scsi_error.c 50.3 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
/*
 *  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>
23
#include <linux/kthread.h>
L
Linus Torvalds 已提交
24 25 26 27 28
#include <linux/interrupt.h>
#include <linux/blkdev.h>
#include <linux/delay.h>

#include <scsi/scsi.h>
29
#include <scsi/scsi_cmnd.h>
L
Linus Torvalds 已提交
30 31 32
#include <scsi/scsi_dbg.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_eh.h>
33
#include <scsi/scsi_transport.h>
L
Linus Torvalds 已提交
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
#include <scsi/scsi_host.h>
#include <scsi/scsi_ioctl.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) {
54
		wake_up_process(shost->ehandler);
L
Linus Torvalds 已提交
55 56 57 58
		SCSI_LOG_ERROR_RECOVERY(5,
				printk("Waking error handler thread\n"));
	}
}
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

/**
 * scsi_schedule_eh - schedule EH for SCSI host
 * @shost:	SCSI host to invoke error handling on.
 *
 * Schedule SCSI EH without scmd.
 **/
void scsi_schedule_eh(struct Scsi_Host *shost)
{
	unsigned long flags;

	spin_lock_irqsave(shost->host_lock, flags);

	if (scsi_host_set_state(shost, SHOST_RECOVERY) == 0 ||
	    scsi_host_set_state(shost, SHOST_CANCEL_RECOVERY) == 0) {
		shost->host_eh_scheduled++;
		scsi_eh_wakeup(shost);
	}

	spin_unlock_irqrestore(shost->host_lock, flags);
}
EXPORT_SYMBOL_GPL(scsi_schedule_eh);
L
Linus Torvalds 已提交
81 82 83 84 85 86 87 88 89 90 91 92 93

/**
 * 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;
94
	int ret = 0;
L
Linus Torvalds 已提交
95

96
	if (!shost->ehandler)
L
Linus Torvalds 已提交
97 98 99
		return 0;

	spin_lock_irqsave(shost->host_lock, flags);
100 101 102
	if (scsi_host_set_state(shost, SHOST_RECOVERY))
		if (scsi_host_set_state(shost, SHOST_CANCEL_RECOVERY))
			goto out_unlock;
L
Linus Torvalds 已提交
103

104
	ret = 1;
105
	scmd->eh_eflags |= eh_flag;
L
Linus Torvalds 已提交
106 107 108
	list_add_tail(&scmd->eh_entry, &shost->eh_cmd_q);
	shost->host_failed++;
	scsi_eh_wakeup(shost);
109
 out_unlock:
L
Linus Torvalds 已提交
110
	spin_unlock_irqrestore(shost->host_lock, flags);
111
	return ret;
L
Linus Torvalds 已提交
112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186
}

/**
 * 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);
}

/**
 * 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;
}

/**
 * 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)
{
187 188
	enum scsi_eh_timer_return (* eh_timed_out)(struct scsi_cmnd *);

L
Linus Torvalds 已提交
189 190
	scsi_log_completion(scmd, TIMEOUT_ERROR);

191
	if (scmd->device->host->transportt->eh_timed_out)
192 193 194 195 196 197 198 199
		eh_timed_out = scmd->device->host->transportt->eh_timed_out;
	else if (scmd->device->host->hostt->eh_timed_out)
		eh_timed_out = scmd->device->host->hostt->eh_timed_out;
	else
		eh_timed_out = NULL;

	if (eh_timed_out)
		switch (eh_timed_out(scmd)) {
L
Linus Torvalds 已提交
200 201 202 203 204 205 206 207 208 209 210 211
		case EH_HANDLED:
			__scsi_done(scmd);
			return;
		case EH_RESET_TIMER:
			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))) {
212 213
		scmd->result |= DID_TIME_OUT << 16;
		__scsi_done(scmd);
L
Linus Torvalds 已提交
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231
	}
}

/**
 * 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;

232
	wait_event(sdev->host->host_wait, !scsi_host_in_recovery(sdev->host));
L
Linus Torvalds 已提交
233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262

