#include <linux/string.h>
#include <linux/slab.h>
#include <linux/kernel.h>
+#include <linux/kthread.h>
#include <linux/interrupt.h>
#include <linux/blkdev.h>
#include <linux/delay.h>
void scsi_eh_wakeup(struct Scsi_Host *shost)
{
if (shost->host_busy == shost->host_failed) {
- up(shost->eh_wait);
+ wake_up_process(shost->ehandler);
SCSI_LOG_ERROR_RECOVERY(5,
printk("Waking error handler thread\n"));
}
{
struct Scsi_Host *shost = scmd->device->host;
unsigned long flags;
+ int ret = 0;
- if (shost->eh_wait == NULL)
+ if (!shost->ehandler)
return 0;
spin_lock_irqsave(shost->host_lock, flags);
+ if (scsi_host_set_state(shost, SHOST_RECOVERY))
+ if (scsi_host_set_state(shost, SHOST_CANCEL_RECOVERY))
+ goto out_unlock;
- 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;
+ ret = 1;
+ scmd->eh_eflags |= eh_flag;
list_add_tail(&scmd->eh_entry, &shost->eh_cmd_q);
- set_bit(SHOST_RECOVERY, &shost->shost_state);
shost->host_failed++;
scsi_eh_wakeup(shost);
+ out_unlock:
spin_unlock_irqrestore(shost->host_lock, flags);
- return 1;
+ return ret;
}
/**
add_timer(&scmd->eh_timeout);
}
-EXPORT_SYMBOL(scsi_add_timer);
/**
* scsi_delete_timer - Delete/cancel timer for a given function.
return rtn;
}
-EXPORT_SYMBOL(scsi_delete_timer);
/**
* scsi_times_out - Timeout function for normal scsi commands.
}
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__);
+ scmd->result |= DID_TIME_OUT << 16;
+ __scsi_done(scmd);
}
}
{
int online;
- wait_event(sdev->host->host_wait, (!test_bit(SHOST_RECOVERY, &sdev->host->shost_state)));
+ wait_event(sdev->host->host_wait, !scsi_host_in_recovery(sdev->host));
online = scsi_device_online(sdev);
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))
+ if (scmd->eh_eflags & 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));
+ sdev_printk(KERN_INFO, sdev,
+ "%s: cmds failed: %d, cancel: %d\n",
+ __FUNCTION__, cmd_failed,
+ cmd_cancel));
cmd_cancel = 0;
cmd_failed = 0;
++devices_failed;
**/
static void scsi_eh_times_out(struct scsi_cmnd *scmd)
{
- scsi_eh_eflags_set(scmd, SCSI_EH_REC_TIMEOUT);
+ scmd->eh_eflags |= SCSI_EH_REC_TIMEOUT;
SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd:%p\n", __FUNCTION__,
scmd));
*/
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));
* 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 (sdev->scsi_level <= SCSI_2)
scmd->cmnd[1] = (scmd->cmnd[1] & 0x1f) |
(sdev->lun << 5 & 0xe0);
* 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;
+ if (scmd->eh_eflags & SCSI_EH_REC_TIMEOUT) {
+ scmd->eh_eflags &= ~SCSI_EH_REC_TIMEOUT;
/*
* as far as the low level driver is
shost->hostt->eh_abort_handler(scmd);
scmd->request->rq_status = RQ_SCSI_DONE;
- scmd->owner = SCSI_OWNER_ERROR_HANDLER;
-
rtn = FAILED;
}
struct list_head *done_q)
{
scmd->device->host->host_failed--;
- scmd->state = SCSI_STATE_BHQUEUE;
-
- scsi_eh_eflags_clr_all(scmd);
+ scmd->eh_eflags = 0;
/*
* set this back so that the upper level can correctly free up
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;
+ struct scsi_cmnd *scmd, *next;
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) ||
+ list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
+ if ((scmd->eh_eflags & SCSI_EH_CANCEL_CMD) ||
SCSI_SENSE_VALID(scmd))
continue;
*/
if (scmd->serial_number == 0)
return SUCCESS;
-
- scmd->owner = SCSI_OWNER_LOWLEVEL;
-
return scmd->device->host->hostt->eh_abort_handler(scmd);
}
__FUNCTION__, scmd, rtn));
if (rtn == SUCCESS)
return 0;
- else if (rtn == NEEDS_RETRY)
+ else if (rtn == NEEDS_RETRY) {
if (retry_cnt--)
goto retry_tur;
+ return 0;
+ }
return 1;
}
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;
