diff options
Diffstat (limited to 'drivers/scsi/scsi_error.c')
| -rw-r--r-- | drivers/scsi/scsi_error.c | 2050 |
1 files changed, 2050 insertions, 0 deletions
diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c new file mode 100644 index 00000000000..9bc597bd13b --- /dev/null +++ b/drivers/scsi/scsi_error.c @@ -0,0 +1,2050 @@ +/* + * scsi_error.c Copyright (C) 1997 Eric Youngdale + * + * SCSI error/timeout handling + * Initial versions: Eric Youngdale. Based upon conversations with + * Leonard Zubkoff and David Miller at Linux Expo, + * ideas originating from all over the place. + * + * Restructured scsi_unjam_host and associated functions. + * September 04, 2002 Mike Anderson (andmike@us.ibm.com) + * + * Forward port of Russell King's (rmk@arm.linux.org.uk) changes and + * minor cleanups. + * September 30, 2002 Mike Anderson (andmike@us.ibm.com) + */ + +#include <linux/module.h> +#include <linux/sched.h> +#include <linux/timer.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/kernel.h> +#include <linux/interrupt.h> +#include <linux/blkdev.h> +#include <linux/delay.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_dbg.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_eh.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_ioctl.h> +#include <scsi/scsi_request.h> + +#include "scsi_priv.h" +#include "scsi_logging.h" + +#define SENSE_TIMEOUT (10*HZ) +#define START_UNIT_TIMEOUT (30*HZ) + +/* + * These should *probably* be handled by the host itself. + * Since it is allowed to sleep, it probably should. + */ +#define BUS_RESET_SETTLE_TIME (10) +#define HOST_RESET_SETTLE_TIME (10) + +/* called with shost->host_lock held */ +void scsi_eh_wakeup(struct Scsi_Host *shost) +{ + if (shost->host_busy == shost->host_failed) { + up(shost->eh_wait); + SCSI_LOG_ERROR_RECOVERY(5, + printk("Waking error handler thread\n")); + } +} + +/** + * scsi_eh_scmd_add - add scsi cmd to error handling. + * @scmd: scmd to run eh on. + * @eh_flag: optional SCSI_EH flag. + * + * Return value: + * 0 on failure. + **/ +int scsi_eh_scmd_add(struct scsi_cmnd *scmd, int eh_flag) +{ + struct Scsi_Host *shost = scmd->device->host; + unsigned long flags; + + if (shost->eh_wait == NULL) + return 0; + + spin_lock_irqsave(shost->host_lock, flags); + + scsi_eh_eflags_set(scmd, eh_flag); + /* + * FIXME: Can we stop setting owner and state. + */ + scmd->owner = SCSI_OWNER_ERROR_HANDLER; + scmd->state = SCSI_STATE_FAILED; + /* + * Set the serial_number_at_timeout to the current + * serial_number + */ + scmd->serial_number_at_timeout = scmd->serial_number; + list_add_tail(&scmd->eh_entry, &shost->eh_cmd_q); + set_bit(SHOST_RECOVERY, &shost->shost_state); + shost->host_failed++; + scsi_eh_wakeup(shost); + spin_unlock_irqrestore(shost->host_lock, flags); + return 1; +} + +/** + * scsi_add_timer - Start timeout timer for a single scsi command. + * @scmd: scsi command that is about to start running. + * @timeout: amount of time to allow this command to run. + * @complete: timeout function to call if timer isn't canceled. + * + * Notes: + * This should be turned into an inline function. Each scsi command + * has its own timer, and as it is added to the queue, we set up the + * timer. When the command completes, we cancel the timer. + **/ +void scsi_add_timer(struct scsi_cmnd *scmd, int timeout, + void (*complete)(struct scsi_cmnd *)) +{ + + /* + * If the clock was already running for this command, then + * first delete the timer. The timer handling code gets rather + * confused if we don't do this. + */ + if (scmd->eh_timeout.function) + del_timer(&scmd->eh_timeout); + + scmd->eh_timeout.data = (unsigned long)scmd; + scmd->eh_timeout.expires = jiffies + timeout; + scmd->eh_timeout.function = (void (*)(unsigned long)) complete; + + SCSI_LOG_ERROR_RECOVERY(5, printk("%s: scmd: %p, time:" + " %d, (%p)\n", __FUNCTION__, + scmd, timeout, complete)); + + add_timer(&scmd->eh_timeout); +} +EXPORT_SYMBOL(scsi_add_timer); + +/** + * scsi_delete_timer - Delete/cancel timer for a given function. + * @scmd: Cmd that we are canceling timer for + * + * Notes: + * This should be turned into an inline function. + * + * Return value: + * 1 if we were able to detach the timer. 