EXT), .timeouts = ata_eh_flush_timeouts }, + { .commands = CMDS(ATA_CMD_VERIFY), + .timeouts = ata_eh_reset_timeouts }, }; #undef CMDS @@ -173,8 +164,8 @@ static void ata_eh_handle_port_resume(struct ata_port *ap) { } #endif /* CONFIG_PM */ -static void __ata_ehi_pushv_desc(struct ata_eh_info *ehi, const char *fmt, - va_list args) +static __printf(2, 0) void __ata_ehi_pushv_desc(struct ata_eh_info *ehi, + const char *fmt, va_list args) { ehi->desc_len += vscnprintf(ehi->desc + ehi->desc_len, ATA_EH_DESC_LEN - ehi->desc_len, @@ -199,6 +190,7 @@ void __ata_ehi_push_desc(struct ata_eh_info *ehi, const char *fmt, ...) __ata_ehi_pushv_desc(ehi, fmt, args); va_end(args); } +EXPORT_SYMBOL_GPL(__ata_ehi_push_desc); /** * ata_ehi_push_desc - push error description with separator @@ -222,6 +214,7 @@ void ata_ehi_push_desc(struct ata_eh_info *ehi, const char *fmt, ...) __ata_ehi_pushv_desc(ehi, fmt, args); va_end(args); } +EXPORT_SYMBOL_GPL(ata_ehi_push_desc); /** * ata_ehi_clear_desc - clean error description @@ -237,6 +230,7 @@ void ata_ehi_clear_desc(struct ata_eh_info *ehi) ehi->desc[0] = '\0'; ehi->desc_len = 0; } +EXPORT_SYMBOL_GPL(ata_ehi_clear_desc); /** * ata_port_desc - append port description @@ -264,9 +258,9 @@ void ata_port_desc(struct ata_port *ap, const char *fmt, ...) __ata_ehi_pushv_desc(&ap->link.eh_info, fmt, args); va_end(args); } +EXPORT_SYMBOL_GPL(ata_port_desc); #ifdef CONFIG_PCI - /** * ata_port_pbar_desc - append PCI BAR description * @ap: target ATA port @@ -303,7 +297,7 @@ void ata_port_pbar_desc(struct ata_port *ap, int bar, ssize_t offset, ata_port_desc(ap, "%s 0x%llx", name, start + (unsigned long long)offset); } - +EXPORT_SYMBOL_GPL(ata_port_pbar_desc); #endif /* CONFIG_PCI */ static int ata_lookup_timeout_table(u8 cmd) @@ -334,7 +328,7 @@ static int ata_lookup_timeout_table(u8 cmd) * RETURNS: * Determined timeout. */ -unsigned long ata_internal_cmd_timeout(struct ata_device *dev, u8 cmd) +unsigned int ata_internal_cmd_timeout(struct ata_device *dev, u8 cmd) { struct ata_eh_context *ehc = &dev->link->eh_context; int ent = ata_lookup_timeout_table(cmd); @@ -369,7 +363,7 @@ void ata_internal_cmd_timed_out(struct ata_device *dev, u8 cmd) return; idx = ehc->cmd_timeout_idx[dev->devno][ent]; - if (ata_eh_cmd_timeout_table[ent].timeouts[idx + 1] != ULONG_MAX) + if (ata_eh_cmd_timeout_table[ent].timeouts[idx + 1] != UINT_MAX) ehc->cmd_timeout_idx[dev->devno][ent]++; } @@ -498,54 +492,19 @@ void ata_eh_release(struct ata_port *ap) mutex_unlock(&ap->host->eh_mutex); } -/** - * ata_scsi_timed_out - SCSI layer time out callback - * @cmd: timed out SCSI command - * - * Handles SCSI layer timeout. We race with normal completion of - * the qc for @cmd. If the qc is already gone, we lose and let - * the scsi command finish (EH_HANDLED). Otherwise, the qc has - * timed out and EH should be invoked. Prevent ata_qc_complete() - * from finishing it by setting EH_SCHEDULED and return - * EH_NOT_HANDLED. - * - * TODO: kill this function once old EH is gone. - * - * LOCKING: - * Called from timer context - * - * RETURNS: - * EH_HANDLED or EH_NOT_HANDLED - */ -enum blk_eh_timer_return ata_scsi_timed_out(struct scsi_cmnd *cmd) +static void ata_eh_dev_disable(struct ata_device *dev) { - struct Scsi_Host *host = cmd->device->host; - struct ata_port *ap = ata_shost_to_port(host); - unsigned long flags; - struct ata_queued_cmd *qc; - enum blk_eh_timer_return ret; - - DPRINTK("ENTER\n"); - - if (ap->ops->error_handler) { - ret = BLK_EH_NOT_HANDLED; - goto out; - } + ata_acpi_on_disable(dev); + ata_down_xfermask_limit(dev, ATA_DNXFER_FORCE_PIO0 | ATA_DNXFER_QUIET); + dev->class++; - ret = BLK_EH_HANDLED; - spin_lock_irqsave(ap->lock, flags); - qc = ata_qc_from_tag(ap, ap->link.active_tag); - if (qc) { - WARN_ON(qc->scsicmd != cmd); - qc->flags |= ATA_QCFLAG_EH_SCHEDULED; - qc->err_mask |= AC_ERR_TIMEOUT; - ret = BLK_EH_NOT_HANDLED; - } - spin_unlock_irqrestore(ap->lock, flags); + /* + * From now till the next successful probe, ering is used to + * track probe failures. Clear accumulated device error info. + */ + ata_ering_clear(&dev->ering); - out: - DPRINTK("EXIT, ret=%d\n", ret); - return ret; + ata_dev_free_resources(dev); } static void ata_eh_unload(struct ata_port *ap) @@ -554,13 +513,25 @@ static void ata_eh_unload(struct ata_port *ap) struct ata_device *dev; unsigned long flags; - /* Restore SControl IPM and SPD for the next driver and + /* + * Unless we are restarting, transition all enabled devices to + * standby power mode. + */ + if (system_state != SYSTEM_RESTART) { + ata_for_each_link(link, ap, PMP_FIRST) { + ata_for_each_dev(dev, link, ENABLED) + ata_dev_power_set_standby(dev); + } + } + + /* + * Restore SControl IPM and SPD for the next driver and * disable attached devices. */ ata_for_each_link(link, ap, PMP_FIRST) { sata_scr_write(link, SCR_CONTROL, link->saved_scontrol & 0xff0); - ata_for_each_dev(dev, link, ALL) - ata_dev_disable(dev); + ata_for_each_dev(dev, link, ENABLED) + ata_eh_dev_disable(dev); } /* freeze and set UNLOADED */ @@ -591,8 +562,6 @@ void ata_scsi_error(struct Scsi_Host *host) unsigned long flags; LIST_HEAD(eh_work_q); - DPRINTK("ENTER\n"); - spin_lock_irqsave(host->host_lock, flags); list_splice_init(&host->eh_cmd_q, &eh_work_q); spin_unlock_irqrestore(host->host_lock, flags); @@ -604,9 +573,8 @@ void ata_scsi_error(struct Scsi_Host *host) ata_scsi_port_error_handler(host, ap); /* finish or retry handled scmd's and clean up */ - WARN_ON(host->host_failed || !list_empty(&eh_work_q)); + WARN_ON(!list_empty(&eh_work_q)); - DPRINTK("EXIT\n"); } /** @@ -624,17 +592,20 @@ void ata_scsi_cmd_error_handler(struct Scsi_Host *host, struct ata_port *ap, { int i; unsigned long flags; + struct scsi_cmnd *scmd, *tmp; + int nr_timedout = 0; /* make sure sff pio task is not running */ ata_sff_flush_pio_task(ap); /* synchronize with host lock and sort out timeouts */ - /* For new EH, all qcs are finished in one of three ways - + /* + * For EH, all qcs are finished in one of three ways - * normal completion, error completion, and SCSI timeout. * Both completions can race against SCSI timeout. When normal * completion wins, the qc never reaches EH. When error - * completion wins, the qc has ATA_QCFLAG_FAILED set. + * completion wins, the qc has ATA_QCFLAG_EH set. * * When SCSI timeout wins, things are a bit more complex. * Normal or error completion can occur after the timeout but @@ -642,68 +613,71 @@ void ata_scsi_cmd_error_handler(struct Scsi_Host *host, struct ata_port *ap, * completions are honored. A scmd is determined to have * timed out iff its associated qc is active and not failed. */ - if (ap->ops->error_handler) { - struct scsi_cmnd *scmd, *tmp; - int nr_timedout = 0; - - spin_lock_irqsave(ap->lock, flags); - - /* This must occur under the ap->lock as we don't want - a polled recovery to race the real interrupt handler - - The lost_interrupt handler checks for any completed but - non-notified command and completes much like an IRQ handler. + spin_lock_irqsave(ap->lock, flags); - We then fall into the error recovery code which will treat - this as if normal completion won the race */ + /* + * This must occur under the ap->lock as we don't want + * a polled recovery to race the real interrupt handler + * + * The lost_interrupt handler checks for any completed but + * non-notified command and completes much like an IRQ handler. + * + * We then fall into the error recovery code which will treat + * this as if normal completion won the race + */ + if (ap->ops->lost_interrupt) + ap->ops->lost_interrupt(ap); - if (ap->ops->lost_interrupt) - ap->ops->lost_interrupt(ap); + list_for_each_entry_safe(scmd, tmp, eh_work_q, eh_entry) { + struct ata_queued_cmd *qc; - list_for_each_entry_safe(scmd, tmp, eh_work_q, eh_entry) { - struct ata_queued_cmd *qc; + /* + * If the scmd was added to EH, via ata_qc_schedule_eh() -> + * scsi_timeout() -> scsi_eh_scmd_add(), scsi_timeout() will + * have set DID_TIME_OUT (since libata does not have an abort + * handler). Thus, to clear DID_TIME_OUT, clear the host byte. + */ + set_host_byte(scmd, DID_OK); - for (i = 0; i < ATA_MAX_QUEUE; i++) { - qc = __ata_qc_from_tag(ap, i); - if (qc->flags & ATA_QCFLAG_ACTIVE && - qc->scsicmd == scmd) - break; - } + ata_qc_for_each_raw(ap, qc, i) { + if (qc->flags & ATA_QCFLAG_ACTIVE && + qc->scsicmd == scmd) + break; + } - if (i < ATA_MAX_QUEUE) { - /* the scmd has an associated qc */ - if (!(qc->flags & ATA_QCFLAG_FAILED)) { - /* which hasn't failed yet, timeout */ - qc->err_mask |= AC_ERR_TIMEOUT; - qc->flags |= ATA_QCFLAG_FAILED; - nr_timedout++; - } - } else { - /* Normal completion occurred after - * SCSI timeout but before this point. - * Successfully complete it. - */ - scmd->retries = scmd->allowed; - scsi_eh_finish_cmd(scmd, &ap->eh_done_q); + if (i < ATA_MAX_QUEUE) { + /* the scmd has an associated qc */ + if (!(qc->flags & ATA_QCFLAG_EH)) { + /* which hasn't failed yet, timeout */ + set_host_byte(scmd, DID_TIME_OUT); + qc->err_mask |= AC_ERR_TIMEOUT; + qc->flags |= ATA_QCFLAG_EH; + nr_timedout++; } + } else { + /* Normal completion occurred after + * SCSI timeout but before this point. + * Successfully complete it. + */ + scmd->retries = scmd->allowed; + scsi_eh_finish_cmd(scmd, &ap->eh_done_q); } + } - /* If we have timed out qcs. They belong to EH from - * this point but the state of the controller is - * unknown. Freeze the port to make sure the IRQ - * handler doesn't diddle with those qcs. This must - * be done atomically w.r.t. setting QCFLAG_FAILED. - */ - if (nr_timedout) - __ata_port_freeze(ap); - - spin_unlock_irqrestore(ap->lock, flags); + /* + * If we have timed out qcs. They belong to EH from + * this point but the state of the controller is + * unknown. Freeze the port to make sure the IRQ + * handler doesn't diddle with those qcs. This must + * be done atomically w.r.t. setting ATA_QCFLAG_EH. + */ + if (nr_timedout) + __ata_port_freeze(ap); - /* initialize eh_tries */ - ap->eh_tries = ATA_EH_MAX_TRIES; - } else - spin_unlock_wait(ap->lock); + /* initialize eh_tries */ + ap->eh_tries = ATA_EH_MAX_TRIES; + spin_unlock_irqrestore(ap->lock, flags); } EXPORT_SYMBOL(ata_scsi_cmd_error_handler); @@ -718,103 +692,105 @@ EXPORT_SYMBOL(ata_scsi_cmd_error_handler); void ata_scsi_port_error_handler(struct Scsi_Host *host, struct ata_port *ap) { unsigned long flags; + struct ata_link *link; - /* invoke error handler */ - if (ap->ops->error_handler) { - struct ata_link *link; - - /* acquire EH ownership */ - ata_eh_acquire(ap); + /* acquire EH ownership */ + ata_eh_acquire(ap); repeat: - /* kill fast drain timer */ - del_timer_sync(&ap->fastdrain_timer); + /* kill fast drain timer */ + timer_delete_sync(&ap->fastdrain_timer); - /* process port resume request */ - ata_eh_handle_port_resume(ap); + /* process port resume request */ + ata_eh_handle_port_resume(ap); - /* fetch & clear EH info */ - spin_lock_irqsave(ap->lock, flags); + /* fetch & clear EH info */ + spin_lock_irqsave(ap->lock, flags); - ata_for_each_link(link, ap, HOST_FIRST) { - struct ata_eh_context *ehc = &link->eh_context; - struct ata_device *dev; + ata_for_each_link(link, ap, HOST_FIRST) { + struct ata_eh_context *ehc = &link->eh_context; + struct ata_device *dev; + + memset(&link->eh_context, 0, sizeof(link->eh_context)); + link->eh_context.i = link->eh_info; + memset(&link->eh_info, 0, sizeof(link->eh_info)); - memset(&link->eh_context, 0, sizeof(link->eh_context)); - link->eh_context.i = link->eh_info; - memset(&link->eh_info, 0, sizeof(link->eh_info)); + ata_for_each_dev(dev, link, ENABLED) { + int devno = dev->devno; - ata_for_each_dev(dev, link, ENABLED) { - int devno = dev->devno; + ehc->saved_xfer_mode[devno] = dev->xfer_mode; + if (ata_ncq_enabled(dev)) + ehc->saved_ncq_enabled |= 1 << devno; - ehc->saved_xfer_mode[devno] = dev->xfer_mode; - if (ata_ncq_enabled(dev)) - ehc->saved_ncq_enabled |= 1 << devno; + /* If we are resuming, wake up the device */ + if (ap->pflags & ATA_PFLAG_RESUMING) { + dev->flags |= ATA_DFLAG_RESUMING; + ehc->i.dev_action[devno] |= ATA_EH_SET_ACTIVE; } } + } - ap->pflags |= ATA_PFLAG_EH_IN_PROGRESS; - ap->pflags &= ~ATA_PFLAG_EH_PENDING; - ap->excl_link = NULL; /* don't maintain exclusion over EH */ + ap->pflags |= ATA_PFLAG_EH_IN_PROGRESS; + ap->pflags &= ~ATA_PFLAG_EH_PENDING; + ap->excl_link = NULL; /* don't maintain exclusion over EH */ - spin_unlock_irqrestore(ap->lock, flags); + spin_unlock_irqrestore(ap->lock, flags); - /* invoke EH, skip if unloading or suspended */ - if (!(ap->pflags & (ATA_PFLAG_UNLOADING | ATA_PFLAG_SUSPENDED))) - ap->ops->error_handler(ap); - else { - /* if unloading, commence suicide */ - if ((ap->pflags & ATA_PFLAG_UNLOADING) && - !(ap->pflags & ATA_PFLAG_UNLOADED)) - ata_eh_unload(ap); - ata_eh_finish(ap); - } + /* invoke EH, skip if unloading or suspended */ + if (!(ap->pflags & (ATA_PFLAG_UNLOADING | ATA_PFLAG_SUSPENDED))) + ap->ops->error_handler(ap); + else { + /* if unloading, commence suicide */ + if ((ap->pflags & ATA_PFLAG_UNLOADING) && + !(ap->pflags & ATA_PFLAG_UNLOADED)) + ata_eh_unload(ap); + ata_eh_finish(ap); + } - /* process port suspend request */ - ata_eh_handle_port_suspend(ap); + /* process port suspend request */ + ata_eh_handle_port_suspend(ap); - /* Exception might have happened after ->error_handler - * recovered the port but before this point. Repeat - * EH in such case. - */ - spin_lock_irqsave(ap->lock, flags); + /* + * Exception might have happened after ->error_handler recovered the + * port but before this point. Repeat EH in such case. + */ + spin_lock_irqsave(ap->lock, flags); - if (ap->pflags & ATA_PFLAG_EH_PENDING) { - if (--ap->eh_tries) { - spin_unlock_irqrestore(ap->lock, flags); - goto repeat; - } - ata_port_err(ap, - "EH pending after %d tries, giving up\n", - ATA_EH_MAX_TRIES); - ap->pflags &= ~ATA_PFLAG_EH_PENDING; + if (ap->pflags & ATA_PFLAG_EH_PENDING) { + if (--ap->eh_tries) { + spin_unlock_irqrestore(ap->lock, flags); + goto repeat; } + ata_port_err(ap, + "EH pending after %d tries, giving up\n", + ATA_EH_MAX_TRIES); + ap->pflags &= ~ATA_PFLAG_EH_PENDING; + } - /* this run is complete, make sure EH info is clear */ - ata_for_each_link(link, ap, HOST_FIRST) - memset(&link->eh_info, 0, sizeof(link->eh_info)); + /* this run is complete, make sure EH info is clear */ + ata_for_each_link(link, ap, HOST_FIRST) + memset(&link->eh_info, 0, sizeof(link->eh_info)); - /* end eh (clear host_eh_scheduled) while holding - * ap->lock such that if exception occurs after this - * point but before EH completion, SCSI midlayer will - * re-initiate EH. - */ - ap->ops->end_eh(ap); + /* + * end eh (clear host_eh_scheduled) while holding ap->lock such that if + * exception occurs after this point but before EH completion, SCSI + * midlayer will re-initiate EH. + */ + ap->ops->end_eh(ap); - spin_unlock_irqrestore(ap->lock, flags); - ata_eh_release(ap); - } else { - WARN_ON(ata_qc_from_tag(ap, ap->link.active_tag) == NULL); - ap->ops->eng_timeout(ap); - } + spin_unlock_irqrestore(ap->lock, flags); + ata_eh_release(ap); scsi_eh_flush_done_q(&ap->eh_done_q); /* clean up */ spin_lock_irqsave(ap->lock, flags); + ap->pflags &= ~ATA_PFLAG_RESUMING; + if (ap->pflags & ATA_PFLAG_LOADING) ap->pflags &= ~ATA_PFLAG_LOADING; - else if (ap->pflags & ATA_PFLAG_SCSI_HOTPLUG) + else if ((ap->pflags & ATA_PFLAG_SCSI_HOTPLUG) && + !(ap->flags & ATA_FLAG_SAS_HOST)) schedule_delayed_work(&ap->hotplug_task, 0); if (ap->pflags & ATA_PFLAG_RECOVERED) @@ -847,7 +823,7 @@ void ata_port_wait_eh(struct ata_port *ap) retry: spin_lock_irqsave(ap->lock, flags); - while (ap->pflags & (ATA_PFLAG_EH_PENDING | ATA_PFLAG_EH_IN_PROGRESS)) { + while (ata_port_eh_scheduled(ap)) { prepare_to_wait(&ap->eh_wait_q, &wait, TASK_UNINTERRUPTIBLE); spin_unlock_irqrestore(ap->lock, flags); schedule(); @@ -865,24 +841,26 @@ void ata_port_wait_eh(struct ata_port *ap) } EXPORT_SYMBOL_GPL(ata_port_wait_eh); -static int ata_eh_nr_in_flight(struct ata_port *ap) +static unsigned int ata_eh_nr_in_flight(struct ata_port *ap) { + struct ata_queued_cmd *qc; unsigned int tag; - int nr = 0; + unsigned int nr = 0; /* count only non-internal commands */ - for (tag = 0; tag < ATA_MAX_QUEUE - 1; tag++) - if (ata_qc_from_tag(ap, tag)) + ata_qc_for_each(ap, qc, tag) { + if (qc) nr++; + } return nr; } -void ata_eh_fastdrain_timerfn(unsigned long arg) +void ata_eh_fastdrain_timerfn(struct timer_list *t) { - struct ata_port *ap = (void *)arg; + struct ata_port *ap = timer_container_of(ap, t, fastdrain_timer); unsigned long flags; - int cnt; + unsigned int cnt; spin_lock_irqsave(ap->lock, flags); @@ -893,13 +871,13 @@ void ata_eh_fastdrain_timerfn(unsigned long arg) goto out_unlock; if (cnt == ap->fastdrain_cnt) { + struct ata_queued_cmd *qc; unsigned int tag; /* No progress during the last interval, tag all * in-flight qcs as timed out and freeze the port. */ - for (tag = 0; tag < ATA_MAX_QUEUE - 1; tag++) { - struct ata_queued_cmd *qc = ata_qc_from_tag(ap, tag); + ata_qc_for_each(ap, qc, tag) { if (qc) qc->err_mask |= AC_ERR_TIMEOUT; } @@ -929,9 +907,9 @@ void ata_eh_fastdrain_timerfn(unsigned long arg) * LOCKING: * spin_lock_irqsave(host lock) */ -static void ata_eh_set_pending(struct ata_port *ap, int fastdrain) +static void ata_eh_set_pending(struct ata_port *ap, bool fastdrain) { - int cnt; + unsigned int cnt; /* already scheduled? */ if (ap->pflags & ATA_PFLAG_EH_PENDING) @@ -967,22 +945,16 @@ static void ata_eh_set_pending(struct ata_port *ap, int fastdrain) void ata_qc_schedule_eh(struct ata_queued_cmd *qc) { struct ata_port *ap = qc->ap; - struct request_queue *q = qc->scsicmd->device->request_queue; - unsigned long flags; - - WARN_ON(!ap->ops->error_handler); - qc->flags |= ATA_QCFLAG_FAILED; - ata_eh_set_pending(ap, 1); + qc->flags |= ATA_QCFLAG_EH; + ata_eh_set_pending(ap, true); /* The following will fail if timeout has already expired. * ata_scsi_error() takes care of such scmds on EH entry. - * Note that ATA_QCFLAG_FAILED is unconditionally set after + * Note that ATA_QCFLAG_EH is unconditionally set after * this function completes. */ - spin_lock_irqsave(q->queue_lock, flags); - blk_abort_request(qc->scsicmd->request); - spin_unlock_irqrestore(q->queue_lock, flags); + blk_abort_request(scsi_cmd_to_rq(qc->scsicmd)); } /** @@ -994,15 +966,13 @@ void ata_qc_schedule_eh(struct ata_queued_cmd *qc) */ void ata_std_sched_eh(struct ata_port *ap) { - WARN_ON(!ap->ops->error_handler); - if (ap->pflags & ATA_PFLAG_INITIALIZING) return; - ata_eh_set_pending(ap, 1); + ata_eh_set_pending(ap, true); scsi_schedule_eh(ap->scsi_host); - DPRINTK("port EH scheduled\n"); + trace_ata_std_sched_eh(ap); } EXPORT_SYMBOL_GPL(ata_std_sched_eh); @@ -1042,21 +1012,20 @@ void ata_port_schedule_eh(struct ata_port *ap) /* see: ata_std_sched_eh, unless you know better */ ap->ops->sched_eh(ap); } +EXPORT_SYMBOL_GPL(ata_port_schedule_eh); static int ata_do_link_abort(struct ata_port *ap, struct ata_link *link) { + struct ata_queued_cmd *qc; int tag, nr_aborted = 0; - WARN_ON(!ap->ops->error_handler); - /* we're gonna abort all commands, no need for fast drain */ - ata_eh_set_pending(ap, 0); - - for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { - struct ata_queued_cmd *qc = ata_qc_from_tag(ap, tag); + ata_eh_set_pending(ap, false); + /* include internal tag in iteration */ + ata_qc_for_each_with_internal(ap, qc, tag) { if (qc && (!link || qc->dev->link == link)) { - qc->flags |= ATA_QCFLAG_FAILED; + qc->flags |= ATA_QCFLAG_EH; ata_qc_complete(qc); nr_aborted++; } @@ -1084,6 +1053,7 @@ int ata_link_abort(struct ata_link *link) { return ata_do_link_abort(link->ap, link); } +EXPORT_SYMBOL_GPL(ata_link_abort); /** * ata_port_abort - abort all qc's on the port @@ -1101,6 +1071,7 @@ int ata_port_abort(struct ata_port *ap) { return ata_do_link_abort(ap, NULL); } +EXPORT_SYMBOL_GPL(ata_port_abort); /** * __ata_port_freeze - freeze port @@ -1122,14 +1093,12 @@ int ata_port_abort(struct ata_port *ap) */ static void __ata_port_freeze(struct ata_port *ap) { - WARN_ON(!ap->ops->error_handler); - if (ap->ops->freeze) ap->ops->freeze(ap); ap->pflags |= ATA_PFLAG_FROZEN; - DPRINTK("ata%u port frozen\n", ap->print_id); + trace_ata_port_freeze(ap); } /** @@ -1148,88 +1117,11 @@ static void __ata_port_freeze(struct ata_port *ap) */ int ata_port_freeze(struct ata_port *ap) { - int nr_aborted; - - WARN_ON(!ap->ops->error_handler); - __ata_port_freeze(ap); - nr_aborted = ata_port_abort(ap); - - return nr_aborted; -} - -/** - * sata_async_notification - SATA async notification handler - * @ap: ATA port where async notification is received - * - * Handler to be called when async notification via SDB FIS is - * received. This function schedules EH if necessary. - * - * LOCKING: - * spin_lock_irqsave(host lock) - * - * RETURNS: - * 1 if EH is scheduled, 0 otherwise. - */ -int sata_async_notification(struct ata_port *ap) -{ - u32 sntf; - int rc; - - if (!