+ struct scsi_cmnd *scmd, *next;
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))
+ list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
+ if (!(scmd->eh_eflags & 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);
+ scmd->eh_eflags &= ~SCSI_EH_CANCEL_CMD;
if (!scsi_device_online(scmd->device) ||
!scsi_eh_tur(scmd)) {
scsi_eh_finish_cmd(scmd, done_q);
if (!scmd->device->host->hostt->eh_device_reset_handler)
return FAILED;
- scmd->owner = SCSI_OWNER_LOWLEVEL;
-
rtn = scmd->device->host->hostt->eh_device_reset_handler(scmd);
-
if (rtn == SUCCESS) {
scmd->device->was_reset = 1;
scmd->device->expecting_cc_ua = 1;
struct list_head *work_q,
struct list_head *done_q)
{
- struct list_head *lh, *lh_sf;
- struct scsi_cmnd *scmd, *stu_scmd;
+ struct scsi_cmnd *scmd, *stu_scmd, *next;
struct scsi_device *sdev;
shost_for_each_device(sdev, shost) {
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);
+ list_for_each_entry_safe(scmd, next,
+ work_q, eh_entry) {
if (scmd->device == sdev)
scsi_eh_finish_cmd(scmd, done_q);
}
struct list_head *work_q,
struct list_head *done_q)
{
- struct list_head *lh, *lh_sf;
- struct scsi_cmnd *scmd, *bdr_scmd;
+ struct scsi_cmnd *scmd, *bdr_scmd, *next;
struct scsi_device *sdev;
int rtn;
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);
+ list_for_each_entry_safe(scmd, next,
+ work_q, eh_entry) {
if (scmd->device == sdev)
scsi_eh_finish_cmd(scmd,
done_q);
SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Snd Bus RST\n",
__FUNCTION__));
- scmd->owner = SCSI_OWNER_LOWLEVEL;
if (!scmd->device->host->hostt->eh_bus_reset_handler)
return FAILED;
SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Snd Host RST\n",
__FUNCTION__));
- scmd->owner = SCSI_OWNER_LOWLEVEL;
if (!scmd->device->host->hostt->eh_host_reset_handler)
return FAILED;
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;
+ struct scsi_cmnd *scmd, *chan_scmd, *next;
unsigned int channel;
int rtn;
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);
+ list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
if (channel == scmd->device->channel)
if (!scsi_device_online(scmd->device) ||
!scsi_eh_tur(scmd))
static int scsi_eh_host_reset(struct list_head *work_q,
struct list_head *done_q)
{
+ struct scsi_cmnd *scmd, *next;
int rtn;
- struct list_head *lh, *lh_sf;
- struct scsi_cmnd *scmd;
if (!list_empty(work_q)) {
scmd = list_entry(work_q->next,
rtn = scsi_try_host_reset(scmd);
if (rtn == SUCCESS) {
- list_for_each_safe(lh, lh_sf, work_q) {
- scmd = list_entry(lh, struct scsi_cmnd, eh_entry);
+ list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
if (!scsi_device_online(scmd->device) ||
(!scsi_eh_try_stu(scmd) && !scsi_eh_tur(scmd)) ||
!scsi_eh_tur(scmd))
static void scsi_eh_offline_sdevs(struct list_head *work_q,
struct list_head *done_q)
{
- struct list_head *lh, *lh_sf;
- struct scsi_cmnd *scmd;
+ struct scsi_cmnd *scmd, *next;
- list_for_each_safe(lh, lh_sf, work_q) {
- scmd = list_entry(lh, struct scsi_cmnd, eh_entry);
- printk(KERN_INFO "scsi: Device offlined - not"
- " ready after error recovery: host"
- " %d channel %d id %d lun %d\n",
- scmd->device->host->host_no,
- scmd->device->channel,
- scmd->device->id,
- scmd->device->lun);
+ list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
+ sdev_printk(KERN_INFO, scmd->device,
+ "scsi: Device offlined - not"
+ " ready after error recovery\n");
scsi_device_set_state(scmd->device, SDEV_OFFLINE);
- if (scsi_eh_eflags_chk(scmd, SCSI_EH_CANCEL_CMD)) {
+ if (scmd->eh_eflags & SCSI_EH_CANCEL_CMD) {
/*
* FIXME: Handle lost cmds.