	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;
263
				if (scmd->eh_eflags & SCSI_EH_CANCEL_CMD)
L
Linus Torvalds 已提交
264 265 266 267 268 269 270 271
					++cmd_cancel;
				else 
					++cmd_failed;
			}
		}

		if (cmd_cancel || cmd_failed) {
			SCSI_LOG_ERROR_RECOVERY(3,
272 273 274 275
				sdev_printk(KERN_INFO, sdev,
					    "%s: cmds failed: %d, cancel: %d\n",
					    __FUNCTION__, cmd_failed,
					    cmd_cancel));
L
Linus Torvalds 已提交
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
			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:
371 372 373 374 375
		if (sshdr.asc == 0x11 || /* UNRECOVERED READ ERR */
		    sshdr.asc == 0x13 || /* AMNF DATA FIELD */
		    sshdr.asc == 0x14) { /* RECORD NOT FOUND */
			return SUCCESS;
		}
L
Linus Torvalds 已提交
376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457
		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_done - Completion function for error handling.
 * @scmd:	Cmd that is done.
 **/
static void scsi_eh_done(struct scsi_cmnd *scmd)
{
458 459
	struct completion     *eh_action;

460 461 462
	SCSI_LOG_ERROR_RECOVERY(3,
		printk("%s scmd: %p result: %x\n",
			__FUNCTION__, scmd, scmd->result));
463 464 465 466

	eh_action = scmd->device->host->eh_action;
	if (eh_action)
		complete(eh_action);
L
Linus Torvalds 已提交
467 468
}

469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
/**
 * 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__));

	if (!scmd->device->host->hostt->eh_host_reset_handler)
		return FAILED;

	rtn = scmd->device->host->hostt->eh_host_reset_handler(scmd);

	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_channel(scmd));
		spin_unlock_irqrestore(scmd->device->host->host_lock, flags);
	}

	return rtn;
}

/**
 * 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__));

	if (!scmd->device->host->hostt->eh_bus_reset_handler)
		return FAILED;

	rtn = scmd->device->host->hostt->eh_bus_reset_handler(scmd);

	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_channel(scmd));
		spin_unlock_irqrestore(scmd->device->host->host_lock, flags);
	}

	return rtn;
}

/**
 * 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)
{
	int rtn;

	if (!scmd->device->host->hostt->eh_device_reset_handler)
		return FAILED;

	rtn = scmd->device->host->hostt->eh_device_reset_handler(scmd);
	if (rtn == SUCCESS) {
		scmd->device->was_reset = 1;
		scmd->device->expecting_cc_ua = 1;
	}

	return rtn;
}

static int __scsi_try_to_abort_cmd(struct scsi_cmnd *scmd)
{
	if (!scmd->device->host->hostt->eh_abort_handler)
		return FAILED;

	return scmd->device->host->hostt->eh_abort_handler(scmd);
}

/**
 * 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)
{
	/*
	 * 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;
	return __scsi_try_to_abort_cmd(scmd);
}

static void scsi_abort_eh_cmnd(struct scsi_cmnd *scmd)
{
	if (__scsi_try_to_abort_cmd(scmd) != SUCCESS)
		if (scsi_try_bus_device_reset(scmd) != SUCCESS)
			if (scsi_try_bus_reset(scmd) != SUCCESS)
				scsi_try_host_reset(scmd);
}

L
Linus Torvalds 已提交
591
/**
592 593 594 595 596 597 598 599 600 601 602 603
 * scsi_send_eh_cmnd  - submit a scsi command as part of error recory
 * @scmd:       SCSI command structure to hijack
 * @cmnd:       CDB to send
 * @cmnd_size:  size in bytes of @cmnd
 * @timeout:    timeout for this request
 * @copy_sense: request sense data if set to 1
 *
 * This function is used to send a scsi command down to a target device
 * as part of the error recovery process.  If @copy_sense is 0 the command
 * sent must be one that does not transfer any data.  If @copy_sense is 1
 * the command must be REQUEST_SENSE and this functions copies out the
 * sense buffer it got into @scmd->sense_buffer.
L
Linus Torvalds 已提交
604 605 606 607
 *
 * Return value:
 *    SUCCESS or FAILED or NEEDS_RETRY
 **/
608 609
static int scsi_send_eh_cmnd(struct scsi_cmnd *scmd, unsigned char *cmnd,
			     int cmnd_size, int timeout, int copy_sense)
L
Linus Torvalds 已提交
610
{
已提交
611 612
	struct scsi_device *sdev = scmd->device;
	struct Scsi_Host *shost = sdev->host;
613
	int old_result = scmd->result;
614
	DECLARE_COMPLETION_ONSTACK(done);
615
	unsigned long timeleft;
L
Linus Torvalds 已提交
616
	unsigned long flags;
617
	struct scatterlist sgl;
618 619 620 621 622 623
	unsigned char old_cmnd[MAX_COMMAND_SIZE];
	enum dma_data_direction old_data_direction;
	unsigned short old_use_sg;
	unsigned char old_cmd_len;
	unsigned old_bufflen;
	void *old_buffer;
624
	int rtn;
L
Linus Torvalds 已提交
625

