X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=drivers%2Fata%2Flibata-eh.c;h=f77a67303f8b3d0d724e4fccd3510a269a72b0c8;hb=f47451c45fe0032ef491aaf3e0623fa0154e156d;hp=33ac5ea4f531fb26d648d4bdab36472666ec4e9c;hpb=45fabbb77bd95adff7a80bde1c7a0ace1075fde6;p=safe%2Fjmp%2Flinux-2.6 diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index 33ac5ea..f77a673 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -33,12 +33,14 @@ */ #include +#include #include #include #include #include #include #include +#include #include "../scsi/scsi_transport_api.h" #include @@ -79,6 +81,12 @@ enum { */ ATA_EH_PRERESET_TIMEOUT = 10000, ATA_EH_FASTDRAIN_INTERVAL = 3000, + + ATA_EH_UA_TRIES = 5, + + /* probe speed down parameters, see ata_eh_schedule_probe() */ + ATA_EH_PROBE_TRIAL_INTERVAL = 60000, /* 1 min */ + ATA_EH_PROBE_TRIALS = 2, }; /* The following table determines how we sequence resets. Each entry @@ -102,6 +110,13 @@ static const unsigned long ata_eh_identify_timeouts[] = { ULONG_MAX, }; +static const unsigned long ata_eh_flush_timeouts[] = { + 15000, /* be generous with flush */ + 15000, /* ditto */ + 30000, /* and even more generous */ + ULONG_MAX, +}; + static const unsigned long ata_eh_other_timeouts[] = { 5000, /* same rationale as identify timeout */ 10000, /* ditto */ @@ -139,6 +154,8 @@ ata_eh_cmd_timeout_table[ATA_EH_CMD_TIMEOUT_TABLE_SIZE] = { .timeouts = ata_eh_other_timeouts, }, { .commands = CMDS(ATA_CMD_INIT_DEV_PARAMS), .timeouts = ata_eh_other_timeouts, }, + { .commands = CMDS(ATA_CMD_FLUSH, ATA_CMD_FLUSH_EXT), + .timeouts = ata_eh_flush_timeouts }, }; #undef CMDS @@ -419,7 +436,7 @@ static void ata_eh_clear_action(struct ata_link *link, struct ata_device *dev, if (!dev) { ehi->action &= ~action; - ata_link_for_each_dev(tdev, link) + ata_for_each_dev(tdev, link, ALL) ehi->dev_action[tdev->devno] &= ~action; } else { /* doesn't make sense for port-wide EH actions */ @@ -427,7 +444,7 @@ static void ata_eh_clear_action(struct ata_link *link, struct ata_device *dev, /* break ehi->action into ehi->dev_action */ if (ehi->action & action) { - ata_link_for_each_dev(tdev, link) + ata_for_each_dev(tdev, link, ALL) ehi->dev_action[tdev->devno] |= ehi->action & action; ehi->action &= ~action; @@ -457,29 +474,29 @@ static void ata_eh_clear_action(struct ata_link *link, struct ata_device *dev, * RETURNS: * EH_HANDLED or EH_NOT_HANDLED */ -enum scsi_eh_timer_return ata_scsi_timed_out(struct scsi_cmnd *cmd) +enum blk_eh_timer_return ata_scsi_timed_out(struct scsi_cmnd *cmd) { 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 scsi_eh_timer_return ret; + enum blk_eh_timer_return ret; DPRINTK("ENTER\n"); if (ap->ops->error_handler) { - ret = EH_NOT_HANDLED; + ret = BLK_EH_NOT_HANDLED; goto out; } - ret = EH_HANDLED; + 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 = EH_NOT_HANDLED; + ret = BLK_EH_NOT_HANDLED; } spin_unlock_irqrestore(ap->lock, flags); @@ -488,6 +505,31 @@ enum scsi_eh_timer_return ata_scsi_timed_out(struct scsi_cmnd *cmd) return ret; } +static void ata_eh_unload(struct ata_port *ap) +{ + struct ata_link *link; + struct ata_device *dev; + unsigned long flags; + + /* 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); + } + + /* freeze and set UNLOADED */ + spin_lock_irqsave(ap->lock, flags); + + ata_port_freeze(ap); /* won't be thawed */ + ap->pflags &= ~ATA_PFLAG_EH_PENDING; /* clear pending from freeze */ + ap->pflags |= ATA_PFLAG_UNLOADED; + + spin_unlock_irqrestore(ap->lock, flags); +} + /** * ata_scsi_error - SCSI layer error handler callback * @host: SCSI host on which error occurred @@ -508,14 +550,14 @@ void ata_scsi_error(struct Scsi_Host *host) DPRINTK("ENTER\n"); - /* synchronize with port task */ - ata_port_flush_task(ap); + /* 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 - * normal completion, error completion, and SCSI timeout. - * Both cmpletions can race against SCSI timeout. When normal + * 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. * @@ -530,7 +572,19 @@ void ata_scsi_error(struct Scsi_Host *host) 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. + + 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); + list_for_each_entry_safe(scmd, tmp, &host->eh_cmd_q, eh_entry) { struct ata_queued_cmd *qc; @@ -574,6 +628,9 @@ void ata_scsi_error(struct Scsi_Host *host) ap->eh_tries = ATA_EH_MAX_TRIES; } else spin_unlock_wait(ap->lock); + + /* If we timed raced normal completion and there is nothing to + recover nr_timedout == 0 why exactly are we doing error recovery ? */ repeat: /* invoke error handler */ @@ -589,7 +646,7 @@ void ata_scsi_error(struct Scsi_Host *host) /* fetch & clear EH info */ spin_lock_irqsave(ap->lock, flags); - __ata_port_for_each_link(link, ap) { + ata_for_each_link(link, ap, HOST_FIRST) { struct ata_eh_context *ehc = &link->eh_context; struct ata_device *dev; @@ -597,16 +654,13 @@ void ata_scsi_error(struct Scsi_Host *host) link->eh_context.i = link->eh_info; memset(&link->eh_info, 0, sizeof(link->eh_info)); - ata_link_for_each_dev(dev, link) { + 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; } - - /* set last reset timestamp to some time in the past */ - ehc->last_reset = jiffies - 60 * HZ; } ap->pflags |= ATA_PFLAG_EH_IN_PROGRESS; @@ -618,8 +672,13 @@ void ata_scsi_error(struct Scsi_Host *host) /* invoke EH, skip if unloading or suspended */ if (!