0 if we blew it, and the + * timer function has already started to run. + **/ +int scsi_delete_timer(struct scsi_cmnd *scmd) +{ + int rtn; + + rtn = del_timer(&scmd->eh_timeout); + + SCSI_LOG_ERROR_RECOVERY(5, printk("%s: scmd: %p," + " rtn: %d\n", __FUNCTION__, + scmd, rtn)); + + scmd->eh_timeout.data = (unsigned long)NULL; + scmd->eh_timeout.function = NULL; + + return rtn; +} +EXPORT_SYMBOL(scsi_delete_timer); + +/** + * scsi_times_out - Timeout function for normal scsi commands. + * @scmd: Cmd that is timing out. + * + * Notes: + * We do not need to lock this. There is the potential for a race + * only in that the normal completion handling might run, but if the + * normal completion function determines that the timer has already + * fired, then it mustn't do anything. + **/ +void scsi_times_out(struct scsi_cmnd *scmd) +{ + scsi_log_completion(scmd, TIMEOUT_ERROR); + + if (scmd->device->host->hostt->eh_timed_out) + switch (scmd->device->host->hostt->eh_timed_out(scmd)) { + case EH_HANDLED: + __scsi_done(scmd); + return; + case EH_RESET_TIMER: + /* This allows a single retry even of a command + * with allowed == 0 */ + if (scmd->retries++ > scmd->allowed) + break; + scsi_add_timer(scmd, scmd->timeout_per_command, + scsi_times_out); + return; + case EH_NOT_HANDLED: + break; + } + + if (unlikely(!scsi_eh_scmd_add(scmd, SCSI_EH_CANCEL_CMD))) { + panic("Error handler thread not present at %p %p %s %d", + scmd, scmd->device->host, __FILE__, __LINE__); + } +} + +/** + * scsi_block_when_processing_errors - Prevent cmds from being queued. + * @sdev: Device on which we are performing recovery. + * + * Description: + * We block until the host is out of error recovery, and then check to + * see whether the host or the device is offline. + * + * Return value: + * 0 when dev was taken offline by error recovery. 1 OK to proceed. + **/ +int scsi_block_when_processing_errors(struct scsi_device *sdev) +{ + int online; + + wait_event(sdev->host->host_wait, (!test_bit(SHOST_RECOVERY, &sdev->host->shost_state))); + + online = scsi_device_online(sdev); + + SCSI_LOG_ERROR_RECOVERY(5, printk("%s: rtn: %d\n", __FUNCTION__, + online)); + + return online; +} +EXPORT_SYMBOL(scsi_block_when_processing_errors); + +#ifdef CONFIG_SCSI_LOGGING +/** + * scsi_eh_prt_fail_stats - Log info on failures. + * @shost: scsi host being recovered. + * @work_q: Queue of scsi cmds to process. + **/ +static inline void scsi_eh_prt_fail_stats(struct Scsi_Host *shost, + struct list_head *work_q) +{ + struct scsi_cmnd *scmd; + struct scsi_device *sdev; + int total_failures = 0; + int cmd_failed = 0; + int cmd_cancel = 0; + int devices_failed = 0; + + shost_for_each_device(sdev, shost) { + list_for_each_entry(scmd, work_q, eh_entry) { + if (scmd->device == sdev) { + ++total_failures; + if (scsi_eh_eflags_chk(scmd, + SCSI_EH_CANCEL_CMD)) + ++cmd_cancel; + else + ++cmd_failed; + } + } + + if (cmd_cancel || cmd_failed) { + SCSI_LOG_ERROR_RECOVERY(3, + printk("%s: %d:%d:%d:%d cmds failed: %d," + " cancel: %d\n", + __FUNCTION__, shost->host_no, + sdev->channel, sdev->id, sdev->lun, + cmd_failed, cmd_cancel)); + cmd_cancel = 0; + cmd_failed = 0; + ++devices_failed; + } + } + + SCSI_LOG_ERROR_RECOVERY(2, printk("Total of %d commands on %d" + " devices require eh work\n", + total_failures, devices_failed)); +} +#endif + +/** + * scsi_check_sense - Examine scsi cmd sense + * @scmd: Cmd to have sense checked. + * + * Return value: + * SUCCESS or FAILED or NEEDS_RETRY + * + * Notes: + * When a deferred error is detected the current command has + * not been executed and needs retrying. + **/ +static int scsi_check_sense(struct scsi_cmnd *scmd) +{ + struct scsi_sense_hdr sshdr; + + if (! scsi_command_normalize_sense(scmd, &sshdr)) + return FAILED; /* no valid sense data */ + + if (scsi_sense_is_deferred(&sshdr)) + return