(ap->flags & ATA_FLAG_AN)) - return 0; - - rc = sata_scr_read(&ap->link, SCR_NOTIFICATION, &sntf); - if (rc == 0) - sata_scr_write(&ap->link, SCR_NOTIFICATION, sntf); - - if (!sata_pmp_attached(ap) || rc) { - /* PMP is not attached or SNTF is not available */ - if (!sata_pmp_attached(ap)) { - /* PMP is not attached. Check whether ATAPI - * AN is configured. If so, notify media - * change. - */ - struct ata_device *dev = ap->link.device; - - if ((dev->class == ATA_DEV_ATAPI) && - (dev->flags & ATA_DFLAG_AN)) - ata_scsi_media_change_notify(dev); - return 0; - } else { - /* PMP is attached but SNTF is not available. - * ATAPI async media change notification is - * not used. The PMP must be reporting PHY - * status change, schedule EH. - */ - ata_port_schedule_eh(ap); - return 1; - } - } else { - /* PMP is attached and SNTF is available */ - struct ata_link *link; - - /* check and notify ATAPI AN */ - ata_for_each_link(link, ap, EDGE) { - if (!(sntf & (1 << link->pmp))) - continue; - - if ((link->device->class == ATA_DEV_ATAPI) && - (link->device->flags & ATA_DFLAG_AN)) - ata_scsi_media_change_notify(link->device); - } - - /* If PMP is reporting that PHY status of some - * downstream ports has changed, schedule EH. - */ - if (sntf & (1 << SATA_PMP_CTRL_PORT)) { - ata_port_schedule_eh(ap); - return 1; - } - return 0; - } + return ata_port_abort(ap); } +EXPORT_SYMBOL_GPL(ata_port_freeze); /** * ata_eh_freeze_port - EH helper to freeze port @@ -1244,16 +1136,14 @@ void ata_eh_freeze_port(struct ata_port *ap) { unsigned long flags; - if (!ap->ops->error_handler) - return; - spin_lock_irqsave(ap->lock, flags); __ata_port_freeze(ap); spin_unlock_irqrestore(ap->lock, flags); } +EXPORT_SYMBOL_GPL(ata_eh_freeze_port); /** - * ata_port_thaw_port - EH helper to thaw port + * ata_eh_thaw_port - EH helper to thaw port * @ap: ATA port to thaw * * Thaw frozen port @ap. @@ -1265,9 +1155,6 @@ void ata_eh_thaw_port(struct ata_port *ap) { unsigned long flags; - if (!ap->ops->error_handler) - return; - spin_lock_irqsave(ap->lock, flags); ap->pflags &= ~ATA_PFLAG_FROZEN; @@ -1277,7 +1164,7 @@ void ata_eh_thaw_port(struct ata_port *ap) spin_unlock_irqrestore(ap->lock, flags); - DPRINTK("ata%u port thawed\n", ap->print_id); + trace_ata_port_thaw(ap); } static void ata_eh_scsidone(struct scsi_cmnd *scmd) @@ -1322,14 +1209,14 @@ void ata_eh_qc_complete(struct ata_queued_cmd *qc) * should be retried. To be used from EH. * * SCSI midlayer limits the number of retries to scmd->allowed. - * scmd->retries is decremented for commands which get retried + * scmd->allowed is incremented for commands which get retried * due to unrelated failures (qc->err_mask is zero). */ void ata_eh_qc_retry(struct ata_queued_cmd *qc) { struct scsi_cmnd *scmd = qc->scsicmd; - if (!qc->err_mask && scmd->retries) - scmd->retries--; + if (!qc->err_mask) + scmd->allowed++; __ata_eh_qc_complete(qc); } @@ -1347,17 +1234,11 @@ void ata_dev_disable(struct ata_device *dev) if (!ata_dev_enabled(dev)) return; - if (ata_msg_drv(dev->link->ap)) - ata_dev_warn(dev, "disabled\n"); - ata_acpi_on_disable(dev); - ata_down_xfermask_limit(dev, ATA_DNXFER_FORCE_PIO0 | ATA_DNXFER_QUIET); - dev->class++; + ata_dev_warn(dev, "disable device\n"); - /* From now till the next successful probe, ering is used to - * track probe failures. Clear accumulated device error info. - */ - ata_ering_clear(&dev->ering); + ata_eh_dev_disable(dev); } +EXPORT_SYMBOL_GPL(ata_dev_disable); /** * ata_eh_detach_dev - detach ATA device @@ -1375,7 +1256,14 @@ void ata_eh_detach_dev(struct ata_device *dev) struct ata_eh_context *ehc = &link->eh_context; unsigned long flags; - ata_dev_disable(dev); + /* + * If the device is still enabled, transition it to standby power mode + * (i.e. spin down HDDs) and disable it. + */ + if (ata_dev_enabled(dev)) { + ata_dev_power_set_standby(dev); + ata_eh_dev_disable(dev); + } spin_lock_irqsave(ap->lock, flags); @@ -1416,6 +1304,8 @@ void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev, struct ata_eh_context *ehc = &link->eh_context; unsigned long flags; + trace_ata_eh_about_to_do(link, dev ? dev->devno : 0, action); + spin_lock_irqsave(ap->lock, flags); ata_eh_clear_action(link, dev, ehi, action); @@ -1431,7 +1321,7 @@ void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev, /** * ata_eh_done - EH action complete -* @ap: target ATA port + * @link: ATA link for which EH actions are complete * @dev: target ATA dev for per-dev action (can be NULL) * @action: action just completed * @@ -1446,6 +1336,8 @@ void ata_eh_done(struct ata_link *link, struct ata_device *dev, { struct ata_eh_context *ehc = &link->eh_context; + trace_ata_eh_done(link, dev ? dev->devno : 0, action); + ata_eh_clear_action(link, dev, &ehc->i, action); } @@ -1481,132 +1373,134 @@ static const char *ata_err_string(unsigned int err_mask) return "invalid argument"; if (err_mask & AC_ERR_DEV) return "device error"; + if (err_mask & AC_ERR_NCQ) + return "NCQ error"; + if (err_mask & AC_ERR_NODEV_HINT) + return "Polling detection error"; return "unknown error"; } /** - * ata_read_log_page - read a specific log page - * @dev: target device - * @log: log to read - * @page: page to read - * @buf: buffer to store read page - * @sectors: number of sectors to read + * atapi_eh_tur - perform ATAPI TEST_UNIT_READY + * @dev: target ATAPI device + * @r_sense_key: out parameter for sense_key * - * Read log page using READ_LOG_EXT command. + * Perform ATAPI TEST_UNIT_READY. * * LOCKING: - * Kernel thread context (may sleep). + * EH context (may sleep). * * RETURNS: - * 0 on success, AC_ERR_* mask otherwise. + * 0 on success, AC_ERR_* mask on failure. */ -unsigned int ata_read_log_page(struct ata_device *dev, u8 log, - u8 page, void *buf, unsigned int sectors) +unsigned int atapi_eh_tur(struct ata_device *dev, u8 *r_sense_key) { + u8 cdb[ATAPI_CDB_LEN] = { TEST_UNIT_READY, 0, 0, 0, 0, 0 }; struct ata_taskfile tf; unsigned int err_mask; - DPRINTK("read log page - log 0x%x, page 0x%x\n", log, page); - ata_tf_init(dev, &tf); - tf.command = ATA_CMD_READ_LOG_EXT; - tf.lbal = log; - tf.lbam = page; - tf.nsect = sectors; - tf.hob_nsect = sectors >> 8; - tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_LBA48 | ATA_TFLAG_DEVICE; - tf.protocol = ATA_PROT_PIO; - - err_mask = ata_exec_internal(dev, &tf, NULL, DMA_FROM_DEVICE, - buf, sectors * ATA_SECT_SIZE, 0); - - DPRINTK("EXIT, err_mask=%x\n", err_mask); + + tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE; + tf.command = ATA_CMD_PACKET; + tf.protocol = ATAPI_PROT_NODATA; + + err_mask = ata_exec_internal(dev, &tf, cdb, DMA_NONE, NULL, 0, 0); + if (err_mask == AC_ERR_DEV) + *r_sense_key = tf.error >> 4; return err_mask; } /** - * ata_eh_read_log_10h - Read log page 10h for NCQ error details - * @dev: Device to read log page 10h from - * @tag: Resulting tag of the failed command - * @tf: Resulting taskfile registers of the failed command - * - * Read log page 10h to obtain NCQ error details and clear error - * condition. + * ata_eh_decide_disposition - Disposition a qc based on sense data + * @qc: qc to examine + * + * For a regular SCSI command, the SCSI completion callback (scsi_done()) + * will call scsi_complete(), which will call scsi_decide_disposition(), + * which will call scsi_check_sense(). scsi_complete() finally calls + * scsi_finish_command(). This is fine for SCSI, since any eventual sense + * data is usually returned in the completion itself (without invoking SCSI + * EH). However, for a QC, we always need to fetch the sense data + * explicitly using SCSI EH. + * + * A command that is completed via SCSI EH will instead be completed using + * scsi_eh_flush_done_q(), which will call scsi_finish_command() directly + * (without ever calling scsi_check_sense()). + * + * For a command that went through SCSI EH, it is the responsibility of the + * SCSI EH strategy handler to call scsi_decide_disposition(), see e.g. how + * scsi_eh_get_sense() calls scsi_decide_disposition() for SCSI LLDDs that + * do not get the sense data as part of the completion. + * + * Thus, for QC commands that went via SCSI EH, we need to call + * scsi_check_sense() ourselves, similar to how scsi_eh_get_sense() calls + * scsi_decide_disposition(), which calls scsi_check_sense(), in order to + * set the correct SCSI ML byte (if any). * * LOCKING: - * Kernel thread context (may sleep). + * EH context. * * RETURNS: - * 0 on success, -errno otherwise. + * SUCCESS or FAILED or NEEDS_RETRY or ADD_TO_MLQUEUE */ -static int ata_eh_read_log_10h(struct ata_device *dev, - int *tag, struct ata_taskfile *tf) +enum scsi_disposition ata_eh_decide_disposition(struct ata_queued_cmd *qc) { - u8 *buf = dev->link->ap->sector_buf; - unsigned int err_mask; - u8 csum; - int i; - - err_mask = ata_read_log_page(dev, ATA_LOG_SATA_NCQ, 0, buf, 1); - if (err_mask) - return -EIO; - - csum = 0; - for (i = 0; i < ATA_SECT_SIZE; i++) - csum += buf[i]; - if (csum) - ata_dev_warn(dev, "invalid checksum 0x%x on log page 10h\n", - csum); - - if (buf[0] & 0x80) - return -ENOENT; - - *tag = buf[0] & 0x1f; - - tf->command = buf[2]; - tf->feature = buf[3]; - tf->lbal = buf[4]; - tf->lbam = buf[5]; - tf->lbah = buf[6]; - tf->device = buf[7]; - tf->hob_lbal = buf[8]; - tf->hob_lbam = buf[9]; - tf->hob_lbah = buf[10]; - tf->nsect = buf[12]; - tf->hob_nsect = buf[13]; - - return 0; + return scsi_check_sense(qc->scsicmd); } /** - * atapi_eh_tur - perform ATAPI TEST_UNIT_READY - * @dev: target ATAPI device - * @r_sense_key: out parameter for sense_key + * ata_eh_request_sense - perform REQUEST_SENSE_DATA_EXT + * @qc: qc to perform REQUEST_SENSE_SENSE_DATA_EXT to * - * Perform ATAPI TEST_UNIT_READY. + * Perform REQUEST_SENSE_DATA_EXT after the device reported CHECK + * SENSE. This function is an EH helper. * * LOCKING: - * EH context (may sleep). + * Kernel thread context (may sleep). * * RETURNS: - * 0 on success, AC_ERR_* mask on failure. + * true if sense data could be fetched, false otherwise. */ -unsigned int atapi_eh_tur(struct ata_device *dev, u8 *r_sense_key) +static bool ata_eh_request_sense(struct ata_queued_cmd *qc) { - u8 cdb[ATAPI_CDB_LEN] = { TEST_UNIT_READY, 0, 0, 0, 0, 0 }; + struct scsi_cmnd *cmd = qc->scsicmd; + struct ata_device *dev = qc->dev; struct ata_taskfile tf; unsigned int err_mask; - ata_tf_init(dev, &tf); + if (ata_port_is_frozen(qc->ap)) { + ata_dev_warn(dev, "sense data available but port frozen\n"); + return false; + } + + if (!ata_id_sense_reporting_enabled(dev->id)) { + ata_dev_warn(qc->dev, "sense data reporting disabled\n"); + return false; + } + ata_tf_init(dev, &tf); tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE; - tf.command = ATA_CMD_PACKET; - tf.protocol = ATAPI_PROT_NODATA; + tf.flags |= ATA_TFLAG_LBA | ATA_TFLAG_LBA48; + tf.command = ATA_CMD_REQ_SENSE_DATA; + tf.protocol = ATA_PROT_NODATA; - err_mask = ata_exec_internal(dev, &tf, cdb, DMA_NONE, NULL, 0, 0); - if (err_mask == AC_ERR_DEV) - *r_sense_key = tf.feature >> 4; - return err_mask; + err_mask = ata_exec_internal(dev, &tf, NULL, DMA_NONE, NULL, 0, 0); + /* Ignore err_mask; ATA_ERR might be set */ + if (tf.status & ATA_SENSE) { + if (ata_scsi_sense_is_valid(tf.lbah, tf.lbam, tf.lbal)) { + /* Set sense without also setting scsicmd->result */ + scsi_build_sense_buffer(dev->flags & ATA_DFLAG_D_SENSE, + cmd->sense_buffer, tf.lbah, + tf.lbam, tf.lbal); + qc->flags |= ATA_QCFLAG_SENSE_VALID; + return true; + } + } else { + ata_dev_warn(dev, "request sense failed stat %02x emask %x\n", + tf.status, err_mask); + } + + return false; } /** @@ -1632,9 +1526,6 @@ unsigned int atapi_eh_request_sense(struct ata_device *dev, struct ata_port *ap = dev->link->ap; struct ata_taskfile tf; - DPRINTK("ATAPI request sense\n"); - - /* FIXME: is this needed? */ memset(sense_buf, 0, SCSI_SENSE_BUFFERSIZE); /* initialize sense_buf with the error register, @@ -1649,8 +1540,15 @@ unsigned int atapi_eh_request_sense(struct ata_device *dev, tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE; tf.command = ATA_CMD_PACKET; - /* is it pointless to prefer PIO for "safety reasons"? */ - if (ap->flags & ATA_FLAG_PIO_DMA) { + /* + * Do not use DMA if the connected device only supports PIO, even if the + * port prefers PIO commands via DMA. + * + * Ideally, we should call atapi_check_dma() to check if it is safe for + * the LLD to use DMA for REQUEST_SENSE, but we don't have a qc. + * Since we can't check the command, perhaps we should only use pio? + */ + if ((ap->flags & ATA_FLAG_PIO_DMA) && !