(ap->pflags & (ATA_PFLAG_UNLOADING | ATA_PFLAG_SUSPENDED))) ap->ops->error_handler(ap); - else + 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); @@ -641,7 +700,7 @@ void ata_scsi_error(struct Scsi_Host *host) } /* this run is complete, make sure EH info is clear */ - __ata_port_for_each_link(link, ap) + ata_for_each_link(link, ap, HOST_FIRST) memset(&link->eh_info, 0, sizeof(link->eh_info)); /* Clear host_eh_scheduled while holding ap->lock such @@ -820,6 +879,8 @@ 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); @@ -831,7 +892,9 @@ void ata_qc_schedule_eh(struct ata_queued_cmd *qc) * Note that ATA_QCFLAG_FAILED is unconditionally set after * this function completes. */ - scsi_req_abort_cmd(qc->scsicmd); + spin_lock_irqsave(q->queue_lock, flags); + blk_abort_request(qc->scsicmd->request); + spin_unlock_irqrestore(q->queue_lock, flags); } /** @@ -950,7 +1013,9 @@ static void __ata_port_freeze(struct ata_port *ap) * ata_port_freeze - abort & freeze port * @ap: ATA port to freeze * - * Abort and freeze @ap. + * Abort and freeze @ap. The freeze operation must be called + * first, because some hardware requires special operations + * before the taskfile registers are accessible. * * LOCKING: * spin_lock_irqsave(host lock) @@ -964,8 +1029,8 @@ int ata_port_freeze(struct ata_port *ap) WARN_ON(!ap->ops->error_handler); - nr_aborted = ata_port_abort(ap); __ata_port_freeze(ap); + nr_aborted = ata_port_abort(ap); return nr_aborted; } @@ -1022,7 +1087,7 @@ int sata_async_notification(struct ata_port *ap) struct ata_link *link; /* check and notify ATAPI AN */ - ata_port_for_each_link(link, ap) { + ata_for_each_link(link, ap, EDGE) { if (!(sntf & (1 << link->pmp))) continue; @@ -1146,6 +1211,32 @@ void ata_eh_qc_retry(struct ata_queued_cmd *qc) } /** + * ata_dev_disable - disable ATA device + * @dev: ATA device to disable + * + * Disable @dev. + * + * Locking: + * EH context. + */ +void ata_dev_disable(struct ata_device *dev) +{ + if (!ata_dev_enabled(dev)) + return; + + if (ata_msg_drv(dev->link->ap)) + ata_dev_printk(dev, KERN_WARNING, "disabled\n"); + ata_acpi_on_disable(dev); + ata_down_xfermask_limit(dev, ATA_DNXFER_FORCE_PIO0 | ATA_DNXFER_QUIET); + dev->class++; + + /* 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_detach_dev - detach ATA device * @dev: ATA device to detach * @@ -1158,6 +1249,7 @@ void ata_eh_detach_dev(struct ata_device *dev) { struct ata_link *link = dev->link; struct ata_port *ap = link->ap; + struct ata_eh_context *ehc = &link->eh_context; unsigned long flags; ata_dev_disable(dev); @@ -1171,9 +1263,11 @@ void ata_eh_detach_dev(struct ata_device *dev) ap->pflags |= ATA_PFLAG_SCSI_HOTPLUG; } - /* clear per-dev EH actions */ + /* clear per-dev EH info */ ata_eh_clear_action(link, dev, &link->eh_info, ATA_EH_PERDEV_MASK); ata_eh_clear_action(link, dev, &link->eh_context.i, ATA_EH_PERDEV_MASK); + ehc->saved_xfer_mode[dev->devno] = 0; + ehc->saved_ncq_enabled &= ~(1 << dev->devno); spin_unlock_irqrestore(ap->lock, flags); } @@ -1203,7 +1297,10 @@ void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev, ata_eh_clear_action(link, dev, ehi, action); - if (!(ehc->i.flags & ATA_EHI_QUIET)) + /* About to take EH action, set RECOVERED. Ignore actions on + * slave links as master will do them again. + */ + if (!(ehc->i.flags & ATA_EHI_QUIET) && link != ap->slave_link) ap->pflags |= ATA_PFLAG_RECOVERED; spin_unlock_irqrestore(ap->lock, flags); @@ -1357,6 +1454,37 @@ static int ata_eh_read_log_10h(struct ata_device *dev, } /** + * atapi_eh_tur - perform ATAPI TEST_UNIT_READY + * @dev: target ATAPI device + * @r_sense_key: out parameter for sense_key + * + * Perform ATAPI TEST_UNIT_READY. + * + * LOCKING: + * EH context (may sleep). + * + * RETURNS: + * 0 on success, AC_ERR_* mask on failure. + */ +static 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; + + ata_tf_init(dev, &tf); + + 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.feature >> 4; + return err_mask; +} + +/** * atapi_eh_request_sense - perform ATAPI REQUEST_SENSE * @dev: device to perform