NEEDS_RETRY; + + /* + * Previous logic looked for FILEMARK, EOM or ILI which are + * mainly associated with tapes and returned SUCCESS. + */ + if (sshdr.response_code == 0x70) { + /* fixed format */ + if (scmd->sense_buffer[2] & 0xe0) + return SUCCESS; + } else { + /* + * descriptor format: look for "stream commands sense data + * descriptor" (see SSC-3). Assume single sense data + * descriptor. Ignore ILI from SBC-2 READ LONG and WRITE LONG. + */ + if ((sshdr.additional_length > 3) && + (scmd->sense_buffer[8] == 0x4) && + (scmd->sense_buffer[11] & 0xe0)) + return SUCCESS; + } + + switch (sshdr.sense_key) { + case NO_SENSE: + return SUCCESS; + case RECOVERED_ERROR: + return /* soft_error */ SUCCESS; + + case ABORTED_COMMAND: + return NEEDS_RETRY; + case NOT_READY: + case UNIT_ATTENTION: + /* + * if we are expecting a cc/ua because of a bus reset that we + * performed, treat this just as a retry. otherwise this is + * information that we should pass up to the upper-level driver + * so that we can deal with it there. + */ + if (scmd->device->expecting_cc_ua) { + scmd->device->expecting_cc_ua = 0; + return NEEDS_RETRY; + } + /* + * if the device is in the process of becoming ready, we + * should retry. + */ + if ((sshdr.asc == 0x04) && (sshdr.ascq == 0x01)) + return NEEDS_RETRY; + /* + * if the device is not started, we need to wake + * the error handler to start the motor + */ + if (scmd->device->allow_restart && + (sshdr.asc == 0x04) && (sshdr.ascq == 0x02)) + return FAILED; + return SUCCESS; + + /* these three are not supported */ + case COPY_ABORTED: + case VOLUME_OVERFLOW: + case MISCOMPARE: + return SUCCESS; + + case MEDIUM_ERROR: + return NEEDS_RETRY; + + case HARDWARE_ERROR: + if (scmd->device->retry_hwerror) + return NEEDS_RETRY; + else + return SUCCESS; + + case ILLEGAL_REQUEST: + case BLANK_CHECK: + case DATA_PROTECT: + default: + return SUCCESS; + } +} + +/** + * scsi_eh_completed_normally - Disposition a eh cmd on return from LLD. + * @scmd: SCSI cmd to examine. + * + * Notes: + * This is *only* called when we are examining the status of commands + * queued during error recovery. the main difference here is that we + * don't allow for the possibility of retries here, and we are a lot + * more restrictive about what we consider acceptable. + **/ +static int scsi_eh_completed_normally(struct scsi_cmnd *scmd) +{ + /* + * first check the host byte, to see if there is anything in there + * that would indicate what we need to do. + */ + if (host_byte(scmd->result) == DID_RESET) { + /* + * rats. we are already in the error handler, so we now + * get to try and figure out what to do next. if the sense + * is valid, we have a pretty good idea of what to do. + * if not, we mark it as FAILED. + */ + return scsi_check_sense(scmd); + } + if (host_byte(scmd->result) != DID_OK) + return FAILED; + + /* + * next, check the message byte. + */ + if (msg_byte(scmd->result) != COMMAND_COMPLETE) + return FAILED; + + /* + * now, check the status byte to see if this indicates + * anything special. + */ + switch (status_byte(scmd->result)) { + case GOOD: + case COMMAND_TERMINATED: + return SUCCESS; + case CHECK_CONDITION: + return scsi_check_sense(scmd); + case CONDITION_GOOD: + case INTERMEDIATE_GOOD: + case INTERMEDIATE_C_GOOD: + /* + * who knows? FIXME(eric) + */ + return SUCCESS; + case BUSY: + case QUEUE_FULL: + case RESERVATION_CONFLICT: + default: + return FAILED; + } + return FAILED; +} + +/** + * scsi_eh_times_out - timeout function for error handling. + * @scmd: Cmd that is timing out. + * + * Notes: + * During error handling, the kernel thread will be sleeping waiting + * for some action to complete on the device. our only job is to + * record that it timed out, and to wake up the thread. + **/ +static void scsi_eh_times_out(struct scsi_cmnd *scmd) +{ + scsi_eh_eflags_set(scmd, SCSI_EH_REC_TIMEOUT); + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd:%p\n", __FUNCTION__, + scmd)); + + if (scmd->device->host->eh_action) + up(scmd->device->host->eh_action); +} + +/** + * scsi_eh_done - Completion function for error handling. + * @scmd: Cmd that is done. + **/ +static void scsi_eh_done(struct scsi_cmnd *scmd) +{ + /* + * if the timeout handler is already running, then just set the + * flag which says we finished late, and return. we have no + * way of stopping the timeout handler from running, so we must + * always defer to it. + */ + if (del_timer(&scmd->eh_timeout)) { + scmd->request->rq_status = RQ_SCSI_DONE; + scmd->owner = SCSI_OWNER_ERROR_HANDLER; + + SCSI_LOG_ERROR_RECOVERY(3, printk("%s scmd: %p result: %x\n", + __FUNCTION__, scmd, scmd->result)); + + if (scmd->device->host->eh_action) + up(scmd->device->host->eh_action); + } +} + +/** + * scsi_send_eh_cmnd - send a cmd to a device as part of error recovery. + * @scmd: SCSI Cmd to send. + * @timeout: Timeout for cmd. + * + * Notes: + * The initialization of the structures is quite a bit different in + * this case, and furthermore, there is a different completion handler + * vs scsi_dispatch_cmd. + * Return value: + * SUCCESS or FAILED or NEEDS_RETRY + **/ +static int scsi_send_eh_cmnd(struct scsi_cmnd *scmd, int timeout) +{ + struct Scsi_Host *host = scmd->device->host; + DECLARE_MUTEX_LOCKED(sem); + unsigned long flags; + int rtn = SUCCESS; + + /* + * we will use a queued command if possible, otherwise we will + * emulate the queuing and calling of completion function ourselves. + */ + scmd->owner = SCSI_OWNER_LOWLEVEL; + + if (scmd->device->scsi_level <= SCSI_2) + scmd->cmnd[1] = (scmd->cmnd[1] & 0x1f) | + (scmd->device->lun << 5 & 0xe0); + + scsi_add_timer(scmd, timeout, scsi_eh_times_out); + + /* + * set up the semaphore so we wait for the command to complete. + */ + scmd->device->host->eh_action = &sem; + scmd->request->rq_status = RQ_SCSI_BUSY; + + spin_lock_irqsave(scmd->device->host->host_lock, flags); + scsi_log_send(scmd); + host->hostt->queuecommand(scmd, scsi_eh_done); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + + down(&sem); + scsi_log_completion(scmd, SUCCESS); + + scmd->device->host->eh_action = NULL; + + /* + * see if timeout. if so, tell the host to forget about it. + * in other words, we don't want a callback any more. + */ + if (scsi_eh_eflags_chk(scmd, SCSI_EH_REC_TIMEOUT)) { + scsi_eh_eflags_clr(scmd, SCSI_EH_REC_TIMEOUT); + scmd->owner = SCSI_OWNER_LOWLEVEL; + + /* + * as far as the low level driver is + * concerned, this command is still active, so + * we must give the low level driver a chance + * to abort it. (db) + * + * FIXME(eric) - we are not tracking whether we could + * abort a timed out command or not. not sure how + * we should treat them differently anyways. + */ + spin_lock_irqsave(scmd->device->host->host_lock, flags); + if (scmd->device->host->hostt->eh_abort_handler) + scmd->device->host->hostt->eh_abort_handler(scmd); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + + scmd->request->rq_status = RQ_SCSI_DONE; + scmd->owner = SCSI_OWNER_ERROR_HANDLER; + + rtn = FAILED; + } + + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd: %p, rtn:%x\n", + __FUNCTION__, scmd, rtn)); + + /* + * now examine the actual status codes to see whether the command + * actually did complete normally. + */ + if (rtn == SUCCESS) { + rtn = scsi_eh_completed_normally(scmd); + SCSI_LOG_ERROR_RECOVERY(3, + printk("%s: scsi_eh_completed_normally %x\n", + __FUNCTION__, rtn)); + switch (rtn) { + case SUCCESS: + case NEEDS_RETRY: + case FAILED: + break; + default: + rtn = FAILED; + break; + } + } + + return rtn; +} + +/** + * scsi_request_sense - Request sense data from a particular target. + * @scmd: SCSI cmd for request sense. + * + * Notes: + * Some hosts automatically obtain this information, others require + * that we obtain it on our own. This function will *not* return until + * the command either times out, or it completes. + **/ +static int scsi_request_sense(struct scsi_cmnd *scmd) +{ + static unsigned char generic_sense[6] = + {REQUEST_SENSE, 0, 0, 0, 252, 0}; + unsigned char *scsi_result; + int saved_result; + int rtn; + + memcpy(scmd->cmnd, generic_sense, sizeof(generic_sense)); + + scsi_result = kmalloc(252, GFP_ATOMIC | (scmd->device->host->hostt->unchecked_isa_dma) ? __GFP_DMA : 0); + + + if (unlikely(!scsi_result)) { + printk(KERN_ERR "%s: cannot allocate scsi_result.