(dev->flags & ATA_DFLAG_PIO)) { tf.protocol = ATAPI_PROT_DMA; tf.feature |= ATAPI_PKT_DMA; } else { @@ -1713,72 +1611,8 @@ static void ata_eh_analyze_serror(struct ata_link *link) } /** - * ata_eh_analyze_ncq_error - analyze NCQ error - * @link: ATA link to analyze NCQ error for - * - * Read log page 10h, determine the offending qc and acquire - * error status TF. For NCQ device errors, all LLDDs have to do - * is setting AC_ERR_DEV in ehi->err_mask. This function takes - * care of the rest. - * - * LOCKING: - * Kernel thread context (may sleep). - */ -void ata_eh_analyze_ncq_error(struct ata_link *link) -{ - struct ata_port *ap = link->ap; - struct ata_eh_context *ehc = &link->eh_context; - struct ata_device *dev = link->device; - struct ata_queued_cmd *qc; - struct ata_taskfile tf; - int tag, rc; - - /* if frozen, we can't do much */ - if (ap->pflags & ATA_PFLAG_FROZEN) - return; - - /* is it NCQ device error? */ - if (!link->sactive || !(ehc->i.err_mask & AC_ERR_DEV)) - return; - - /* has LLDD analyzed already? */ - for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { - qc = __ata_qc_from_tag(ap, tag); - - if (!(qc->flags & ATA_QCFLAG_FAILED)) - continue; - - if (qc->err_mask) - return; - } - - /* okay, this error is ours */ - memset(&tf, 0, sizeof(tf)); - rc = ata_eh_read_log_10h(dev, &tag, &tf); - if (rc) { - ata_link_err(link, "failed to read log page 10h (errno=%d)\n", - rc); - return; - } - - if (!(link->sactive & (1 << tag))) { - ata_link_err(link, "log page 10h reported inactive tag %d\n", - tag); - return; - } - - /* we've got the perpetrator, condemn it */ - qc = __ata_qc_from_tag(ap, tag); - memcpy(&qc->result_tf, &tf, sizeof(tf)); - qc->result_tf.flags = ATA_TFLAG_ISADDR | ATA_TFLAG_LBA | ATA_TFLAG_LBA48; - qc->err_mask |= AC_ERR_DEV | AC_ERR_NCQ; - ehc->i.err_mask &= ~AC_ERR_DEV; -} - -/** * ata_eh_analyze_tf - analyze taskfile of a failed qc * @qc: qc to analyze - * @tf: Taskfile registers to analyze * * Analyze taskfile of @qc and further determine cause of * failure. This function also requests ATAPI sense data if @@ -1790,51 +1624,81 @@ void ata_eh_analyze_ncq_error(struct ata_link *link) * RETURNS: * Determined recovery action */ -static unsigned int ata_eh_analyze_tf(struct ata_queued_cmd *qc, - const struct ata_taskfile *tf) +static unsigned int ata_eh_analyze_tf(struct ata_queued_cmd *qc) { + const struct ata_taskfile *tf = &qc->result_tf; unsigned int tmp, action = 0; - u8 stat = tf->command, err = tf->feature; + u8 stat = tf->status, err = tf->error; if ((stat & (ATA_BUSY | ATA_DRQ | ATA_DRDY)) != ATA_DRDY) { qc->err_mask |= AC_ERR_HSM; return ATA_EH_RESET; } - if (stat & (ATA_ERR | ATA_DF)) + if (stat & (ATA_ERR | ATA_DF)) { qc->err_mask |= AC_ERR_DEV; - else + /* + * Sense data reporting does not work if the + * device fault bit is set. + */ + if (stat & ATA_DF) + stat &= ~ATA_SENSE; + } else { return 0; + } switch (qc->dev->class) { case ATA_DEV_ATA: + case ATA_DEV_ZAC: + /* + * Fetch the sense data explicitly if: + * -It was a non-NCQ command that failed, or + * -It was a NCQ command that failed, but the sense data + * was not included in the NCQ command error log + * (i.e. NCQ autosense is not supported by the device). + */ + if (!(qc->flags & ATA_QCFLAG_SENSE_VALID) && + (stat & ATA_SENSE) && ata_eh_request_sense(qc)) + set_status_byte(qc->scsicmd, SAM_STAT_CHECK_CONDITION); if (err & ATA_ICRC) qc->err_mask |= AC_ERR_ATA_BUS; - if (err & ATA_UNC) + if (err & (ATA_UNC | ATA_AMNF)) qc->err_mask |= AC_ERR_MEDIA; if (err & ATA_IDNF) qc->err_mask |= AC_ERR_INVALID; break; case ATA_DEV_ATAPI: - if (!(qc->ap->pflags & ATA_PFLAG_FROZEN)) { + if (!ata_port_is_frozen(qc->ap)) { tmp = atapi_eh_request_sense(qc->dev, qc->scsicmd->sense_buffer, - qc->result_tf.feature >> 4); - if (!tmp) { - /* ATA_QCFLAG_SENSE_VALID is used to - * tell atapi_qc_complete() that sense - * data is already valid. - * - * TODO: interpret sense data and set - * appropriate err_mask. - */ + qc->result_tf.error >> 4); + if (!tmp) qc->flags |= ATA_QCFLAG_SENSE_VALID; - } else + else qc->err_mask |= tmp; } } + if (qc->flags & ATA_QCFLAG_SENSE_VALID) { + enum scsi_disposition ret = ata_eh_decide_disposition(qc); + + /* + * SUCCESS here means that the sense code could be + * evaluated and should be passed to the upper layers + * for correct evaluation. + * FAILED means the sense code could not be interpreted + * and the device would need to be reset. + * NEEDS_RETRY and ADD_TO_MLQUEUE means that the + * command would need to be retried. + */ + if (ret == NEEDS_RETRY || ret == ADD_TO_MLQUEUE) { + qc->flags |= ATA_QCFLAG_RETRY; + qc->err_mask |= AC_ERR_OTHER; + } else if (ret != SUCCESS) { + qc->err_mask |= AC_ERR_HSM; + } + } if (qc->err_mask & (AC_ERR_HSM | AC_ERR_TIMEOUT | AC_ERR_ATA_BUS)) action |= ATA_EH_RESET; @@ -1923,7 +1787,7 @@ static int speed_down_verdict_cb(struct ata_ering_entry *ent, void *void_arg) * This is to expedite speed down decisions right after device is * initially configured. * - * The followings are speed down rules. #1 and #2 deal with + * The following are speed down rules. #1 and #2 deal with * DUBIOUS errors. * * 1. If more than one DUBIOUS_ATA_BUS or DUBIOUS_TOUT_HSM errors @@ -2024,9 +1888,7 @@ static unsigned int ata_eh_speed_down(struct ata_device *dev, verdict = ata_eh_speed_down_verdict(dev); /* turn off NCQ? */ - if ((verdict & ATA_EH_SPDN_NCQ_OFF) && - (dev->flags & (ATA_DFLAG_PIO | ATA_DFLAG_NCQ | - ATA_DFLAG_NCQ_OFF)) == ATA_DFLAG_NCQ) { + if ((verdict & ATA_EH_SPDN_NCQ_OFF) && ata_ncq_enabled(dev)) { dev->flags |= ATA_DFLAG_NCQ_OFF; ata_dev_warn(dev, "NCQ disabled due to excessive errors\n"); goto done; @@ -2104,6 +1966,294 @@ static inline int ata_eh_worth_retry(struct ata_queued_cmd *qc) } /** + * ata_eh_quiet - check if we need to be quiet about a command error + * @qc: qc to check + * + * Look at the qc flags anbd its scsi command request flags to determine + * if we need to be quiet about the command failure. + */ +static inline bool ata_eh_quiet(struct ata_queued_cmd *qc) +{ + if (qc->scsicmd && scsi_cmd_to_rq(qc->scsicmd)->rq_flags & RQF_QUIET) + qc->flags |= ATA_QCFLAG_QUIET; + return qc->flags & ATA_QCFLAG_QUIET; +} + +static int ata_eh_get_non_ncq_success_sense(struct ata_link *link) +{ + struct ata_port *ap = link->ap; + struct ata_queued_cmd *qc; + + qc = __ata_qc_from_tag(ap, link->active_tag); + if (!qc) + return -EIO; + + if (!(qc->flags & ATA_QCFLAG_EH) || + !(qc->flags & ATA_QCFLAG_EH_SUCCESS_CMD) || + qc->err_mask) + return -EIO; + + if (!ata_eh_request_sense(qc)) + return -EIO; + + /* + * No point in checking the return value, since the command has already + * completed successfully. + */ + ata_eh_decide_disposition(qc); + + return 0; +} + +static void ata_eh_get_success_sense(struct ata_link *link) +{ + struct ata_eh_context *ehc = &link->eh_context; + struct ata_device *dev = link->device; + struct ata_port *ap = link->ap; + struct ata_queued_cmd *qc; + int tag, ret = 0; + + if (!(ehc->i.dev_action[dev->devno] & ATA_EH_GET_SUCCESS_SENSE)) + return; + + /* if frozen, we can't do much */ + if (ata_port_is_frozen(ap)) { + ata_dev_warn(dev, + "successful sense data available but port frozen\n"); + goto out; + } + + /* + * If the link has sactive set, then we have outstanding NCQ commands + * and have to read the Successful NCQ Commands log to get the sense + * data. Otherwise, we are dealing with a non-NCQ command and use + * request sense ext command to retrieve the sense data. + */ + if (link->sactive) + ret = ata_eh_get_ncq_success_sense(link); + else + ret = ata_eh_get_non_ncq_success_sense(link); + if (ret) + goto out; + + ata_eh_done(link, dev, ATA_EH_GET_SUCCESS_SENSE); + return; + +out: + /* + * If we failed to get sense data for a successful command that ought to + * have sense data, we cannot simply return BLK_STS_OK to user space. + * This is because we can't know if the sense data that we couldn't get + * was actually "DATA CURRENTLY UNAVAILABLE". Reporting such a command + * as success to user space would result in a silent data corruption. + * Thus, add a bogus ABORTED_COMMAND sense data to such commands, such + * that SCSI will report these commands as BLK_STS_IOERR to user space. + */ + ata_qc_for_each_raw(ap, qc, tag) { + if (!(qc->flags & ATA_QCFLAG_EH) || + !(qc->flags & ATA_QCFLAG_EH_SUCCESS_CMD) || + qc->err_mask || + ata_dev_phys_link(qc->dev) != link) + continue; + + /* We managed to get sense for this success command, skip. */ + if (qc->flags & ATA_QCFLAG_SENSE_VALID) + continue; + + /* This success command did not have any sense data, skip. */ + if (!