REQUEST_SENSE to * @sense_buf: result sense data buffer (SCSI_SENSE_BUFFERSIZE bytes long) @@ -1500,6 +1628,7 @@ void ata_eh_analyze_ncq_error(struct ata_link *link) } /* okay, this error is ours */ + memset(&tf, 0, sizeof(tf)); rc = ata_eh_read_log_10h(dev, &tag, &tf); if (rc) { ata_link_printk(link, KERN_ERR, "failed to read log page 10h " @@ -1782,7 +1911,7 @@ static unsigned int ata_eh_speed_down(struct ata_device *dev, /* speed down? */ if (verdict & ATA_EH_SPDN_SPEED_DOWN) { /* speed down SATA link speed if possible */ - if (sata_down_spd_limit(link) == 0) { + if (sata_down_spd_limit(link, 0) == 0) { action |= ATA_EH_RESET; goto done; } @@ -1904,8 +2033,9 @@ static void ata_eh_link_autopsy(struct ata_link *link) qc->err_mask &= ~(AC_ERR_DEV | AC_ERR_OTHER); /* determine whether the command is worth retrying */ - if (!(qc->err_mask & AC_ERR_INVALID) && - ((qc->flags & ATA_QCFLAG_IO) || qc->err_mask != AC_ERR_DEV)) + if (qc->flags & ATA_QCFLAG_IO || + (!(qc->err_mask & AC_ERR_INVALID) && + qc->err_mask != AC_ERR_DEV)) qc->flags |= ATA_QCFLAG_RETRY; /* accumulate error info */ @@ -1965,7 +2095,7 @@ void ata_eh_autopsy(struct ata_port *ap) { struct ata_link *link; - ata_port_for_each_link(link, ap) + ata_for_each_link(link, ap, EDGE) ata_eh_link_autopsy(link); /* Handle the frigging slave link. Autopsy is done similarly @@ -1976,8 +2106,13 @@ void ata_eh_autopsy(struct ata_port *ap) struct ata_eh_context *mehc = &ap->link.eh_context; struct ata_eh_context *sehc = &ap->slave_link->eh_context; + /* transfer control flags from master to slave */ + sehc->i.flags |= mehc->i.flags & ATA_EHI_TO_SLAVE_MASK; + + /* perform autopsy on the slave link */ ata_eh_link_autopsy(ap->slave_link); + /* transfer actions from slave to master and clear slave */ ata_eh_about_to_do(ap->slave_link, NULL, ATA_EH_ALL_ACTIONS); mehc->i.action |= sehc->i.action; mehc->i.dev_action[1] |= sehc->i.dev_action[1]; @@ -1993,6 +2128,116 @@ 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 + * + * Return a textual description of the given command, or NULL if the + * command is not known. + * + * LOCKING: + * None + */ +const char *ata_get_cmd_descript(u8 command) +{ +#ifdef CONFIG_ATA_VERBOSE_ERROR + static const struct + { + 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_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_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_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_PIO_READ, "READ SECTOR(S)" }, + { ATA_CMD_PIO_READ_EXT, "READ SECTOR(S) EXT" }, + { ATA_CMD_PIO_WRITE, "WRITE SECTOR(S)" }, + { ATA_CMD_PIO_WRITE_EXT, "WRITE SECTOR(S) EXT" }, + { ATA_CMD_READ_MULTI, "READ MULTIPLE" }, + { 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_SET_FEATURES, "SET FEATURES" }, + { ATA_CMD_SET_MULTI, "SET MULTIPLE MODE" }, + { ATA_CMD_VERIFY, "READ VERIFY SECTOR(S)" }, + { ATA_CMD_VERIFY_EXT, "READ VERIFY SECTOR(S) EXT" }, + { ATA_CMD_WRITE_UNCORR_EXT, "WRITE UNCORRECTABLE EXT" }, + { ATA_CMD_STANDBYNOW1, "STANDBY IMMEDIATE" }, + { ATA_CMD_IDLEIMMEDIATE, "IDLE IMMEDIATE" }, + { ATA_CMD_SLEEP, "SLEEP" }, + { ATA_CMD_INIT_DEV_PARAMS, "INITIALIZE DEVICE PARAMETERS" }, + { ATA_CMD_READ_NATIVE_MAX, "READ NATIVE MAX ADDRESS" }, + { ATA_CMD_READ_NATIVE_MAX_EXT, "READ NATIVE MAX ADDRESS EXT" }, + { ATA_CMD_SET_MAX, "SET MAX ADDRESS" }, + { ATA_CMD_SET_MAX_EXT, "SET MAX ADDRESS EXT" }, + { 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_TRUSTED_RCV, "TRUSTED RECEIVE" }, + { ATA_CMD_TRUSTED_RCV_DMA, "TRUSTED RECEIVE DMA" }, + { ATA_CMD_TRUSTED_SND, "TRUSTED SEND" }, + { ATA_CMD_TRUSTED_SND_DMA, "TRUSTED SEND DMA" }, + { ATA_CMD_PMP_READ, "READ BUFFER" }, + { ATA_CMD_PMP_WRITE, "WRITE BUFFER" }, + { ATA_CMD_CONF_OVERLAY, "DEVICE CONFIGURATION OVERLAY" }, + { ATA_CMD_SEC_SET_PASS, "SECURITY SET PASSWORD" }, + { ATA_CMD_SEC_UNLOCK, "SECURITY UNLOCK" }, + { ATA_CMD_SEC_ERASE_PREP, "SECURITY ERASE PREPARE" }, + { ATA_CMD_SEC_ERASE_UNIT, "SECURITY ERASE UNIT" }, + { ATA_CMD_SEC_FREEZE_LOCK, "SECURITY FREEZE LOCK" }, + { ATA_CMD_SEC_DISABLE_PASS, "SECURITY DISABLE PASSWORD" }, + { ATA_CMD_CONFIG_STREAM, "CONFIGURE STREAM" }, + { ATA_CMD_SMART, "SMART" }, + { ATA_CMD_MEDIA_LOCK, "DOOR LOCK" }, + { ATA_CMD_MEDIA_UNLOCK, "DOOR UNLOCK" }, + { 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_READ_LONG, "READ LONG (with retries)" }, + { ATA_CMD_READ_LONG_ONCE, "READ LONG (without retries)" }, + { ATA_CMD_WRITE_LONG, "WRITE LONG (with retries)" }, + { ATA_CMD_WRITE_LONG_ONCE, "WRITE LONG (without retries)" }, + { ATA_CMD_RESTORE, "RECALIBRATE" }, + { 0, NULL } /* terminate list */ + }; + + unsigned int i; + for (i = 0; cmd_descr[i].text; i++) + if (cmd_descr[i].command == command) + return