\n", + __FUNCTION__); + return FAILED; + } + + /* + * zero the sense buffer. some host adapters automatically always + * request sense, so it is not a good idea that + * scmd->request_buffer and scmd->sense_buffer point to the same + * address (db). 0 is not a valid sense code. + */ + memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); + memset(scsi_result, 0, 252); + + saved_result = scmd->result; + scmd->request_buffer = scsi_result; + scmd->request_bufflen = 252; + scmd->use_sg = 0; + scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); + scmd->sc_data_direction = DMA_FROM_DEVICE; + scmd->underflow = 0; + + rtn = scsi_send_eh_cmnd(scmd, SENSE_TIMEOUT); + + /* last chance to have valid sense data */ + if(!SCSI_SENSE_VALID(scmd)) { + memcpy(scmd->sense_buffer, scmd->request_buffer, + sizeof(scmd->sense_buffer)); + } + + kfree(scsi_result); + + /* + * when we eventually call scsi_finish, we really wish to complete + * the original request, so let's restore the original data. (db) + */ + scsi_setup_cmd_retry(scmd); + scmd->result = saved_result; + return rtn; +} + +/** + * scsi_eh_finish_cmd - Handle a cmd that eh is finished with. + * @scmd: Original SCSI cmd that eh has finished. + * @done_q: Queue for processed commands. + * + * Notes: + * We don't want to use the normal command completion while we are are + * still handling errors - it may cause other commands to be queued, + * and that would disturb what we are doing. thus we really want to + * keep a list of pending commands for final completion, and once we + * are ready to leave error handling we handle completion for real. + **/ +static void scsi_eh_finish_cmd(struct scsi_cmnd *scmd, + struct list_head *done_q) +{ + scmd->device->host->host_failed--; + scmd->state = SCSI_STATE_BHQUEUE; + + scsi_eh_eflags_clr_all(scmd); + + /* + * set this back so that the upper level can correctly free up + * things. + */ + scsi_setup_cmd_retry(scmd); + list_move_tail(&scmd->eh_entry, done_q); +} + +/** + * scsi_eh_get_sense - Get device sense data. + * @work_q: Queue of commands to process. + * @done_q: Queue of proccessed commands.. + * + * Description: + * See if we need to request sense information. if so, then get it + * now, so we have a better idea of what to do. + * + * Notes: + * This has the unfortunate side effect that if a shost adapter does + * not automatically request sense information, that we end up shutting + * it down before we request it. + * + * All drivers should request sense information internally these days, + * so for now all I have to say is tough noogies if you end up in here. + * + * XXX: Long term this code should go away, but that needs an audit of + * all LLDDs first. + **/ +static int scsi_eh_get_sense(struct list_head *work_q, + struct list_head *done_q) +{ + struct list_head *lh, *lh_sf; + struct scsi_cmnd *scmd; + int rtn; + + list_for_each_safe(lh, lh_sf, work_q) { + scmd = list_entry(lh, struct scsi_cmnd, eh_entry); + if (scsi_eh_eflags_chk(scmd, SCSI_EH_CANCEL_CMD) || + SCSI_SENSE_VALID(scmd)) + continue; + + SCSI_LOG_ERROR_RECOVERY(2, printk("%s: requesting sense" + " for id: %d\n", + current->comm, + scmd->device->id)); + rtn = scsi_request_sense(scmd); + if (rtn != SUCCESS) + continue; + + SCSI_LOG_ERROR_RECOVERY(3, printk("sense requested for %p" + " result %x\n", scmd, + scmd->result)); + SCSI_LOG_ERROR_RECOVERY(3, scsi_print_sense("bh", scmd)); + + rtn = scsi_decide_disposition(scmd); + + /* + * if the result was normal, then just pass it along to the + * upper level. + */ + if (rtn == SUCCESS) + /* we don't want this command reissued, just + * finished with the sense data, so set + * retries to the max allowed to ensure it + * won't get reissued */ + scmd->retries = scmd->allowed; + else if (rtn != NEEDS_RETRY) + continue; + + scsi_eh_finish_cmd(scmd, done_q); + } + + return list_empty(work_q); +} + +/** + * scsi_try_to_abort_cmd - Ask host to abort a running command. + * @scmd: SCSI cmd to abort from Lower Level. + * + * Notes: + * This function will not return until the user's completion function + * has been called. there is no timeout on this operation. if the + * author of the low-level driver wishes this operation to be timed, + * they can provide this facility themselves. helper functions in + * scsi_error.c can be supplied to make this easier to do. + **/ +static int scsi_try_to_abort_cmd(struct scsi_cmnd *scmd) +{ + unsigned long flags; + int rtn = FAILED; + + if (!scmd->device->host->hostt->eh_abort_handler) + return rtn; + + /* + * scsi_done was called just after the command timed out and before + * we had a chance to process it. (db) + */ + if (scmd->serial_number == 0) + return SUCCESS; + + scmd->owner = SCSI_OWNER_LOWLEVEL; + + spin_lock_irqsave(scmd->device->host->host_lock, flags); + rtn = scmd->device->host->hostt->eh_abort_handler(scmd); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + + return rtn; +} + +/** + * scsi_eh_tur - Send TUR to device. + * @scmd: Scsi cmd to send TUR + * + * Return value: + * 0 - Device is ready. 1 - Device NOT ready. + **/ +static int scsi_eh_tur(struct scsi_cmnd *scmd) +{ + static unsigned char tur_command[6] = {TEST_UNIT_READY, 0, 0, 0, 0, 0}; + int retry_cnt = 1, rtn; + +retry_tur: + memcpy(scmd->cmnd, tur_command, sizeof(tur_command)); + + /* + * zero the sense buffer. the scsi spec mandates that any + * untransferred sense data should be interpreted as being zero. + */ + memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); + + scmd->request_buffer = NULL; + scmd->request_bufflen = 0; + scmd->use_sg = 0; + scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); + scmd->underflow = 0; + scmd->sc_data_direction = DMA_NONE; + + rtn = scsi_send_eh_cmnd(scmd, SENSE_TIMEOUT); + + /* + * when we eventually call scsi_finish, we really wish to complete + * the original request, so let's restore the original data. (db) + */ + scsi_setup_cmd_retry(scmd); + + /* + * hey, we are done. let's look to see what happened. + */ + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd %p rtn %x\n", + __FUNCTION__, scmd, rtn)); + if (rtn == SUCCESS) + return 0; + else if (rtn == NEEDS_RETRY) + if (retry_cnt--) + goto retry_tur; + return 1; +} + +/** + * scsi_eh_abort_cmds - abort canceled commands. + * @shost: scsi host being recovered. + * @eh_done_q: list_head for processed commands. + * + * Decription: + * Try and see whether or not it makes sense to try and abort the + * running command. this only works out to be the case if we have one + * command that has timed out. if the command simply failed, it makes + * no sense to try and abort the command, since as far as the shost + * adapter is concerned, it isn't running. + **/ +static int scsi_eh_abort_cmds(struct list_head *work_q, + struct list_head *done_q) +{ + struct list_head *lh, *lh_sf; + struct scsi_cmnd *scmd; + int rtn; + + list_for_each_safe(lh, lh_sf, work_q) { + scmd = list_entry(lh, struct scsi_cmnd, eh_entry); + if (!scsi_eh_eflags_chk(scmd, SCSI_EH_CANCEL_CMD)) + continue; + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: aborting cmd:" + "0x%p\n", current->comm, + scmd)); + rtn = scsi_try_to_abort_cmd(scmd); + if (rtn == SUCCESS) { + scsi_eh_eflags_clr(scmd, SCSI_EH_CANCEL_CMD); + if (!scsi_device_online(scmd->device) || + !scsi_eh_tur(scmd)) { + scsi_eh_finish_cmd(scmd, done_q); + } + + } else + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: aborting" + " cmd failed:" + "0x%p\n", + current->comm, + scmd)); + } + + return list_empty(work_q); +} + +/** + * scsi_try_bus_device_reset - Ask host to perform a BDR on a dev + * @scmd: SCSI cmd used to send BDR + * + * Notes: + * There is no timeout for this operation. if this operation is + * unreliable for a given host, then