*/
return SUCCESS;
case RESERVATION_CONFLICT:
- printk(KERN_INFO "scsi: reservation conflict: host"
- " %d channel %d id %d lun %d\n",
- scmd->device->host->host_no, scmd->device->channel,
- scmd->device->id, scmd->device->lun);
+ sdev_printk(KERN_INFO, scmd->device,
+ "reservation conflict\n");
return SUCCESS; /* causes immediate i/o error */
default:
return FAILED;
static void scsi_restart_operations(struct Scsi_Host *shost)
{
struct scsi_device *sdev;
+ unsigned long flags;
/*
* If the door was locked, we need to insert a door lock request
SCSI_LOG_ERROR_RECOVERY(3, printk("%s: waking up host to restart\n",
__FUNCTION__));
- clear_bit(SHOST_RECOVERY, &shost->shost_state);
+ spin_lock_irqsave(shost->host_lock, flags);
+ if (scsi_host_set_state(shost, SHOST_RUNNING))
+ if (scsi_host_set_state(shost, SHOST_CANCEL))
+ BUG_ON(scsi_host_set_state(shost, SHOST_DEL));
+ spin_unlock_irqrestore(shost->host_lock, flags);
wake_up(&shost->host_wait);
**/
static void scsi_eh_flush_done_q(struct list_head *done_q)
{
- struct list_head *lh, *lh_sf;
- struct scsi_cmnd *scmd;
+ struct scsi_cmnd *scmd, *next;
- list_for_each_safe(lh, lh_sf, done_q) {
- scmd = list_entry(lh, struct scsi_cmnd, eh_entry);
- list_del_init(lh);
+ list_for_each_entry_safe(scmd, next, done_q, eh_entry) {
+ list_del_init(&scmd->eh_entry);
if (scsi_device_online(scmd->device) &&
!blk_noretry_request(scmd->request) &&
(++scmd->retries < scmd->allowed)) {
{
struct Scsi_Host *shost = (struct Scsi_Host *) data;
int rtn;
- DECLARE_MUTEX_LOCKED(sem);
-
- /*
- * Flush resources
- */
-
- daemonize("scsi_eh_%d", shost->host_no);
current->flags |= PF_NOFREEZE;
- shost->eh_wait = &sem;
- shost->ehandler = current;
-
+
/*
- * Wake up the thread that created us.
+ * Note - we always use TASK_INTERRUPTIBLE even if the module
+ * was loaded as part of the kernel. The reason is that
+ * UNINTERRUPTIBLE would cause this thread to be counted in
+ * the load average as a running process, and an interruptible
+ * wait doesn't.
*/
- SCSI_LOG_ERROR_RECOVERY(3, printk("Wake up parent of"
- " scsi_eh_%d\n",shost->host_no));
-
- complete(shost->eh_notify);
-
- while (1) {
- /*
- * If we get a signal, it means we are supposed to go
- * away and die. This typically happens if the user is
- * trying to unload a module.
- */
- SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler"
- " scsi_eh_%d"
- " sleeping\n",shost->host_no));
-
- /*
- * Note - we always use down_interruptible with the semaphore
- * even if the module was loaded as part of the kernel. The
- * reason is that down() will cause this thread to be counted
- * in the load average as a running process, and down
- * interruptible doesn't. Given that we need to allow this
- * thread to die if the driver was loaded as a module, using
- * semaphores isn't unreasonable.
- */
- down_interruptible(&sem);
- if (shost->eh_kill)
- break;
+ set_current_state(TASK_INTERRUPTIBLE);
+ while (!kthread_should_stop()) {
+ if (shost->host_failed == 0 ||
+ shost->host_failed != shost->host_busy) {
+ SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler"
+ " scsi_eh_%d"
+ " sleeping\n",
+ shost->host_no));
+ schedule();
+ set_current_state(TASK_INTERRUPTIBLE);
+ continue;
+ }
+ __set_current_state(TASK_RUNNING);
SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler"
" scsi_eh_%d waking"
" up\n",shost->host_no));
* which are still online.
*/
scsi_restart_operations(shost);
-
+ set_current_state(TASK_INTERRUPTIBLE);
}
+ __set_current_state(TASK_RUNNING);
+
SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler scsi_eh_%d"
" exiting\n",shost->host_no));
/*
* Make sure that nobody tries to wake us up again.
*/
- shost->eh_wait = NULL;
-
- /*
- * Knock this down too. From this point on, the host is flying
- * without a pilot. If this is because the module is being unloaded,
- * that's fine. If the user sent a signal to this thing, we are
- * potentially in real danger.
- */
- shost->eh_active = 0;
shost->ehandler = NULL;
-
- /*
- * If anyone is waiting for us to exit (i.e. someone trying to unload
- * a driver), then wake up that process to let them know we are on
- * the way out the door.
- */
- complete_and_exit(shost->eh_notify, 0);
return 0;
}
scmd->request = &req;
memset(&scmd->eh_timeout, 0, sizeof(scmd->eh_timeout));
scmd->request->rq_status = RQ_SCSI_BUSY;
- scmd->state = SCSI_STATE_INITIALIZING;
- scmd->owner = SCSI_OWNER_MIDLEVEL;
-
+
memset(&scmd->cmnd, '\0', sizeof(scmd->cmnd));
scmd->scsi_done = scsi_reset_provider_done_command;
scmd->bufflen = 0;
scmd->request_buffer = NULL;
scmd->request_bufflen = 0;
- scmd->abort_reason = DID_ABORT;
scmd->cmd_len = 0;
int scsi_normalize_sense(const u8 *sense_buffer, int sb_len,
struct scsi_sense_hdr *sshdr)
{
- if (!sense_buffer || !sb_len || (sense_buffer[0] & 0x70) != 0x70)
+ if (!sense_buffer || !sb_len)
return 0;
memset(sshdr, 0, sizeof(struct scsi_sense_hdr));
sshdr->response_code = (sense_buffer[0] & 0x7f);
+
+ if (!scsi_sense_valid(sshdr))
+ return 0;
+
if (sshdr->response_code >= 0x72) {
/*
* descriptor format