cmd_descr[i].text; +#endif + + return NULL; +} + +/** * ata_eh_link_report - report error handling to user * @link: ATA link EH is going on * @@ -2058,6 +2303,7 @@ static void ata_eh_link_report(struct ata_link *link) ata_link_printk(link, KERN_ERR, "%s\n", desc); } +#ifdef CONFIG_ATA_VERBOSE_ERROR if (ehc->i.serror) ata_link_printk(link, KERN_ERR, "SError: { %s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s}\n", @@ -2078,6 +2324,7 @@ static void ata_eh_link_report(struct ata_link *link) ehc->i.serror & SERR_TRANS_ST_ERROR ? "TrStaTrns " : "", ehc->i.serror & SERR_UNRECOG_FIS ? "UnrecFIS " : "", 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); @@ -2109,14 +2356,23 @@ static void ata_eh_link_report(struct ata_link *link) dma_str[qc->dma_dir]); } - if (ata_is_atapi(qc->tf.protocol)) - snprintf(cdb_buf, sizeof(cdb_buf), + 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_printk(qc->dev, KERN_ERR, + "failed command: %s\n", descr); + } ata_dev_printk(qc->dev, KERN_ERR, "cmd %02x/%02x:%02x:%02x:%02x:%02x/%02x:%02x:%02x:%02x:%02x/%02x " @@ -2135,6 +2391,7 @@ static void ata_eh_link_report(struct ata_link *link) res->device, qc->err_mask, ata_err_string(qc->err_mask), qc->err_mask & AC_ERR_NCQ ? " " : ""); +#ifdef CONFIG_ATA_VERBOSE_ERROR if (res->command & (ATA_BUSY | ATA_DRDY | ATA_DF | ATA_DRQ | ATA_ERR)) { if (res->command & ATA_BUSY) @@ -2158,6 +2415,7 @@ static void ata_eh_link_report(struct ata_link *link) res->feature & ATA_UNC ? "UNC " : "", res->feature & ATA_IDNF ? "IDNF " : "", res->feature & ATA_ABORTED ? "ABRT " : ""); +#endif } } @@ -2174,7 +2432,7 @@ void ata_eh_report(struct ata_port *ap) { struct ata_link *link; - __ata_port_for_each_link(link, ap) + ata_for_each_link(link, ap, HOST_FIRST) ata_eh_link_report(link); } @@ -2185,7 +2443,7 @@ static int ata_do_reset(struct ata_link *link, ata_reset_fn_t reset, struct ata_device *dev; if (clear_classes) - ata_link_for_each_dev(dev, link) + ata_for_each_dev(dev, link, ALL) classes[dev->devno] = ATA_DEV_UNKNOWN; return reset(link, classes, deadline); @@ -2210,7 +2468,7 @@ int ata_eh_reset(struct ata_link *link, int classify, 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->eh_context; + struct ata_eh_context *sehc = slave ? &slave->eh_context : NULL; unsigned int *classes = ehc->classes; unsigned int lflags = link->flags; int verbose = !(ehc->i.flags & ATA_EHI_QUIET); @@ -2233,19 +2491,23 @@ int ata_eh_reset(struct ata_link *link, int classify, if (link->flags & ATA_LFLAG_NO_SRST) softreset = NULL; - now = jiffies; - deadline = ata_deadline(ehc->last_reset, ATA_EH_RESET_COOL_DOWN); - if (time_before(now, deadline)) - schedule_timeout_uninterruptible(deadline - now); + /* make sure each reset attemp is at least COOL_DOWN apart */ + if (ehc->i.flags & ATA_EHI_DID_RESET) { + now = jiffies; + WARN_ON(time_after(ehc->last_reset, now)); + deadline = ata_deadline(ehc->last_reset, + ATA_EH_RESET_COOL_DOWN); + if (time_before(now, deadline)) + schedule_timeout_uninterruptible(deadline - now); + } spin_lock_irqsave(ap->lock, flags); ap->pflags |= ATA_PFLAG_RESETTING; spin_unlock_irqrestore(ap->lock, flags); ata_eh_about_to_do(link, NULL, ATA_EH_RESET); - ehc->last_reset = jiffies; - ata_link_for_each_dev(dev, link) { + ata_for_each_dev(dev, link, ALL) { /* If we issue an SRST then an ATA drive (not ATAPI) * may change configuration and be in PIO0 timing. If * we do a hard reset (or are coming from power on) @@ -2306,7 +2568,7 @@ int ata_eh_reset(struct ata_link *link, int classify, "port disabled. ignoring.\n"); ehc->i.action &= ~ATA_EH_RESET; - ata_link_for_each_dev(dev, link) + ata_for_each_dev(dev, link, ALL) classes[dev->devno] = ATA_DEV_NONE; rc = 0; @@ -2317,11 +2579,14 @@ int ata_eh_reset(struct ata_link *link, int classify, } /* prereset() might have cleared ATA_EH_RESET. If so, - * bang classes and return. + * bang classes, thaw and return. */ if (reset && !(ehc->i.action & ATA_EH_RESET)) { - ata_link_for_each_dev(dev, link) + ata_for_each_dev(dev, link, ALL) classes[dev->devno] = ATA_DEV_NONE; + if ((ap->pflags & ATA_PFLAG_FROZEN) && + ata_is_host_link(link)) + ata_eh_thaw_port(ap); rc = 0; goto out; } @@ -2331,7 +2596,6 @@ int ata_eh_reset(struct ata_link *link, int classify, /* * Perform reset */ - ehc->last_reset = jiffies; if (ata_is_host_link(link)) ata_eh_freeze_port(ap); @@ -2343,6 +2607,7 @@ int ata_eh_reset(struct ata_link *link, int classify, reset == softreset ? "soft" : "hard"); /* mark that this EH session started with reset */ + ehc->last_reset = jiffies; if (reset == hardreset) ehc->i.flags |= ATA_EHI_DID_HARDRESET; else @@ -2393,6 +2658,10 @@ int ata_eh_reset(struct ata_link *link, int classify, ata_eh_about_to_do(link, NULL, ATA_EH_RESET); rc = ata_do_reset(link, reset, classes, deadline, true); + if (rc) { + failed_link = link; + goto fail; + } } } else { if (verbose) @@ -2405,7 +2674,7 @@ int ata_eh_reset(struct ata_link *link, int classify, /* * Post-reset processing */ - ata_link_for_each_dev(dev, link) { + ata_for_each_dev(dev, link, ALL) { /* After the reset, the device state is PIO 0 and the * controller state is undefined. Reset also wakes up * drives from sleeping mode. @@ -2420,7 +2689,7 @@ int ata_eh_reset(struct ata_link *link, int classify, if (lflags & ATA_LFLAG_ASSUME_ATA) classes[dev->devno] = ATA_DEV_ATA; else if (lflags & ATA_LFLAG_ASSUME_SEMB) - classes[dev->devno] = ATA_DEV_SEMB_UNSUP; /* not yet */ + classes[dev->devno] = ATA_DEV_SEMB_UNSUP; } /* record current link speed */ @@ -2446,48 +2715,68 @@ int ata_eh_reset(struct ata_link *link, int classify, postreset(slave, classes); } - /* clear cached SError */ + /* + * Some controllers can't be frozen very well and may set + * spuruious error conditions during reset. Clear accumulated + * error information. As reset is the final recovery action, + * nothing is lost by doing this. + */ spin_lock_irqsave(link->ap->lock, flags); - link->eh_info.serror = 0; + memset(&link->eh_info, 0, sizeof(link->eh_info)); if (slave) - slave->eh_info.serror = 0; + memset(&slave->eh_info, 0, sizeof(link->eh_info)); + ap->pflags &= ~ATA_PFLAG_EH_PENDING; spin_unlock_irqrestore(link->ap->lock, flags); - /* Make sure onlineness and classification result correspond. + /* + * Make sure onlineness and classification result correspond. * Hotplug could have happened during reset and some * controllers fail to wait while a drive is spinning up after * being hotplugged causing misdetection. By cross checking - * link onlineness and classification result, those conditions - * can be reliably detected and retried. + * link on/offlineness and classification result, those + * conditions can be reliably detected and retried. */ nr_unknown = 0; - ata_link_for_each_dev(dev, link) { - /* convert all ATA_DEV_UNKNOWN to ATA_DEV_NONE */ - if (classes[dev->devno] == ATA_DEV_UNKNOWN) { - classes[dev->devno] = ATA_DEV_NONE; - if (ata_phys_link_online(ata_dev_phys_link(dev))) + ata_for_each_dev(dev, link, ALL) { + if (ata_phys_link_online(ata_dev_phys_link(dev))) { + if (classes[dev->devno] == ATA_DEV_UNKNOWN) { + ata_dev_printk(dev, KERN_DEBUG, "link online " + "but device misclassifed\n"); + classes[dev->devno] = ATA_DEV_NONE; nr_unknown++; + } + } else if (ata_phys_link_offline(ata_dev_phys_link(dev))) { + if (ata_class_enabled(classes[dev->devno])) + ata_dev_printk(dev, KERN_DEBUG, "link offline, " + "clearing class %d to NONE\n", + classes[dev->devno]); + classes[dev->devno] = ATA_DEV_NONE; + } else if (classes[dev->devno] == ATA_DEV_UNKNOWN) { + ata_dev_printk(dev, KERN_DEBUG, "link status unknown, " + "clearing UNKNOWN to NONE\n"); + classes[dev->devno] = ATA_DEV_NONE; } } if (classify && nr_unknown) { if (try < max_tries) { ata_link_printk(link, KERN_WARNING, "link online but " - "device misclassified, retrying\n"); + "%d devices misclassified, retrying\n", + nr_unknown); failed_link = link; rc = -EAGAIN; goto fail; } ata_link_printk(link, KERN_WARNING, - "link online but device misclassified, " - "device detection might fail\n"); + "link online but %d devices misclassified, " + "device detection might fail\n", nr_unknown); } /* reset successful, schedule revalidation */ ata_eh_done(link, NULL, ATA_EH_RESET); if (slave) ata_eh_done(slave, NULL, ATA_EH_RESET); - ehc->last_reset = jiffies; + ehc->last_reset = jiffies; /* update to completion time */ ehc->i.action |= ATA_EH_REVALIDATE; rc = 0; @@ -2525,11 +2814,11 @@ int ata_eh_reset(struct ata_link *link, int classify, } if (try == max_tries - 1) { - sata_down_spd_limit(link); + sata_down_spd_limit(link, 0); if (slave) - sata_down_spd_limit(slave); + sata_down_spd_limit(slave, 0); } else if (rc == -EPIPE) - sata_down_spd_limit(failed_link); + sata_down_spd_limit(failed_link, 0); if (hardreset) reset = hardreset; @@ -2570,8 +2859,8 @@ static inline void ata_eh_pull_park_action(struct ata_port *ap) spin_lock_irqsave(ap->lock, flags); INIT_COMPLETION(ap->park_req_pending); - ata_port_for_each_link(link, ap) { - ata_link_for_each_dev(dev, link) { + ata_for_each_link(link, ap, EDGE) { + ata_for_each_dev(dev, link, ALL) { struct ata_eh_info *ehi = &link->eh_info; link->eh_context.i.dev_action[dev->devno] |= @@ -2626,7 +2915,7 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, * be done backwards such that PDIAG- is released by the slave * device before the master device is identified. */ - ata_link_for_each_dev_reverse(dev, link) { + ata_for_each_dev(dev, link, ALL_REVERSE) { unsigned