the host itself needs to put a + * timer on it, and set the host back to a consistent state prior to + * returning. + **/ +static int scsi_try_bus_device_reset(struct scsi_cmnd *scmd) +{ + unsigned long flags; + int rtn = FAILED; + + if (!scmd->device->host->hostt->eh_device_reset_handler) + return rtn; + + scmd->owner = SCSI_OWNER_LOWLEVEL; + + spin_lock_irqsave(scmd->device->host->host_lock, flags); + rtn = scmd->device->host->hostt->eh_device_reset_handler(scmd); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + + if (rtn == SUCCESS) { + scmd->device->was_reset = 1; + scmd->device->expecting_cc_ua = 1; + } + + return rtn; +} + +/** + * scsi_eh_try_stu - Send START_UNIT to device. + * @scmd: Scsi cmd to send START_UNIT + * + * Return value: + * 0 - Device is ready. 1 - Device NOT ready. + **/ +static int scsi_eh_try_stu(struct scsi_cmnd *scmd) +{ + static unsigned char stu_command[6] = {START_STOP, 0, 0, 0, 1, 0}; + int rtn; + + if (!scmd->device->allow_restart) + return 1; + + memcpy(scmd->cmnd, stu_command, sizeof(stu_command)); + + /* + * zero the sense buffer. the scsi spec mandates that any + * untransferred sense data should be interpreted as being zero. + */ + memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); + + scmd->request_buffer = NULL; + scmd->request_bufflen = 0; + scmd->use_sg = 0; + scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); + scmd->underflow = 0; + scmd->sc_data_direction = DMA_NONE; + + rtn = scsi_send_eh_cmnd(scmd, START_UNIT_TIMEOUT); + + /* + * when we eventually call scsi_finish, we really wish to complete + * the original request, so let's restore the original data. (db) + */ + scsi_setup_cmd_retry(scmd); + + /* + * hey, we are done. let's look to see what happened. + */ + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd %p rtn %x\n", + __FUNCTION__, scmd, rtn)); + if (rtn == SUCCESS) + return 0; + return 1; +} + + /** + * scsi_eh_stu - send START_UNIT if needed + * @shost: scsi host being recovered. + * @eh_done_q: list_head for processed commands. + * + * Notes: + * If commands are failing due to not ready, initializing command required, + * try revalidating the device, which will end up sending a start unit. + **/ +static int scsi_eh_stu(struct Scsi_Host *shost, + struct list_head *work_q, + struct list_head *done_q) +{ + struct list_head *lh, *lh_sf; + struct scsi_cmnd *scmd, *stu_scmd; + struct scsi_device *sdev; + + shost_for_each_device(sdev, shost) { + stu_scmd = NULL; + list_for_each_entry(scmd, work_q, eh_entry) + if (scmd->device == sdev && SCSI_SENSE_VALID(scmd) && + scsi_check_sense(scmd) == FAILED ) { + stu_scmd = scmd; + break; + } + + if (!stu_scmd) + continue; + + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Sending START_UNIT to sdev:" + " 0x%p\n", current->comm, sdev)); + + if (!scsi_eh_try_stu(stu_scmd)) { + if (!scsi_device_online(sdev) || + !scsi_eh_tur(stu_scmd)) { + list_for_each_safe(lh, lh_sf, work_q) { + scmd = list_entry(lh, struct scsi_cmnd, eh_entry); + if (scmd->device == sdev) + scsi_eh_finish_cmd(scmd, done_q); + } + } + } else { + SCSI_LOG_ERROR_RECOVERY(3, + printk("%s: START_UNIT failed to sdev:" + " 0x%p\n", current->comm, sdev)); + } + } + + return list_empty(work_q); +} + + +/** + * scsi_eh_bus_device_reset - send bdr if needed + * @shost: scsi host being recovered. + * @eh_done_q: list_head for processed commands. + * + * Notes: + * Try a bus device reset. still, look to see whether we have multiple + * devices that are jammed or not - if we have multiple devices, it + * makes no sense to try bus_device_reset - we really would need to try + * a bus_reset instead. + **/ +static int scsi_eh_bus_device_reset(struct Scsi_Host *shost, + struct list_head *work_q, + struct list_head *done_q) +{ + struct list_head *lh, *lh_sf; + struct scsi_cmnd *scmd, *bdr_scmd; + struct scsi_device *sdev; + int rtn; + + shost_for_each_device(sdev, shost) { + bdr_scmd = NULL; + list_for_each_entry(scmd, work_q, eh_entry) + if (scmd->device == sdev) { + bdr_scmd = scmd; + break; + } + + if (!bdr_scmd) + continue; + + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Sending BDR sdev:" + " 0x%p\n", current->comm, + sdev)); + rtn = scsi_try_bus_device_reset(bdr_scmd); + if (rtn == SUCCESS) { + if (!scsi_device_online(sdev) || + !scsi_eh_tur(bdr_scmd)) { + list_for_each_safe(lh, lh_sf, + work_q) { + scmd = list_entry(lh, struct + scsi_cmnd, + eh_entry); + if (scmd->device == sdev) + scsi_eh_finish_cmd(scmd, + done_q); + } + } + } else { + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: BDR" + " failed sdev:" + "0x%p\n", + current->comm, + sdev)); + } + } + + return list_empty(work_q); +} + +/** + * scsi_try_bus_reset - ask host to perform a bus reset + * @scmd: SCSI cmd to send bus reset. + **/ +static int scsi_try_bus_reset(struct scsi_cmnd *scmd) +{ + unsigned long flags; + int rtn; + + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Snd Bus RST\n", + __FUNCTION__)); + scmd->owner = SCSI_OWNER_LOWLEVEL; + scmd->serial_number_at_timeout = scmd->serial_number; + + if (!scmd->device->host->hostt->eh_bus_reset_handler) + return FAILED; + + spin_lock_irqsave(scmd->device->host->host_lock, flags); + rtn = scmd->device->host->hostt->eh_bus_reset_handler(scmd); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + + if (rtn == SUCCESS) { + if (!scmd->device->host->hostt->skip_settle_delay) + ssleep(BUS_RESET_SETTLE_TIME); + spin_lock_irqsave(scmd->device->host->host_lock, flags); + scsi_report_bus_reset(scmd->device->host, scmd->device->channel); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + } + + return rtn; +} + +/** + * scsi_try_host_reset - ask host adapter to reset itself + * @scmd: SCSI cmd to send hsot reset. + **/ +static int scsi_try_host_reset(struct scsi_cmnd *scmd) +{ + unsigned long flags; + int rtn; + + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Snd Host RST\n", + __FUNCTION__)); + scmd->owner = SCSI_OWNER_LOWLEVEL; + scmd->serial_number_at_timeout = scmd->serial_number; + + if (!scmd->device->host->hostt->eh_host_reset_handler) + return FAILED; + + spin_lock_irqsave(scmd->device->host->host_lock, flags); + rtn = scmd->device->host->hostt->eh_host_reset_handler(scmd); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + + if (rtn == SUCCESS) { + if (!scmd->device->host->hostt->skip_settle_delay) + ssleep(HOST_RESET_SETTLE_TIME); + spin_lock_irqsave(scmd->device->host->host_lock, flags); + scsi_report_bus_reset(scmd->device->host, scmd->device->channel); + spin_unlock_irqrestore(scmd->device->host->host_lock, flags); + } + + return rtn; +} + +/** + * scsi_eh_bus_reset - send a bus reset + * @shost: scsi host being recovered. + * @eh_done_q: list_head for processed commands. + **/ +static int scsi_eh_bus_reset(struct Scsi_Host *shost, + struct list_head *work_q, + struct list_head *done_q) +{ + struct list_head *lh, *lh_sf; + struct scsi_cmnd *scmd; + struct scsi_cmnd *chan_scmd; + unsigned int channel; + int rtn; + + /* + * we really want to loop over the various channels, and do this on + * a channel by channel basis. we should also check to see if any + * of the failed commands are on soft_reset devices, and if so, skip + * the reset. + */ + + for (channel = 0; channel <= shost->max_channel; channel++) { + chan_scmd = NULL; + list_for_each_entry(scmd, work_q, eh_entry) { + if (channel == scmd->device->channel) { + chan_scmd = scmd; + break; + /* + * FIXME add back in some support for + * soft_reset devices. + */ + } + } + + if (!chan_scmd) + continue; + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Sending BRST chan:" + " %d\n", current->comm, + channel)); + rtn = scsi_try_bus_reset(chan_scmd); + if (rtn == SUCCESS) { + list_for_each_safe(lh, lh_sf, work_q) { + scmd = list_entry(lh, struct scsi_cmnd, + eh_entry); + if (channel == scmd->device->channel) + if (!scsi_device_online(scmd->device) || + !scsi_eh_tur(scmd)) + scsi_eh_finish_cmd(scmd, + done_q); + } + } else { + SCSI_LOG_ERROR_RECOVERY(3, printk("%s: BRST" + " failed chan: %d\n", + current->comm, + |