(qc->result_tf.status & ATA_SENSE)) + continue; + + /* This success command had sense data, but we failed to get. */ + ata_scsi_set_sense(dev, qc->scsicmd, ABORTED_COMMAND, 0, 0); + qc->flags |= ATA_QCFLAG_SENSE_VALID; + } + ata_eh_done(link, dev, ATA_EH_GET_SUCCESS_SENSE); +} + +/* + * Check if a link is established. This is a relaxed version of + * ata_phys_link_online() which accounts for the fact that this is potentially + * called after changing the link power management policy, which may not be + * reflected immediately in the SStatus register (e.g., we may still be seeing + * the PHY in partial, slumber or devsleep Partial power management state. + * So check that: + * - A device is still present, that is, DET is 1h (Device presence detected + * but Phy communication not established) or 3h (Device presence detected and + * Phy communication established) + * - Communication is established, that is, IPM is not 0h, indicating that PHY + * is online or in a low power state. + */ +static bool ata_eh_link_established(struct ata_link *link) +{ + u32 sstatus; + u8 det, ipm; + + /* + * For old IDE/PATA adapters that do not have a valid scr_read method, + * or if reading the SStatus register fails, assume that the device is + * present. Device probe will determine if that is really the case. + */ + if (sata_scr_read(link, SCR_STATUS, &sstatus)) + return true; + + det = sstatus & 0x0f; + ipm = (sstatus >> 8) & 0x0f; + + return (det & 0x01) && ipm; +} + +/** + * ata_eh_link_set_lpm - configure SATA interface power management + * @link: link to configure + * @policy: the link power management policy + * @r_failed_dev: out parameter for failed device + * + * Enable SATA Interface power management. This will enable + * Device Interface Power Management (DIPM) for min_power and + * medium_power_with_dipm policies, and then call driver specific + * callbacks for enabling Host Initiated Power management. + * + * LOCKING: + * EH context. + * + * RETURNS: + * 0 on success, -errno on failure. + */ +static int ata_eh_link_set_lpm(struct ata_link *link, + enum ata_lpm_policy policy, + struct ata_device **r_failed_dev) +{ + struct ata_port *ap = ata_is_host_link(link) ? link->ap : NULL; + struct ata_eh_context *ehc = &link->eh_context; + struct ata_device *dev, *link_dev = NULL, *lpm_dev = NULL; + enum ata_lpm_policy old_policy = link->lpm_policy; + bool host_has_dipm = !(link->ap->flags & ATA_FLAG_NO_DIPM); + unsigned int hints = ATA_LPM_EMPTY | ATA_LPM_HIPM; + unsigned int err_mask; + int rc; + + /* if the link or host doesn't do LPM, noop */ + if (!IS_ENABLED(CONFIG_SATA_HOST) || + (link->flags & ATA_LFLAG_NO_LPM) || (ap && !ap->ops->set_lpm)) + return 0; + + /* + * This function currently assumes that it will never be supplied policy + * ATA_LPM_UNKNOWN. + */ + if (WARN_ON_ONCE(policy == ATA_LPM_UNKNOWN)) + return 0; + + ata_link_dbg(link, "Set LPM policy: %d -> %d\n", old_policy, policy); + + /* + * DIPM is enabled only for ATA_LPM_MIN_POWER, + * ATA_LPM_MIN_POWER_WITH_PARTIAL, and ATA_LPM_MED_POWER_WITH_DIPM, as + * some devices misbehave when the host NACKs transition to SLUMBER. + */ + ata_for_each_dev(dev, link, ENABLED) { + bool dev_has_hipm = ata_id_has_hipm(dev->id); + bool dev_has_dipm = ata_id_has_dipm(dev->id); + + /* find the first enabled and LPM enabled devices */ + if (!link_dev) + link_dev = dev; + + if (!lpm_dev && + (dev_has_hipm || (dev_has_dipm && host_has_dipm))) + lpm_dev = dev; + + hints &= ~ATA_LPM_EMPTY; + if (!dev_has_hipm) + hints &= ~ATA_LPM_HIPM; + + /* disable DIPM before changing link config */ + if (dev_has_dipm) { + err_mask = ata_dev_set_feature(dev, + SETFEATURES_SATA_DISABLE, SATA_DIPM); + if (err_mask && err_mask != AC_ERR_DEV) { + ata_dev_warn(dev, + "failed to disable DIPM, Emask 0x%x\n", + err_mask); + rc = -EIO; + goto fail; + } + } + } + + if (ap) { + rc = ap->ops->set_lpm(link, policy, hints); + if (!rc && ap->slave_link) + rc = ap->ops->set_lpm(ap->slave_link, policy, hints); + } else + rc = sata_pmp_set_lpm(link, policy, hints); + + /* + * Attribute link config failure to the first (LPM) enabled + * device on the link. + */ + if (rc) { + if (rc == -EOPNOTSUPP) { + link->flags |= ATA_LFLAG_NO_LPM; + return 0; + } + dev = lpm_dev ? lpm_dev : link_dev; + goto fail; + } + + /* + * Low level driver acked the transition. Issue DIPM command + * with the new policy set. + */ + link->lpm_policy = policy; + if (ap && ap->slave_link) + ap->slave_link->lpm_policy = policy; + + /* + * Host config updated, enable DIPM if transitioning to + * ATA_LPM_MIN_POWER, ATA_LPM_MIN_POWER_WITH_PARTIAL, or + * ATA_LPM_MED_POWER_WITH_DIPM. + */ + ata_for_each_dev(dev, link, ENABLED) { + bool dev_has_dipm = ata_id_has_dipm(dev->id); + + if (policy >= ATA_LPM_MED_POWER_WITH_DIPM && host_has_dipm && + dev_has_dipm) { + err_mask = ata_dev_set_feature(dev, + SETFEATURES_SATA_ENABLE, SATA_DIPM); + if (err_mask && err_mask != AC_ERR_DEV) { + ata_dev_warn(dev, + "failed to enable DIPM, Emask 0x%x\n", + err_mask); + rc = -EIO; + goto fail; + } + } + } + + link->last_lpm_change = jiffies; + link->flags |= ATA_LFLAG_CHANGED; + + return 0; + +fail: + /* restore the old policy */ + link->lpm_policy = old_policy; + if (ap && ap->slave_link) + ap->slave_link->lpm_policy = old_policy; + + /* if no device or only one more chance is left, disable LPM */ + if (!dev || ehc->tries[dev->devno] <= 2) { + ata_link_warn(link, "disabling LPM on the link\n"); + link->flags |= ATA_LFLAG_NO_LPM; + } + if (r_failed_dev) + *r_failed_dev = dev; + return rc; +} + +/** * ata_eh_link_autopsy - analyze error and determine recovery action * @link: host link to perform autopsy on * @@ -2118,14 +2268,13 @@ static void ata_eh_link_autopsy(struct ata_link *link) { struct ata_port *ap = link->ap; struct ata_eh_context *ehc = &link->eh_context; + struct ata_queued_cmd *qc; struct ata_device *dev; unsigned int all_err_mask = 0, eflags = 0; - int tag; + int tag, nr_failed = 0, nr_quiet = 0; u32 serror; int rc; - DPRINTK("ENTER\n"); - if (ehc->i.flags & ATA_EHI_NO_AUTOPSY) return; @@ -2144,16 +2293,24 @@ static void ata_eh_link_autopsy(struct ata_link *link) /* analyze NCQ failure */ ata_eh_analyze_ncq_error(link); + /* + * Check if this was a successful command that simply needs sense data. + * Since the sense data is not part of the completion, we need to fetch + * it using an additional command. Since this can't be done from irq + * context, the sense data for successful commands are fetched by EH. + */ + ata_eh_get_success_sense(link); + /* any real error trumps AC_ERR_OTHER */ if (ehc->i.err_mask & ~AC_ERR_OTHER) ehc->i.err_mask &= ~AC_ERR_OTHER; all_err_mask |= ehc->i.err_mask; - for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { - struct ata_queued_cmd *qc = __ata_qc_from_tag(ap, tag); - - if (!(qc->flags & ATA_QCFLAG_FAILED) || + ata_qc_for_each_raw(ap, qc, tag) { + if (!(qc->flags & ATA_QCFLAG_EH) || + qc->flags & ATA_QCFLAG_RETRY || + qc->flags & ATA_QCFLAG_EH_SUCCESS_CMD || ata_dev_phys_link(qc->dev) != link) continue; @@ -2161,7 +2318,7 @@ static void ata_eh_link_autopsy(struct ata_link *link) qc->err_mask |= ehc->i.err_mask; /* analyze TF */ - ehc->i.action |= ata_eh_analyze_tf(qc, &qc->result_tf); + ehc->i.action |= ata_eh_analyze_tf(qc); /* DEV errors are probably spurious in case of ATA_BUS error */ if (qc->err_mask & AC_ERR_ATA_BUS) @@ -2172,12 +2329,16 @@ static void ata_eh_link_autopsy(struct ata_link *link) if (qc->err_mask & ~AC_ERR_OTHER) qc->err_mask &= ~AC_ERR_OTHER; - /* SENSE_VALID trumps dev/unknown error and revalidation */ + /* + * SENSE_VALID trumps dev/unknown error and revalidation. Upper + * layers will determine whether the command is worth retrying + * based on the sense data and device class/type. Otherwise, + * determine directly if the command is worth retrying using its + * error mask and flags. + */ if (qc->flags & ATA_QCFLAG_SENSE_VALID) qc->err_mask &= ~(AC_ERR_DEV | AC_ERR_OTHER); - - /* determine whether the command is worth retrying */ - if (ata_eh_worth_retry(qc)) + else if (ata_eh_worth_retry(qc)) qc->flags |= ATA_QCFLAG_RETRY; /* accumulate error info */ @@ -2185,10 +2346,20 @@ static void ata_eh_link_autopsy(struct ata_link *link) all_err_mask |= qc->err_mask; if (qc->flags & ATA_QCFLAG_IO) eflags |= ATA_EFLAG_IS_IO; + trace_ata_eh_link_autopsy_qc(qc); + + /* Count quiet errors */ + if (ata_eh_quiet(qc)) + nr_quiet++; + nr_failed++; } + /* If all failed commands requested silence, then be quiet */ + if (nr_quiet == nr_failed) + ehc->i.flags |= ATA_EHI_QUIET; + /* enforce default EH actions */ - if (ap->pflags & ATA_PFLAG_FROZEN || + if (ata_port_is_frozen(ap) || all_err_mask & (AC_ERR_HSM | AC_ERR_TIMEOUT)) ehc->i.action |= ATA_EH_RESET; else if (((eflags & ATA_EFLAG_IS_IO) && all_err_mask) || @@ -2218,9 +2389,8 @@ static void ata_eh_link_autopsy(struct ata_link *link) if (dev->flags & ATA_DFLAG_DUBIOUS_XFER) eflags |= ATA_EFLAG_DUBIOUS_XFER; ehc->i.action |= ata_eh_speed_down(dev, eflags, all_err_mask); + trace_ata_eh_link_autopsy(dev, ehc->i.action, all_err_mask); } - - DPRINTK("EXIT\n"); } /** @@ -2270,16 +2440,15 @@ void ata_eh_autopsy(struct ata_port *ap) } /** - * ata_get_cmd_descript - get description for ATA command - * @command: ATA command code to get description for + * ata_get_cmd_name - get name for ATA command + * @command: ATA command code to get name for * - * Return a textual description of the given command, or NULL if the - * command is not known. + * Return a textual name of the given command or "unknown" * * LOCKING: * None */ -const char *ata_get_cmd_descript(u8 command) +const char *ata_get_cmd_name(u8 command) { #ifdef CONFIG_ATA_VERBOSE_ERROR static const struct @@ -2288,31 +2457,35 @@ const char *ata_get_cmd_descript(u8 command) const char *text; } cmd_descr[] = { { ATA_CMD_DEV_RESET, "DEVICE RESET" }, - { ATA_CMD_CHK_POWER, "CHECK POWER MODE" }, - { ATA_CMD_STANDBY, "STANDBY" }, - { ATA_CMD_IDLE, "IDLE" }, - { ATA_CMD_EDD, "EXECUTE DEVICE DIAGNOSTIC" }, - { ATA_CMD_DOWNLOAD_MICRO, "DOWNLOAD MICROCODE" }, + { ATA_CMD_CHK_POWER, "CHECK POWER MODE" }, + { ATA_CMD_STANDBY, "STANDBY" }, + { ATA_CMD_IDLE, "IDLE" }, + { ATA_CMD_EDD, "EXECUTE DEVICE DIAGNOSTIC" }, + { ATA_CMD_DOWNLOAD_MICRO, "DOWNLOAD MICROCODE" }, + { ATA_CMD_DOWNLOAD_MICRO_DMA, "DOWNLOAD MICROCODE DMA" }, { ATA_CMD_NOP, "NOP" }, - { ATA_CMD_FLUSH, "FLUSH CACHE" }, - { ATA_CMD_FLUSH_EXT, "FLUSH CACHE EXT" }, - { ATA_CMD_ID_ATA, "IDENTIFY DEVICE" }, - { ATA_CMD_ID_ATAPI, "IDENTIFY PACKET DEVICE" }, - { ATA_CMD_SERVICE, "SERVICE" }, - { ATA_CMD_READ, "READ DMA" }, - { ATA_CMD_READ_EXT, "READ DMA EXT" }, - { ATA_CMD_READ_QUEUED, "READ DMA QUEUED" }, - { ATA_CMD_READ_STREAM_EXT, "READ STREAM EXT" }, + { ATA_CMD_FLUSH, "FLUSH CACHE" }, + { ATA_CMD_FLUSH_EXT, "FLUSH CACHE EXT" }, + { ATA_CMD_ID_ATA, "IDENTIFY DEVICE" }, + { ATA_CMD_ID_ATAPI, "IDENTIFY PACKET DEVICE" }, + { ATA_CMD_SERVICE, "SERVICE" }, + { ATA_CMD_READ, "READ DMA" }, + { ATA_CMD_READ_EXT, "READ DMA EXT" }, + { ATA_CMD_READ_QUEUED, "READ DMA QUEUED" }, + { ATA_CMD_READ_STREAM_EXT, "READ STREAM EXT" }, { ATA_CMD_READ_STREAM_DMA_EXT, "READ STREAM DMA EXT" }, - { ATA_CMD_WRITE, "WRITE DMA" }, - { ATA_CMD_WRITE_EXT, "WRITE DMA EXT" }, - { ATA_CMD_WRITE_QUEUED, "WRITE DMA QUEUED EXT" }, - { ATA_CMD_WRITE_STREAM_EXT, "WRITE STREAM EXT" }, + { ATA_CMD_WRITE, "WRITE DMA" }, + { ATA_CMD_WRITE_EXT, "WRITE DMA EXT" }, + { ATA_CMD_WRITE_QUEUED, "WRITE