int action = ata_eh_dev_action(dev); unsigned int readid_flags = 0; @@ -2659,6 +2948,12 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, } else if (dev->class == ATA_DEV_UNKNOWN && ehc->tries[dev->devno] && ata_class_enabled(ehc->classes[dev->devno])) { + /* Temporarily set dev->class, it will be + * permanently set once all configurations are + * complete. This is necessary because new + * device configuration is done in two + * separate loops. + */ dev->class = ehc->classes[dev->devno]; if (dev->class == ATA_DEV_PMP) @@ -2666,20 +2961,25 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, else rc = ata_dev_read_id(dev, &dev->class, readid_flags, dev->id); + + /* read_id might have changed class, store and reset */ + ehc->classes[dev->devno] = dev->class; + dev->class = ATA_DEV_UNKNOWN; + switch (rc) { case 0: + /* clear error info accumulated during probe */ + ata_ering_clear(&dev->ering); new_mask |= 1 << dev->devno; break; case -ENOENT: /* IDENTIFY was issued to non-existent * device. No need to reset. Just - * thaw and kill the device. + * thaw and ignore the device. */ ata_eh_thaw_port(ap); - dev->class = ATA_DEV_UNKNOWN; break; default: - dev->class = ATA_DEV_UNKNOWN; goto err; } } @@ -2695,16 +2995,22 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, /* Configure new devices forward such that user doesn't see * device detection messages backwards. */ - ata_link_for_each_dev(dev, link) { - if (!(new_mask & (1 << dev->devno)) || - dev->class == ATA_DEV_PMP) + ata_for_each_dev(dev, link, ALL) { + if (!(new_mask & (1 << dev->devno))) + continue; + + dev->class = ehc->classes[dev->devno]; + + if (dev->class == ATA_DEV_PMP) continue; ehc->i.flags |= ATA_EHI_PRINTINFO; rc = ata_dev_configure(dev); ehc->i.flags &= ~ATA_EHI_PRINTINFO; - if (rc) + if (rc) { + dev->class = ATA_DEV_UNKNOWN; goto err; + } spin_lock_irqsave(ap->lock, flags); ap->pflags |= ATA_PFLAG_SCSI_HOTPLUG; @@ -2725,7 +3031,7 @@ static int ata_eh_revalidate_and_attach(struct ata_link *link, /** * ata_set_mode - Program timings and issue SET FEATURES - XFER * @link: link on which timings will be programmed - * @r_failed_dev: out paramter for failed device + * @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 @@ -2744,7 +3050,7 @@ int ata_set_mode(struct ata_link *link, struct ata_device **r_failed_dev) int rc; /* if data transfer is verified, clear DUBIOUS_XFER on ering top */ - ata_link_for_each_dev(dev, link) { + ata_for_each_dev(dev, link, ENABLED) { if (!(dev->flags & ATA_DFLAG_DUBIOUS_XFER)) { struct ata_ering_entry *ent; @@ -2761,7 +3067,7 @@ int ata_set_mode(struct ata_link *link, struct ata_device **r_failed_dev) rc = ata_do_set_mode(link, r_failed_dev); /* if transfer mode has changed, set DUBIOUS_XFER on device */ - ata_link_for_each_dev(dev, link) { + ata_for_each_dev(dev, link, ENABLED) { struct ata_eh_context *ehc = &link->eh_context; u8 saved_xfer_mode = ehc->saved_xfer_mode[dev->devno]; u8 saved_ncq = !!(ehc->saved_ncq_enabled & (1 << dev->devno)); @@ -2774,14 +3080,136 @@ int ata_set_mode(struct ata_link *link, struct ata_device **r_failed_dev) return rc; } +/** + * atapi_eh_clear_ua - Clear ATAPI UNIT ATTENTION after reset + * @dev: ATAPI device to clear UA for + * + * Resets and other operations can make an ATAPI device raise + * UNIT ATTENTION which causes the next operation to fail. This + * function clears UA. + * + * LOCKING: + * EH context (may sleep). + * + * RETURNS: + * 0 on success, -errno on failure. + */ +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_key = 0; + unsigned int err_mask; + + err_mask = atapi_eh_tur(dev, &sense_key); + if (err_mask != 0 && err_mask != AC_ERR_DEV) { + ata_dev_printk(dev, KERN_WARNING, "TEST_UNIT_READY " + "failed (err_mask=0x%x)\n", err_mask); + return -EIO; + } + + if (!err_mask || sense_key != UNIT_ATTENTION) + return 0; + + err_mask = atapi_eh_request_sense(dev, sense_buffer, sense_key); + if (err_mask) { + ata_dev_printk(dev, KERN_WARNING, "failed to clear " + "UNIT ATTENTION (err_mask=0x%x)\n", err_mask); + return -EIO; + } + } + + ata_dev_printk(dev, KERN_WARNING, + "UNIT ATTENTION persists after %d tries\n", ATA_EH_UA_TRIES); + + return 0; +} + +/** + * ata_eh_maybe_retry_flush - Retry FLUSH if necessary + * @dev: ATA device which may need FLUSH retry + * + * If @dev failed FLUSH, it needs to be reported upper layer + * immediately as it means that @dev failed to remap and already + * lost at least a sector and further FLUSH retrials won't make + * any difference to the lost sector. However, if FLUSH failed + * for other reasons, for example transmission error, FLUSH needs + * to be retried. + * + * This function determines whether FLUSH failure