626 627 628 629 630 631 632 633 634 635 636 637 638 639
	/*
	 * We need saved copies of a number of fields - this is because
	 * error handling may need to overwrite these with different values
	 * to run different commands, and once error handling is complete,
	 * we will need to restore these values prior to running the actual
	 * command.
	 */
	old_buffer = scmd->request_buffer;
	old_bufflen = scmd->request_bufflen;
	memcpy(old_cmnd, scmd->cmnd, sizeof(scmd->cmnd));
	old_data_direction = scmd->sc_data_direction;
	old_cmd_len = scmd->cmd_len;
	old_use_sg = scmd->use_sg;

640 641 642
	memset(scmd->cmnd, 0, sizeof(scmd->cmnd));
	memcpy(scmd->cmnd, cmnd, cmnd_size);

643
	if (copy_sense) {
A
Al Viro 已提交
644
		gfp_t gfp_mask = GFP_ATOMIC;
645 646 647 648

		if (shost->hostt->unchecked_isa_dma)
			gfp_mask |= __GFP_DMA;

649 650
		sgl.page = alloc_page(gfp_mask);
		if (!sgl.page)
651
			return FAILED;
652 653 654 655 656 657 658
		sgl.offset = 0;
		sgl.length = 252;

		scmd->sc_data_direction = DMA_FROM_DEVICE;
		scmd->request_bufflen = sgl.length;
		scmd->request_buffer = &sgl;
		scmd->use_sg = 1;
659 660 661 662
	} else {
		scmd->request_buffer = NULL;
		scmd->request_bufflen = 0;
		scmd->sc_data_direction = DMA_NONE;
663
		scmd->use_sg = 0;
664 665 666 667 668
	}

	scmd->underflow = 0;
	scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]);

已提交
669
	if (sdev->scsi_level <= SCSI_2)
L
Linus Torvalds 已提交
670
		scmd->cmnd[1] = (scmd->cmnd[1] & 0x1f) |
已提交
671
			(sdev->lun << 5 & 0xe0);
L
Linus Torvalds 已提交
672

673 674 675 676 677 678
	/*
	 * 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));

679
	shost->eh_action = &done;
L
Linus Torvalds 已提交
680

已提交
681
	spin_lock_irqsave(shost->host_lock, flags);
L
Linus Torvalds 已提交
682
	scsi_log_send(scmd);
已提交
683 684
	shost->hostt->queuecommand(scmd, scsi_eh_done);
	spin_unlock_irqrestore(shost->host_lock, flags);
L
Linus Torvalds 已提交
685

686
	timeleft = wait_for_completion_timeout(&done, timeout);
L
Linus Torvalds 已提交
687

已提交
688
	shost->eh_action = NULL;
L
Linus Torvalds 已提交
689

690
	scsi_log_completion(scmd, SUCCESS);
L
Linus Torvalds 已提交
691

692 693 694
	SCSI_LOG_ERROR_RECOVERY(3,
		printk("%s: scmd: %p, timeleft: %ld\n",
			__FUNCTION__, scmd, timeleft));
L
Linus Torvalds 已提交
695 696

	/*
697 698 699 700
	 * If there is time left scsi_eh_done got called, and we will
	 * examine the actual status codes to see whether the command
	 * actually did complete normally, else tell the host to forget
	 * about this command.
L
Linus Torvalds 已提交
701
	 */
702
	if (timeleft) {
L
Linus Torvalds 已提交
703 704 705 706
		rtn = scsi_eh_completed_normally(scmd);
		SCSI_LOG_ERROR_RECOVERY(3,
			printk("%s: scsi_eh_completed_normally %x\n",
			       __FUNCTION__, rtn));
707

L
Linus Torvalds 已提交
708 709 710 711 712 713 714 715 716
		switch (rtn) {
		case SUCCESS:
		case NEEDS_RETRY:
		case FAILED:
			break;
		default:
			rtn = FAILED;
			break;
		}
717
	} else {
718
		scsi_abort_eh_cmnd(scmd);
719
		rtn = FAILED;
L
Linus Torvalds 已提交
720 721
	}