DMA QUEUED EXT" }, + { ATA_CMD_WRITE_STREAM_EXT, "WRITE STREAM EXT" }, { ATA_CMD_WRITE_STREAM_DMA_EXT, "WRITE STREAM DMA EXT" }, { ATA_CMD_WRITE_FUA_EXT, "WRITE DMA FUA EXT" }, { ATA_CMD_WRITE_QUEUED_FUA_EXT, "WRITE DMA QUEUED FUA EXT" }, { ATA_CMD_FPDMA_READ, "READ FPDMA QUEUED" }, { ATA_CMD_FPDMA_WRITE, "WRITE FPDMA QUEUED" }, + { ATA_CMD_NCQ_NON_DATA, "NCQ NON-DATA" }, + { ATA_CMD_FPDMA_SEND, "SEND FPDMA QUEUED" }, + { ATA_CMD_FPDMA_RECV, "RECEIVE FPDMA QUEUED" }, { ATA_CMD_PIO_READ, "READ SECTOR(S)" }, { ATA_CMD_PIO_READ_EXT, "READ SECTOR(S) EXT" }, { ATA_CMD_PIO_WRITE, "WRITE SECTOR(S)" }, @@ -2321,7 +2494,7 @@ const char *ata_get_cmd_descript(u8 command) { ATA_CMD_READ_MULTI_EXT, "READ MULTIPLE EXT" }, { ATA_CMD_WRITE_MULTI, "WRITE MULTIPLE" }, { ATA_CMD_WRITE_MULTI_EXT, "WRITE MULTIPLE EXT" }, - { ATA_CMD_WRITE_MULTI_FUA_EXT, "WRITE MULTIPLE FUA EXT" }, + { ATA_CMD_WRITE_MULTI_FUA_EXT, "WRITE MULTIPLE FUA EXT" }, { ATA_CMD_SET_FEATURES, "SET FEATURES" }, { ATA_CMD_SET_MULTI, "SET MULTIPLE MODE" }, { ATA_CMD_VERIFY, "READ VERIFY SECTOR(S)" }, @@ -2338,13 +2511,16 @@ const char *ata_get_cmd_descript(u8 command) { ATA_CMD_READ_LOG_EXT, "READ LOG EXT" }, { ATA_CMD_WRITE_LOG_EXT, "WRITE LOG EXT" }, { ATA_CMD_READ_LOG_DMA_EXT, "READ LOG DMA EXT" }, - { ATA_CMD_WRITE_LOG_DMA_EXT, "WRITE LOG DMA EXT" }, + { ATA_CMD_WRITE_LOG_DMA_EXT, "WRITE LOG DMA EXT" }, + { ATA_CMD_TRUSTED_NONDATA, "TRUSTED NON-DATA" }, { ATA_CMD_TRUSTED_RCV, "TRUSTED RECEIVE" }, - { ATA_CMD_TRUSTED_RCV_DMA, "TRUSTED RECEIVE DMA" }, + { ATA_CMD_TRUSTED_RCV_DMA, "TRUSTED RECEIVE DMA" }, { ATA_CMD_TRUSTED_SND, "TRUSTED SEND" }, - { ATA_CMD_TRUSTED_SND_DMA, "TRUSTED SEND DMA" }, + { ATA_CMD_TRUSTED_SND_DMA, "TRUSTED SEND DMA" }, { ATA_CMD_PMP_READ, "READ BUFFER" }, + { ATA_CMD_PMP_READ_DMA, "READ BUFFER DMA" }, { ATA_CMD_PMP_WRITE, "WRITE BUFFER" }, + { ATA_CMD_PMP_WRITE_DMA, "WRITE BUFFER DMA" }, { ATA_CMD_CONF_OVERLAY, "DEVICE CONFIGURATION OVERLAY" }, { ATA_CMD_SEC_SET_PASS, "SECURITY SET PASSWORD" }, { ATA_CMD_SEC_UNLOCK, "SECURITY UNLOCK" }, @@ -2357,12 +2533,16 @@ const char *ata_get_cmd_descript(u8 command) { ATA_CMD_MEDIA_LOCK, "DOOR LOCK" }, { ATA_CMD_MEDIA_UNLOCK, "DOOR UNLOCK" }, { ATA_CMD_DSM, "DATA SET MANAGEMENT" }, - { ATA_CMD_CHK_MED_CRD_TYP, "CHECK MEDIA CARD TYPE" }, - { ATA_CMD_CFA_REQ_EXT_ERR, "CFA REQUEST EXTENDED ERROR" }, + { ATA_CMD_CHK_MED_CRD_TYP, "CHECK MEDIA CARD TYPE" }, + { ATA_CMD_CFA_REQ_EXT_ERR, "CFA REQUEST EXTENDED ERROR" }, { ATA_CMD_CFA_WRITE_NE, "CFA WRITE SECTORS WITHOUT ERASE" }, { ATA_CMD_CFA_TRANS_SECT, "CFA TRANSLATE SECTOR" }, { ATA_CMD_CFA_ERASE, "CFA ERASE SECTORS" }, - { ATA_CMD_CFA_WRITE_MULT_NE, "CFA WRITE MULTIPLE WITHOUT ERASE" }, + { ATA_CMD_CFA_WRITE_MULT_NE, "CFA WRITE MULTIPLE WITHOUT ERASE" }, + { ATA_CMD_REQ_SENSE_DATA, "REQUEST SENSE DATA EXT" }, + { ATA_CMD_SANITIZE_DEVICE, "SANITIZE DEVICE" }, + { ATA_CMD_ZAC_MGMT_IN, "ZAC MANAGEMENT IN" }, + { ATA_CMD_ZAC_MGMT_OUT, "ZAC MANAGEMENT OUT" }, { ATA_CMD_READ_LONG, "READ LONG (with retries)" }, { ATA_CMD_READ_LONG_ONCE, "READ LONG (without retries)" }, { ATA_CMD_WRITE_LONG, "WRITE LONG (with retries)" }, @@ -2377,8 +2557,9 @@ const char *ata_get_cmd_descript(u8 command) return cmd_descr[i].text; #endif - return NULL; + return "unknown"; } +EXPORT_SYMBOL_GPL(ata_get_cmd_name); /** * ata_eh_link_report - report error handling to user @@ -2393,8 +2574,9 @@ static void ata_eh_link_report(struct ata_link *link) { struct ata_port *ap = link->ap; struct ata_eh_context *ehc = &link->eh_context; + struct ata_queued_cmd *qc; const char *frozen, *desc; - char tries_buf[6]; + char tries_buf[16] = ""; int tag, nr_failed = 0; if (ehc->i.flags & ATA_EHI_QUIET) @@ -2404,10 +2586,8 @@ static void ata_eh_link_report(struct ata_link *link) if (ehc->i.desc[0] != '\0') desc = ehc->i.desc; - for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { - struct ata_queued_cmd *qc = __ata_qc_from_tag(ap, tag); - - if (!(qc->flags & ATA_QCFLAG_FAILED) || + ata_qc_for_each_raw(ap, qc, tag) { + if (!(qc->flags & ATA_QCFLAG_EH) || ata_dev_phys_link(qc->dev) != link || ((qc->flags & ATA_QCFLAG_QUIET) && qc->err_mask == AC_ERR_DEV)) @@ -2422,12 +2602,11 @@ static void ata_eh_link_report(struct ata_link *link) return; frozen = ""; - if (ap->pflags & ATA_PFLAG_FROZEN) + if (ata_port_is_frozen(ap)) frozen = " frozen"; - memset(tries_buf, 0, sizeof(tries_buf)); if (ap->eh_tries < ATA_EH_MAX_TRIES) - snprintf(tries_buf, sizeof(tries_buf) - 1, " t%d", + snprintf(tries_buf, sizeof(tries_buf), " t%d", ap->eh_tries); if (ehc->i.dev) { @@ -2469,14 +2648,12 @@ static void ata_eh_link_report(struct ata_link *link) ehc->i.serror & SERR_DEV_XCHG ? "DevExch " : ""); #endif - for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { - struct ata_queued_cmd *qc = __ata_qc_from_tag(ap, tag); + ata_qc_for_each_raw(ap, qc, tag) { struct ata_taskfile *cmd = &qc->tf, *res = &qc->result_tf; - const u8 *cdb = qc->cdb; char data_buf[20] = ""; char cdb_buf[70] = ""; - if (!(qc->flags & ATA_QCFLAG_FAILED) || + if (!(qc->flags & ATA_QCFLAG_EH) || ata_dev_phys_link(qc->dev) != link || !qc->err_mask) continue; @@ -2486,36 +2663,54 @@ static void ata_eh_link_report(struct ata_link *link) [DMA_TO_DEVICE] = "out", [DMA_FROM_DEVICE] = "in", }; - static const char *prot_str[] = { - [ATA_PROT_PIO] = "pio", - [ATA_PROT_DMA] = "dma", - [ATA_PROT_NCQ] = "ncq", - [ATAPI_PROT_PIO] = "pio", - [ATAPI_PROT_DMA] = "dma", - }; + const char *prot_str = NULL; + switch (qc->tf.protocol) { + case ATA_PROT_UNKNOWN: + prot_str = "unknown"; + break; + case ATA_PROT_NODATA: + prot_str = "nodata"; + break; + case ATA_PROT_PIO: + prot_str = "pio"; + break; + case ATA_PROT_DMA: + prot_str = "dma"; + break; + case ATA_PROT_NCQ: + prot_str = "ncq dma"; + break; + case ATA_PROT_NCQ_NODATA: + prot_str = "ncq nodata"; + break; + case ATAPI_PROT_NODATA: + prot_str = "nodata"; + break; + case ATAPI_PROT_PIO: + prot_str = "pio"; + break; + case ATAPI_PROT_DMA: + prot_str = "dma"; + break; + } snprintf(data_buf, sizeof(data_buf), " %s %u %s", - prot_str[qc->tf.protocol], qc->nbytes, - dma_str[qc->dma_dir]); + prot_str, qc->nbytes, dma_str[qc->dma_dir]); } if (ata_is_atapi(qc->tf.protocol)) { - if (qc->scsicmd) - scsi_print_command(qc->scsicmd); - else - snprintf(cdb_buf, sizeof(cdb_buf), - "cdb %02x %02x %02x %02x %02x %02x %02x %02x " - "%02x %02x %02x %02x %02x %02x %02x %02x\n ", - cdb[0], cdb[1], cdb[2], cdb[3], - cdb[4], cdb[5], cdb[6], cdb[7], - cdb[8], cdb[9], cdb[10], cdb[11], - cdb[12], cdb[13], cdb[14], cdb[15]); - } else { - const char *descr = ata_get_cmd_descript(cmd->command); - if (descr) - ata_dev_err(qc->dev, "failed command: %s\n", - descr); - } + const u8 *cdb = qc->cdb; + size_t cdb_len = qc->dev->cdb_len; + + if (qc->scsicmd) { + cdb = qc->scsicmd->cmnd; + cdb_len = qc->scsicmd->cmd_len; + } + __scsi_format_command(cdb_buf, sizeof(cdb_buf), + cdb, cdb_len); + } else + ata_dev_err(qc->dev, "failed command: %s\n", + ata_get_cmd_name(cmd->command)); ata_dev_err(qc->dev, "cmd %02x/%02x:%02x:%02x:%02x:%02x/%02x:%02x:%02x:%02x:%02x/%02x " @@ -2527,7 +2722,7 @@ static void ata_eh_link_report(struct ata_link *link) cmd->hob_feature, cmd->hob_nsect, cmd->hob_lbal, cmd->hob_lbam, cmd->hob_lbah, cmd->device, qc->tag, data_buf, cdb_buf, - res->command, res->feature, res->nsect, + res->status, res->error, res->nsect, res->lbal, res->lbam, res->lbah, res->hob_feature, res->hob_nsect, res->hob_lbal, res->hob_lbam, res->hob_lbah, @@ -2535,26 +2730,28 @@ static void ata_eh_link_report(struct ata_link *link) qc->err_mask & AC_ERR_NCQ ? " <F>" : ""); #ifdef CONFIG_ATA_VERBOSE_ERROR - if (res->command & (ATA_BUSY | ATA_DRDY | ATA_DF | ATA_DRQ | - ATA_ERR)) { - if (res->command & ATA_BUSY) + if (res->status & (ATA_BUSY | ATA_DRDY | ATA_DF | ATA_DRQ | + ATA_SENSE | ATA_ERR)) { + if (res->status & ATA_BUSY) ata_dev_err(qc->dev, "status: { Busy }\n"); else - ata_dev_err(qc->dev, "status: { %s%s%s%s}\n", - res->command & ATA_DRDY ? "DRDY " : "", - res->command & ATA_DF ? "DF " : "", - res->command & ATA_DRQ ? "DRQ " : "", - res->command & ATA_ERR ? "ERR " : ""); + ata_dev_err(qc->dev, "status: { %s%s%s%s%s}\n", + res->status & ATA_DRDY ? "DRDY " : "", + res->status & ATA_DF ? "DF " : "", + res->status & ATA_DRQ ? "DRQ " : "", + res->status & ATA_SENSE ? "SENSE " : "", + res->status & ATA_ERR ? "ERR " : ""); } if (cmd->command != ATA_CMD_PACKET && - (res->feature & (ATA_ICRC | ATA_UNC | ATA_IDNF | - ATA_ABORTED))) - ata_dev_err(qc->dev, "error: { %s%s%s%s}\n", - res->feature & ATA_ICRC ? "ICRC " : "", - res->feature & ATA_UNC ? "UNC " : "", - res->feature & ATA_IDNF ? "IDNF " : "", - res->feature & ATA_ABORTED ? "ABRT " : ""); + (res->error & (ATA_ICRC | ATA_UNC | ATA_AMNF | ATA_IDNF | + ATA_ABORTED))) + ata_dev_err(qc->dev, "error: { %s%s%s%s%s}\n", + res->error & ATA_ICRC ? "ICRC " : "", + res->error & ATA_UNC ? "UNC " : "", + res->error & ATA_AMNF ? "AMNF " : "", + res->error & ATA_IDNF ? "IDNF " : "", + res->error & ATA_ABORTED ? "ABRT " : ""); #endif } } @@ -2589,25 +2786,28 @@ static int ata_do_reset(struct ata_link *link, ata_reset_fn_t reset, return reset(link, classes, deadline); } -static int ata_eh_followup_srst_needed(struct ata_link *link, int rc) +static bool ata_eh_followup_srst_needed(struct ata_link *link, int rc) { if ((link->flags & ATA_LFLAG_NO_SRST) || ata_link_offline(link)) - return 0; + return false; if (rc == -EAGAIN) - return 1; + return true; if (sata_pmp_supported(link->ap) && ata_is_host_link(link)) - return 1; - return 0; + return true; + return false; } int ata_eh_reset(struct ata_link *link, int classify, - ata_prereset_fn_t prereset, ata_reset_fn_t softreset, - ata_reset_fn_t hardreset, ata_postreset_fn_t postreset) + struct ata_reset_operations *reset_ops) { struct ata_port *ap = link->ap; struct ata_link *slave = ap->slave_link; struct ata_eh_context *ehc = &link->eh_context; struct ata_eh_context *sehc = slave ? &slave->eh_context : NULL; + ata_reset_fn_t hardreset = reset_ops->hardreset; + ata_reset_fn_t softreset = reset_ops->softreset; + ata_prereset_fn_t prereset = reset_ops->prereset; + ata_postreset_fn_t postreset = reset_ops->postreset; unsigned int *classes = ehc->classes; unsigned int lflags = link->flags; int verbose = !(ehc->i.flags & ATA_EHI_QUIET); @@ -2623,7 +2823,7 @@ int ata_eh_reset(struct ata_link *link, int classify, /* * Prepare to reset */ - while (ata_eh_reset_timeouts[max_tries] != ULONG_MAX) + while (ata_eh_reset_timeouts[max_tries] != UINT_MAX) max_tries++; if (link->flags & ATA_LFLAG_RST_ONCE) max_tries = 1; @@ -2726,8 +2926,7 @@ int ata_eh_reset(struct ata_link *link, int classify, if (reset && !