retry is + * necessary and performs it if so. + * + * RETURNS: + * 0 if EH can continue, -errno if EH needs to be repeated. + */ +static int ata_eh_maybe_retry_flush(struct ata_device *dev) +{ + struct ata_link *link = dev->link; + struct ata_port *ap = link->ap; + struct ata_queued_cmd *qc; + struct ata_taskfile tf; + unsigned int err_mask; + int rc = 0; + + /* did flush fail for this device? */ + if (!ata_tag_valid(link->active_tag)) + return 0; + + qc = __ata_qc_from_tag(ap, link->active_tag); + if (qc->dev != dev || (qc->tf.command != ATA_CMD_FLUSH_EXT && + qc->tf.command != ATA_CMD_FLUSH)) + return 0; + + /* if the device failed it, it should be reported to upper layers */ + if (qc->err_mask & AC_ERR_DEV) + return 0; + + /* flush failed for some other reason, give it another shot */ + ata_tf_init(dev, &tf); + + tf.command = qc->tf.command; + tf.flags |= ATA_TFLAG_DEVICE; + tf.protocol = ATA_PROT_NODATA; + + ata_dev_printk(dev, KERN_WARNING, "retrying FLUSH 0x%x Emask 0x%x\n", + tf.command, qc->err_mask); + + err_mask = ata_exec_internal(dev, &tf, NULL, DMA_NONE, NULL, 0, 0); + if (!err_mask) { + /* + * FLUSH is complete but there's no way to + * successfully complete a failed command from EH. + * Making sure retry is allowed at least once and + * retrying it should do the trick - whatever was in + * the cache is already on the platter and this won't + * cause infinite loop. + */ + qc->scsicmd->allowed = max(qc->scsicmd->allowed, 1); + } else { + ata_dev_printk(dev, KERN_WARNING, "FLUSH failed Emask 0x%x\n", + err_mask); + rc = -EIO; + + /* if device failed it, report it to upper layers */ + if (err_mask & AC_ERR_DEV) { + qc->err_mask |= AC_ERR_DEV; + qc->result_tf = tf; + if (!(ap->pflags & ATA_PFLAG_FROZEN)) + rc = 0; + } + } + return rc; +} + static int ata_link_nr_enabled(struct ata_link *link) { struct ata_device *dev; int cnt = 0; - ata_link_for_each_dev(dev, link) - if (ata_dev_enabled(dev)) - cnt++; + ata_for_each_dev(dev, link, ENABLED) + cnt++; return cnt; } @@ -2790,7 +3218,7 @@ static int ata_link_nr_vacant(struct ata_link *link) struct ata_device *dev; int cnt = 0; - ata_link_for_each_dev(dev, link) + ata_for_each_dev(dev, link, ALL) if (dev->class == ATA_DEV_UNKNOWN) cnt++; return cnt; @@ -2816,7 +3244,7 @@ static int ata_eh_skip_recovery(struct ata_link *link) return 0; /* skip if class codes for all vacant slots are ATA_DEV_NONE */ - ata_link_for_each_dev(dev, link) { + ata_for_each_dev(dev, link, ALL) { if (dev->class == ATA_DEV_UNKNOWN && ehc->classes[dev->devno] != ATA_DEV_NONE) return 0; @@ -2825,9 +3253,24 @@ static int ata_eh_skip_recovery(struct ata_link *link) return 1; } +static int ata_count_probe_trials_cb(struct ata_ering_entry *ent, void *void_arg) +{ + u64 interval = msecs_to_jiffies(ATA_EH_PROBE_TRIAL_INTERVAL); + u64 now = get_jiffies_64(); + int *trials = void_arg; + + if (ent->timestamp < now - min(now, interval)) + return -1; + + (*trials)++; + return 0; +} + static int ata_eh_schedule_probe(struct ata_device *dev) { struct ata_eh_context *ehc = &dev->link->eh_context; + struct ata_link *link = ata_dev_phys_link(dev); + int trials = 0; if (!(ehc->i.probe_mask & (1 << dev->devno)) || (ehc->did_probe_mask & (1 << dev->devno))) @@ -2840,6 +3283,25 @@ static int ata_eh_schedule_probe(struct ata_device *dev) ehc->saved_xfer_mode[dev->devno] = 0; ehc->saved_ncq_enabled &= ~(1 << dev->devno); + /* Record and count probe trials on the ering. The specific + * error mask used is irrelevant. Because a successful device + * detection clears the ering, this count accumulates only if + * there are consecutive failed probes. + * + * If the count is equal to or higher than ATA_EH_PROBE_TRIALS + * in the last ATA_EH_PROBE_TRIAL_INTERVAL, link speed is + * forced to 1.5Gbps. + * + * This is to work around cases where failed link speed + * negotiation results in device misdetection leading to + * infinite DEVXCHG or PHRDY CHG events. + */ + ata_ering_record(&dev->ering, 0, AC_ERR_OTHER); + ata_ering_map(&dev->ering, ata_count_probe_trials_cb, &trials); + + if (trials > ATA_EH_PROBE_TRIALS) + sata_down_spd_limit(link, 1); + return 1; } @@ -2847,7 +3309,11 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) { struct ata_eh_context *ehc = &dev->link->eh_context; - ehc->tries[dev->devno]--; + /* -EAGAIN from EH routine indicates retry without prejudice. + * The requester is responsible for ensuring forward progress. + */ + if (err != -EAGAIN) + ehc->tries[dev->devno]--; switch (err) { case -ENODEV: @@ -2857,12 +3323,13 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err) /* give it just one more chance */ ehc->tries[dev->devno] = min(ehc->tries[dev->devno], 1); case -EIO: - if (ehc->tries[dev->devno] == 1 && dev->pio_mode > XFER_PIO_0) { + if (ehc->tries[dev->devno] == 1) { /* This is the last chance, better to slow * down than lose it. */ - sata_down_spd_limit(ata_dev_phys_link(dev)); - ata_down_xfermask_limit(dev, ATA_DNXFER_PIO); + sata_down_spd_limit(ata_dev_phys_link(dev), 0); + if (dev->pio_mode > XFER_PIO_0) + ata_down_xfermask_limit(dev, ATA_DNXFER_PIO); } } @@ -2924,7 +3391,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, DPRINTK("ENTER\n"); /* prep for recovery */ - ata_port_for_each_link(link, ap) { + ata_for_each_link(link, ap, EDGE) { struct ata_eh_context *ehc = &link->eh_context; /* re-enable link? */ @@ -2936,7 +3403,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, ata_eh_done(link, NULL, ATA_EH_ENABLE_LINK); } - ata_link_for_each_dev(dev, link) { + ata_for_each_dev(dev, link, ALL) { if (link->flags & ATA_LFLAG_NO_RETRY) ehc->tries[dev->devno] = 1; else @@ -2966,19 +3433,19 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, goto out; /* prep for EH */ - ata_port_for_each_link(link, ap) { + ata_for_each_link(link, ap, EDGE) { struct ata_eh_context *ehc = &link->eh_context; /* skip EH if possible. */ if (ata_eh_skip_recovery(link)) ehc->i.action = 0; - ata_link_for_each_dev(dev, link) + ata_for_each_dev(dev, link, ALL) ehc->classes[dev->devno] = ATA_DEV_UNKNOWN; } /* reset */ - ata_port_for_each_link(link, ap) { + ata_for_each_link(link, ap, EDGE) { struct ata_eh_context *ehc = &link->eh_context; if (!(ehc->i.action & ATA_EH_RESET)) @@ -3003,8 +3470,8 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, ata_eh_pull_park_action(ap); deadline = jiffies; - ata_port_for_each_link(link, ap) { - ata_link_for_each_dev(dev, link) { + ata_for_each_link(link, ap, EDGE) { + ata_for_each_dev(dev, link, ALL) { struct ata_eh_context *ehc = &link->eh_context; unsigned long tmp; @@ -3032,8 +3499,8 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, deadline = wait_for_completion_timeout(&ap->park_req_pending, deadline - now); } while (deadline); - ata_port_for_each_link(link, ap) { - ata_link_for_each_dev(dev, link) { + ata_for_each_link(link, ap, EDGE) { + ata_for_each_dev(dev, link, ALL) { if (!(link->eh_context.unloaded_mask & (1 << dev->devno))) continue; @@ -3044,7 +3511,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, } /* the rest */ - ata_port_for_each_link(link, ap) { + ata_for_each_link(link, ap, EDGE) { struct ata_eh_context *ehc = &link->eh_context; /* revalidate existing devices and attach new ones */ @@ -3066,8 +3533,31 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset, ehc->i.flags &= ~ATA_EHI_SETMODE; } + /* If reset has been issued, clear UA to avoid + * disrupting the current users of the device. + */ + if (ehc->i.flags & ATA_EHI_DID_RESET) { + ata_for_each_dev(dev, link, ALL) { + if (dev->class != ATA_DEV_ATAPI) + continue; + rc = atapi_eh_clear_ua(dev); + if (rc) + goto dev_fail; + } + } + + /* retry flush if necessary */ + ata_for_each_dev(dev, link, ALL) { + if (dev->class != ATA_DEV_ATA) + continue; + rc = ata_eh_maybe_retry_flush(dev); + if (rc) + goto dev_fail; + } + + /* configure link power saving */ if (ehc->i.action & ATA_EH_LPM) - ata_link_for_each_dev(dev, link) + ata_for_each_dev(dev, link, ALL) ata_dev_enable_pm(dev, ap->pm_policy); /* this link is okay now */ @@ -3172,7 +3662,7 @@ void ata_do_eh(struct ata_port *ap, ata_prereset_fn_t prereset, rc = ata_eh_recover(ap, prereset, softreset, hardreset, postreset, NULL); if (rc) { - ata_link_for_each_dev(dev, &ap->link) + ata_for_each_dev(dev, &ap->link, ALL) ata_dev_disable(dev); } @@ -3194,7 +3684,7 @@ void ata_std_error_handler(struct ata_port *ap) ata_reset_fn_t hardreset = ops->hardreset; /* ignore built-in hardreset if SCR access is not available */ - if (ata_is_builtin_hardreset(hardreset) && !sata_scr_valid(&ap->link)) + if (hardreset == sata_std_hardreset && !sata_scr_valid(&ap->link)) hardreset = NULL; ata_do_eh(ap, ops->prereset, ops->softreset, hardreset, ops->postreset); @@ -3269,6 +3759,8 @@ static void ata_eh_handle_port_suspend(struct ata_port *ap) */ 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; @@ -3283,6 +3775,17 @@ static void ata_eh_handle_port_resume(struct ata_port *ap) WARN_ON(!(ap->pflags & ATA_PFLAG_SUSPENDED)); + /* + * Error timestamps are in jiffies which doesn't run while + * suspended and PHY events during resume isn't too uncommon. + * When the two are combined, it can lead to unnecessary speed + * downs if the machine is suspended and resumed repeatedly. + * Clear error history. + */ + ata_for_each_link(link, ap, HOST_FIRST) + ata_for_each_dev(dev, link, ALL) + ata_ering_clear(&dev->ering); + ata_acpi_set_state(ap, PMSG_ON); if (ap->ops->port_resume)