(ehc->i.action & ATA_EH_RESET)) { ata_for_each_dev(dev, link, ALL) classes[dev->devno] = ATA_DEV_NONE; - if ((ap->pflags & ATA_PFLAG_FROZEN) && - ata_is_host_link(link)) + if (ata_port_is_frozen(ap) && ata_is_host_link(link)) ata_eh_thaw_port(ap); rc = 0; goto out; @@ -2750,12 +2949,19 @@ int ata_eh_reset(struct ata_link *link, int classify, /* mark that this EH session started with reset */ ehc->last_reset = jiffies; - if (reset == hardreset) + if (reset == hardreset) { ehc->i.flags |= ATA_EHI_DID_HARDRESET; - else + trace_ata_link_hardreset_begin(link, classes, deadline); + } else { ehc->i.flags |= ATA_EHI_DID_SOFTRESET; + trace_ata_link_softreset_begin(link, classes, deadline); + } rc = ata_do_reset(link, reset, classes, deadline, true); + if (reset == hardreset) + trace_ata_link_hardreset_end(link, classes, rc); + else + trace_ata_link_softreset_end(link, classes, rc); if (rc && rc != -EAGAIN) { failed_link = link; goto fail; @@ -2769,11 +2975,15 @@ int ata_eh_reset(struct ata_link *link, int classify, ata_link_info(slave, "hard resetting link\n"); ata_eh_about_to_do(slave, NULL, ATA_EH_RESET); + trace_ata_slave_hardreset_begin(slave, classes, + deadline); tmp = ata_do_reset(slave, reset, classes, deadline, false); + trace_ata_slave_hardreset_end(slave, classes, tmp); switch (tmp) { case -EAGAIN: rc = -EAGAIN; + break; case 0: break; default: @@ -2797,7 +3007,9 @@ int ata_eh_reset(struct ata_link *link, int classify, } ata_eh_about_to_do(link, NULL, ATA_EH_RESET); + trace_ata_link_softreset_begin(link, classes, deadline); rc = ata_do_reset(link, reset, classes, deadline, true); + trace_ata_link_softreset_end(link, classes, rc); if (rc) { failed_link = link; goto fail; @@ -2851,27 +3063,20 @@ int ata_eh_reset(struct ata_link *link, int classify, */ if (postreset) { postreset(link, classes); - if (slave) + trace_ata_link_postreset(link, classes, rc); + if (slave) { postreset(slave, classes); + trace_ata_slave_postreset(slave, classes, rc); + } } - /* - * Some controllers can't be frozen very well and may set spurious - * error conditions during reset. Clear accumulated error - * information and re-thaw the port if frozen. As reset is the - * final recovery action and we cross check link onlineness against - * device classification later, no hotplug event is lost by this. - */ + /* clear cached SError */ spin_lock_irqsave(link->ap->lock, flags); - memset(&link->eh_info, 0, sizeof(link->eh_info)); + link->eh_info.serror = 0; if (slave) - memset(&slave->eh_info, 0, sizeof(link->eh_info)); - ap->pflags &= ~ATA_PFLAG_EH_PENDING; + slave->eh_info.serror = 0; spin_unlock_irqrestore(link->ap->lock, flags); - if (ap->pflags & ATA_PFLAG_FROZEN) - ata_eh_thaw_port(ap); - /* * Make sure onlineness and classification result correspond. * Hotplug could have happened during reset and some @@ -2951,6 +3156,8 @@ int ata_eh_reset(struct ata_link *link, int classify, */ if (ata_is_host_link(link)) ata_eh_thaw_port(ap); + ata_link_warn(link, "%s failed\n", + reset == hardreset ? "hardreset" : "softreset"); goto out; } @@ -3009,7 +3216,7 @@ static inline void ata_eh_pull_park_action(struct ata_port *ap) * ourselves at the beginning of each pass over the loop. * * Additionally, all write accesses to &ap->park_req_pending - * through INIT_COMPLETION() (see below) or complete_all() + * through reinit_completion() (see below) or complete_all() * (see ata_scsi_park_store()) are protected by the host lock. * As a result we have that park_req_pending.done is zero on * exit from this function, i.e. when ATA_EH_PARK actions for @@ -3023,7 +3230,7 @@ static inline void ata_eh_pull_park_action(struct ata_port *ap) */ spin_lock_irqsave(ap->lock, flags); - INIT_COMPLETION(ap->park_req_pending); + reinit_completion(&ap->park_req_pending); ata_for_each_link(link, ap, EDGE) { ata_for_each_dev(dev, link, ALL) { struct ata_eh_info *ehi = &link->eh_info; @@ -3056,7 +3263,7 @@ static void ata_eh_park_issue_cmd(struct ata_device *dev, int park) } tf.flags |= ATA_TFLAG_DEVICE | ATA_TFLAG_ISADDR; - tf.protocol |= ATA_PROT_NODATA; + tf.protocol = ATA_PROT_NODATA; err_mask = ata_exec_internal(dev, &tf, NULL, DMA_NONE, NULL, 0, 0); if (park && (err_mask || tf.lbal != 0xc4)) { ata_dev_err(dev, "head unload failed!\n"); @@ -3074,8 +3281,6 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, unsigned long flags; int rc = 0; - DPRINTK("ENTER\n"); - /* For PATA drive side cable detection to work, IDENTIFY must * be done backwards such that PDIAG- is released by the slave * device before the master device is identified. @@ -3090,7 +3295,24 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, if ((action & ATA_EH_REVALIDATE) && ata_dev_enabled(dev)) { WARN_ON(dev->class == ATA_DEV_PMP); - if (ata_phys_link_offline(ata_dev_phys_link(dev))) { + /* + * The link may be in a deep sleep, wake it up. + * + * If the link is in deep sleep, ata_phys_link_offline() + * will return true, causing the revalidation to fail, + * which leads to a (potentially) needless hard reset. + * + * ata_eh_recover() will later restore the link policy + * to ap->target_lpm_policy after revalidation is done. + */ + if (link->lpm_policy > ATA_LPM_MAX_POWER) { + rc = ata_eh_link_set_lpm(link, ATA_LPM_MAX_POWER, + r_failed_dev); + if (rc) + goto err; + } + + if (!ata_eh_link_established(ata_dev_phys_link(dev))) { rc = -EIO; goto err; } @@ -3109,7 +3331,7 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, ehc->i.flags |= ATA_EHI_SETMODE; /* schedule the scsi_rescan_device() here */ - schedule_work(&(ap->scsi_rescan_task)); + schedule_delayed_work(&ap->scsi_rescan_task, 0); } else if (dev->class == ATA_DEV_UNKNOWN && ehc->tries[dev->devno] && ata_class_enabled(ehc->classes[dev->devno])) { @@ -3188,18 +3410,18 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, return 0; err: + dev->flags &= ~ATA_DFLAG_RESUMING; *r_failed_dev = dev; - DPRINTK("EXIT rc=%d\n", rc); return rc; } /** - * ata_set_mode - Program timings and issue SET FEATURES - XFER + * ata_eh_set_mode - Program timings and issue SET FEATURES - XFER * @link: link on which timings will be programmed * @r_failed_dev: out parameter for failed device * * Set ATA device disk transfer mode (PIO3, UDMA6, etc.). If - * ata_set_mode() fails, pointer to the failing device is + * ata_eh_set_mode() fails, pointer to the failing device is * returned in @r_failed_dev. * * LOCKING: @@ -3208,7 +3430,8 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, * RETURNS: * 0 on success, negative errno otherwise */ -int ata_set_mode(struct ata_link *link, struct ata_device **r_failed_dev) +static int ata_eh_set_mode(struct ata_link *link, + struct ata_device **r_failed_dev) { struct ata_port *ap = link->ap; struct ata_device *dev; @@ -3229,7 +3452,7 @@ int ata_set_mode(struct ata_link *link, struct ata_device **r_failed_dev) if (ap->ops->set_mode) rc = ap->ops->set_mode(link, r_failed_dev); else - rc = ata_do_set_mode(link, r_failed_dev); + rc = ata_set_mode(link, r_failed_dev); /* if transfer mode has changed, set DUBIOUS_XFER on device */ ata_for_each_dev(dev, link, ENABLED) { @@ -3264,7 +3487,7 @@ static int atapi_eh_clear_ua(struct ata_device *dev) int i; for (i = 0; i < ATA_EH_UA_TRIES; i++) { - u8 *sense_buffer = dev->link->ap->sector_buf; + u8 *sense_buffer = dev->sector_buf; u8 sense_key = 0; unsigned int err_mask; @@ -3362,143 +3585,13 @@ static int ata_eh_maybe_retry_flush(struct ata_device *dev) if (err_mask & AC_ERR_DEV) { qc->err_mask |= AC_ERR_DEV; qc->result_tf = tf; - if (!(ap->pflags & ATA_PFLAG_FROZEN)) + if (!ata_port_is_frozen(ap)) rc = 0; } } return rc; } -/** - * ata_eh_set_lpm - configure SATA interface power management - * @link: link to configure power management - * @policy: the link power management policy - * @r_failed_dev: out parameter for failed device - * - * Enable SATA Interface power management. This will enable - * Device Interface Power Management (DIPM) for min_power - * policy, and then call driver specific callbacks for - * enabling Host Initiated Power management. - * - * LOCKING: - * EH context. - * - * RETURNS: - * 0 on success, -errno on failure. - */ -static int ata_eh_set_lpm(struct ata_link *link, enum ata_lpm_policy policy, - struct ata_device **r_failed_dev) -{ - struct ata_port *ap = ata_is_host_link(link) ? link->ap : NULL; - struct ata_eh_context *ehc = &link->eh_context; - struct ata_device *dev, *link_dev = NULL, *lpm_dev = NULL; - enum ata_lpm_policy old_policy = link->lpm_policy; - bool no_dipm = link->ap->flags & ATA_FLAG_NO_DIPM; - unsigned int hints = ATA_LPM_EMPTY | ATA_LPM_HIPM; - unsigned int err_mask; - int rc; - - /* if the link or host doesn't do LPM, noop */ - if ((link->flags & ATA_LFLAG_NO_LPM) || (ap && !ap->ops->set_lpm)) - return 0; - - /* - * DIPM is enabled only for MIN_POWER as some devices - * misbehave when the host NACKs transition to SLUMBER. Order - * device and link configurations such that the host always - * allows DIPM requests. - */ - ata_for_each_dev(dev, link, ENABLED) { - bool hipm = ata_id_has_hipm(dev->id); - bool dipm = ata_id_has_dipm(dev->id) && !no_dipm; - - /* find the first enabled and LPM enabled devices */ - if (!link_dev) - link_dev = dev; - - if (!lpm_dev && (hipm || dipm)) - lpm_dev = dev; - - hints &= ~ATA_LPM_EMPTY; - if (!hipm) - hints &= ~ATA_LPM_HIPM; - - /* disable DIPM before changing link config */ - if (policy != ATA_LPM_MIN_POWER && dipm) { - err_mask = ata_dev_set_feature(dev, - SETFEATURES_SATA_DISABLE, SATA_DIPM); - if (err_mask && err_mask != AC_ERR_DEV) { - ata_dev_warn(dev, - "failed to disable DIPM, Emask 0x%x\n", - err_mask); - rc = -EIO; - goto fail; - } - } - } - - if (ap) { - rc = ap->ops->set_lpm(link, policy, hints); - if (!rc && ap->slave_link) - rc = ap->ops->set_lpm(ap->slave_link, policy, hints); - } else - rc = sata_pmp_set_lpm(link, policy, hints); - - /* - * Attribute link config failure to the first (LPM) enabled - * device on the link. - */ - if (rc) { - if (rc == -EOPNOTSUPP) { - link->flags |= ATA_LFLAG_NO_LPM; - return 0; - } - dev = lpm_dev ? lpm_dev : link_dev; - goto fail; - } - - /* - * Low level driver acked the transition. Issue DIPM command - * with the new policy set. - */ - link->lpm_policy = policy; - if (ap && ap->slave_link) - ap->slave_link->lpm_policy = policy; - - /* host config updated, enable DIPM if transitioning to MIN_POWER */ - ata_for_each_dev(dev, link, ENABLED) { - if (policy == ATA_LPM_MIN_POWER && !no_dipm && - ata_id_has_dipm(dev->id)) { - err_mask = ata_dev_set_feature(dev, - SETFEATURES_SATA_ENABLE, SATA_DIPM); - if (err_mask && err_mask != AC_ERR_DEV) { - ata_dev_warn(dev, - "failed to enable DIPM, Emask 0x%x\n", - err_mask); - rc = -EIO; - goto fail; - } - } - } - - return 0; - -fail: - /* restore the old policy */ - link->lpm_policy = old_policy; - if (ap && ap->slave_link) - ap->slave_link->lpm_policy = old_policy; - - /* if no device or only one more chance is left, disable LPM */ - if (!dev || ehc->tries[dev->devno] <= 2) { - ata_link_warn(link, "disabling LPM on the link\n"); - link->flags |= ATA_LFLAG_NO_LPM; - } - if (r_failed_dev) - *r_failed_dev = dev; - return rc; -} - int ata_link_nr_enabled(struct ata_link *link) { struct ata_device *dev; @@ -3535,7 +3628,7 @@ static int ata_eh_skip_recovery(struct ata_link *link) return 1; /* thaw frozen port and recover failed devices */ - if ((ap->pflags & ATA_PFLAG_FROZEN) || ata_link_nr_enabled(link)) + if (ata_port_is_frozen(ap) || ata_link_nr_enabled(link)) return 0; /* reset at least once if reset is requested */ @@ -3630,9 +3723,11 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) case -ENODEV: /* device missing or wrong IDENTIFY data, schedule probing */ ehc->i.probe_mask |= (1 << dev->devno); + fallthrough; case -EINVAL: /* give it just one more chance */ ehc->tries[dev->devno] = min(ehc->tries[dev->devno], 1); + fallthrough; case -EIO: if (ehc->tries[dev->devno] == 1) { /* This is the last chance, better to slow @@ -3669,10 +3764,7 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) /** * ata_eh_recover - recover host port after error * @ap: host port to recover - * @prereset: prereset method (can be NULL) - * @softreset: softreset method (can be NULL) - * @hardreset: hardreset method (can be NULL) - * @postreset: postreset method (can be NULL) + * @reset_ops: The set of reset operations to use * @r_failed_link: out parameter for failed link * * This is the alpha and omega, eum and yang, heart and soul of @@ -3688,9 +3780,7 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) * RETURNS: * 0 on success, -errno on failure. */ -int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, - ata_reset_fn_t softreset, ata_reset_fn_t hardreset, - ata_postreset_fn_t postreset, +int ata_eh_recover(struct ata_port *ap, struct ata_reset_operations *reset_ops, struct ata_link **r_failed_link) { struct ata_link *link; @@ -3698,8 +3788,6 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, int rc, nr_fails; unsigned long flags, deadline; - DPRINTK("ENTER\n"); - /* prep for recovery */ ata_for_each_link(link, ap, EDGE) { struct ata_eh_context *ehc = &link->eh_context; @@ -3760,8 +3848,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, if (!(ehc->i.action & ATA_EH_RESET)) continue; - rc = ata_eh_reset(link, ata_link_nr_vacant(link), - prereset, softreset, hardreset, postreset); + rc = ata_eh_reset(link, ata_link_nr_vacant(link), reset_ops); if (rc) { ata_link_err(link, "reset failed, giving up\n"); goto out; @@ -3783,7 +3870,8 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, struct ata_eh_context *ehc = &link->eh_context; unsigned long tmp; - if (dev->class != ATA_DEV_ATA) + if (dev->class != ATA_DEV_ATA && + dev->class != ATA_DEV_ZAC) continue; if (!(ehc->i.dev_action[dev->devno] & ATA_EH_PARK)) @@ -3841,7 +3929,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, /* configure transfer mode if necessary */ if (ehc->i.flags & ATA_EHI_SETMODE) { - rc = ata_set_mode(link, &dev); + rc = ata_eh_set_mode(link, &dev); if (rc) goto rest_fail; ehc->i.flags &= ~ATA_EHI_SETMODE; @@ -3862,9 +3950,21 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, } } + /* + * Make sure to transition devices to the active power mode + * if needed (e.g. if we were scheduled on system resume). + */ + ata_for_each_dev(dev, link, ENABLED) { + if (ehc->i.dev_action[dev->devno] & ATA_EH_SET_ACTIVE) { + ata_dev_power_set_active(dev); + ata_eh_done(link, dev, ATA_EH_SET_ACTIVE); + } + } + /* retry flush if necessary */ ata_for_each_dev(dev, link, ALL) { - if (dev->class != ATA_DEV_ATA) + if (dev->class != ATA_DEV_ATA && + dev->class != ATA_DEV_ZAC) continue; rc = ata_eh_maybe_retry_flush(dev); if (rc) @@ -3874,7 +3974,8 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, config_lpm: /* configure link power saving */ if (link->lpm_policy != ap->target_lpm_policy) { - rc = ata_eh_set_lpm(link, ap->target_lpm_policy, &dev); + rc = ata_eh_link_set_lpm(link, ap->target_lpm_policy, + &dev); if (rc) goto rest_fail; } @@ -3888,7 +3989,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, if (dev) ata_eh_handle_dev_fail(dev, rc); - if (ap->pflags & ATA_PFLAG_FROZEN) { + if (ata_port_is_frozen(ap)) { /* PMP reset requires working host port. * Can't retry if it's frozen. */ @@ -3905,7 +4006,6 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, if (rc && r_failed_link) *r_failed_link = link; - DPRINTK("EXIT, rc=%d\n", rc); return rc; } @@ -3921,13 +4021,12 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, */ void ata_eh_finish(struct ata_port *ap) { + struct ata_queued_cmd *qc; int tag; /* retry or finish qcs */ - for (tag = 0; tag < ATA_MAX_QUEUE; tag++) { - struct ata_queued_cmd *qc = __ata_qc_from_tag(ap, tag); - - if (!(qc->flags & ATA_QCFLAG_FAILED)) + ata_qc_for_each_raw(ap, qc, tag) { + if (!(qc->flags & ATA_QCFLAG_EH)) continue; if (qc->err_mask) { @@ -3935,16 +4034,30 @@ void ata_eh_finish(struct ata_port *ap) * generate sense data in this function, * considering both err_mask and tf. */ - if (qc->flags & ATA_QCFLAG_RETRY) + if (qc->flags & ATA_QCFLAG_RETRY) { + /* + * Since qc->err_mask is set, ata_eh_qc_retry() + * will not increment scmd->allowed, so upper + * layer will only retry the command if it has + * not already been retried too many times. + */ ata_eh_qc_retry(qc); - else + } else { ata_eh_qc_complete(qc); + } } else { - if (qc->flags & ATA_QCFLAG_SENSE_VALID) { + if (qc->flags & ATA_QCFLAG_SENSE_VALID || + qc->flags & ATA_QCFLAG_EH_SUCCESS_CMD) { ata_eh_qc_complete(qc); } else { /* feed zero TF to sense generation */ memset(&qc->result_tf, 0, sizeof(qc->result_tf)); + /* + * Since qc->err_mask is not set, + * ata_eh_qc_retry() will increment + * scmd->allowed, so upper layer is guaranteed + * to retry the command. + */ ata_eh_qc_retry(qc); } } @@ -3956,59 +4069,40 @@ void ata_eh_finish(struct ata_port *ap) } /** - * ata_do_eh - do standard error handling + * ata_std_error_handler - standard error handler * @ap: host port to handle error for * - * @prereset: prereset method (can be NULL) - * @softreset: softreset method (can be NULL) - * @hardreset: hardreset method (can be NULL) - * @postreset: postreset method (can be NULL) - * * Perform standard error handling sequence. * * LOCKING: * Kernel thread context (may sleep). */ -void ata_do_eh(struct ata_port *ap, ata_prereset_fn_t prereset, - ata_reset_fn_t softreset, ata_reset_fn_t hardreset, - ata_postreset_fn_t postreset) +void ata_std_error_handler(struct ata_port *ap) { - struct ata_device *dev; + struct ata_reset_operations *reset_ops = &ap->ops->reset; + struct ata_link *link = &ap->link; int rc; + /* Ignore built-in hardresets if SCR access is not available */ + if ((reset_ops->hardreset == sata_std_hardreset || + reset_ops->hardreset == sata_sff_hardreset) && + !sata_scr_valid(link)) + link->flags |= ATA_LFLAG_NO_HRST; + ata_eh_autopsy(ap); ata_eh_report(ap); - rc = ata_eh_recover(ap, prereset, softreset, hardreset, postreset, - NULL); + rc = ata_eh_recover(ap, reset_ops, NULL); if (rc) { - ata_for_each_dev(dev, &ap->link, ALL) + struct ata_device *dev; + + ata_for_each_dev(dev, link, ALL) ata_dev_disable(dev); } ata_eh_finish(ap); } - -/** - * ata_std_error_handler - standard error handler - * @ap: host port to handle error for - * - * Standard error handler - * - * LOCKING: - * Kernel thread context (may sleep). - */ -void ata_std_error_handler(struct ata_port *ap) -{ - struct ata_port_operations *ops = ap->ops; - ata_reset_fn_t hardreset = ops->hardreset; - - /* ignore built-in hardreset if SCR access is not available */ - if (hardreset == sata_std_hardreset && !sata_scr_valid(&ap->link)) - hardreset = NULL; - - ata_do_eh(ap, ops->prereset, ops->softreset, hardreset, ops->postreset); -} +EXPORT_SYMBOL_GPL(ata_std_error_handler); #ifdef CONFIG_PM /** @@ -4025,6 +4119,7 @@ static void ata_eh_handle_port_suspend(struct ata_port *ap) unsigned long flags; int rc = 0; struct ata_device *dev; + struct ata_link *link; /* are we suspending? */ spin_lock_irqsave(ap->lock, flags); @@ -4038,6 +4133,22 @@ static void ata_eh_handle_port_suspend(struct ata_port *ap) WARN_ON(ap->pflags & ATA_PFLAG_SUSPENDED); /* + * We will reach this point for all of the PM events: + * PM_EVENT_SUSPEND (if runtime pm, PM_EVENT_AUTO will also be set) + * PM_EVENT_FREEZE, and PM_EVENT_HIBERNATE. + * + * We do not want to perform disk spin down for PM_EVENT_FREEZE. + * (Spin down will be performed by the subsequent PM_EVENT_HIBERNATE.) + */ + if (!(ap->pm_mesg.event & PM_EVENT_FREEZE)) { + /* Set all devices attached to the port in standby mode */ + ata_for_each_link(link, ap, HOST_FIRST) { + ata_for_each_dev(dev, link, ENABLED) + ata_dev_power_set_standby(dev); + } + } + + /* * If we have a ZPODD attached, check its zero * power ready status before the port is frozen. * Only needed for runtime suspend. @@ -4049,11 +4160,6 @@ static void ata_eh_handle_port_suspend(struct ata_port *ap) } } - /* tell ACPI we're suspending */ - rc = ata_acpi_on_suspend(ap); - if (rc) - goto out; - /* suspend */ ata_eh_freeze_port(ap); @@ -4061,21 +4167,16 @@ static void ata_eh_handle_port_suspend(struct ata_port *ap) rc = ap->ops->port_suspend(ap, ap->pm_mesg); ata_acpi_set_state(ap, ap->pm_mesg); - out: - /* report result */ + + /* update the flags */ spin_lock_irqsave(ap->lock, flags); ap->pflags &= ~ATA_PFLAG_PM_PENDING; if (rc == 0) ap->pflags |= ATA_PFLAG_SUSPENDED; - else if (ap->pflags & ATA_PFLAG_FROZEN) + else if (ata_port_is_frozen(ap)) ata_port_schedule_eh(ap); - if (ap->pm_result) { - *ap->pm_result = rc; - ap->pm_result = NULL; - } - spin_unlock_irqrestore(ap->lock, flags); return; @@ -4095,7 +4196,6 @@ static void ata_eh_handle_port_resume(struct ata_port *ap) struct ata_link *link; struct ata_device *dev; unsigned long flags; - int rc = 0; /* are we resuming? */ spin_lock_irqsave(ap->lock, flags); @@ -4122,18 +4222,15 @@ static void ata_eh_handle_port_resume(struct ata_port *ap) ata_acpi_set_state(ap, ap->pm_mesg); if (ap->ops->port_resume) - rc = ap->ops->port_resume(ap); + ap->ops->port_resume(ap); /* tell ACPI that we're resuming */ ata_acpi_on_resume(ap); - /* report result */ + /* update the flags */ spin_lock_irqsave(ap->lock, flags); ap->pflags &= ~(ATA_PFLAG_PM_PENDING | ATA_PFLAG_SUSPENDED); - if (ap->pm_result) { - *ap->pm_result = rc; - ap->pm_result = NULL; - } + ap->pflags |= ATA_PFLAG_RESUMING; spin_unlock_irqrestore(ap->lock, flags); } #endif /* CONFIG_PM */ |