722 723 724 725 726 727 728 729 730

	/*
	 * Last chance to have valid sense data.
	 */
	if (copy_sense) {
		if (!SCSI_SENSE_VALID(scmd)) {
			memcpy(scmd->sense_buffer, scmd->request_buffer,
			       sizeof(scmd->sense_buffer));
		}
731
		__free_page(sgl.page);
732 733 734 735 736 737 738 739 740 741 742 743 744
	}


	/*
	 * Restore original data
	 */
	scmd->request_buffer = old_buffer;
	scmd->request_bufflen = old_bufflen;
	memcpy(scmd->cmnd, old_cmnd, sizeof(scmd->cmnd));
	scmd->sc_data_direction = old_data_direction;
	scmd->cmd_len = old_cmd_len;
	scmd->use_sg = old_use_sg;
	scmd->result = old_result;
L
Linus Torvalds 已提交
745 746 747 748 749 750 751 752 753 754 755 756 757 758 759
	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] =
760
		{REQUEST_SENSE, 0, 0, 0, 252, 0};
L
Linus Torvalds 已提交
761

762
	return scsi_send_eh_cmnd(scmd, generic_sense, 6, SENSE_TIMEOUT, 1);
L
Linus Torvalds 已提交
763 764 765 766 767 768 769 770 771 772 773 774 775 776
}

/**
 * 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.
 **/
777
void scsi_eh_finish_cmd(struct scsi_cmnd *scmd, struct list_head *done_q)
L
Linus Torvalds 已提交
778 779
{
	scmd->device->host->host_failed--;
780
	scmd->eh_eflags = 0;
L
Linus Torvalds 已提交
781 782
	list_move_tail(&scmd->eh_entry, done_q);
}
783
EXPORT_SYMBOL(scsi_eh_finish_cmd);
L
Linus Torvalds 已提交
784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804

/**
 * 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.
 **/
805 806
int scsi_eh_get_sense(struct list_head *work_q,
		      struct list_head *done_q)
L
Linus Torvalds 已提交
807
{
808
	struct scsi_cmnd *scmd, *next;
L
Linus Torvalds 已提交
809 810
	int rtn;

811
	list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
812
		if ((scmd->eh_eflags & SCSI_EH_CANCEL_CMD) ||
L
Linus Torvalds 已提交
813 814 815
		    SCSI_SENSE_VALID(scmd))
			continue;

816 817 818
		SCSI_LOG_ERROR_RECOVERY(2, scmd_printk(KERN_INFO, scmd,
						  "%s: requesting sense\n",
						  current->comm));
L
Linus Torvalds 已提交
819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847
		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);
}
848
EXPORT_SYMBOL_GPL(scsi_eh_get_sense);
L
Linus Torvalds 已提交
849 850 851 852 853 854 855 856 857 858 859 860 861 862

/**
 * 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:
863
	rtn = scsi_send_eh_cmnd(scmd, tur_command, 6, SENSE_TIMEOUT, 0);
L
Linus Torvalds 已提交
864 865 866

	SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd %p rtn %x\n",
		__FUNCTION__, scmd, rtn));
867 868 869

	switch (rtn) {
	case NEEDS_RETRY:
L
Linus Torvalds 已提交
870 871
		if (retry_cnt--)
			goto retry_tur;
872 873
		/*FALLTHRU*/
	case SUCCESS:
874
		return 0;
875 876
	default:
		return 1;
877
	}
L
Linus Torvalds 已提交
878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
}

/**
 * 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)
{
895
	struct scsi_cmnd *scmd, *next;
L
Linus Torvalds 已提交
896 897
	int rtn;

898
	list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
899
		if (!(scmd->eh_eflags & SCSI_EH_CANCEL_CMD))
L
Linus Torvalds 已提交
900 901 902 903 904 905
			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) {
906
			scmd->eh_eflags &= ~SCSI_EH_CANCEL_CMD;
L
Linus Torvalds 已提交
907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933
			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_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};

934 935
	if (scmd->device->allow_restart) {
		int rtn;
L
Linus Torvalds 已提交
936

937 938
		rtn = scsi_send_eh_cmnd(scmd, stu_command, 6,
					START_UNIT_TIMEOUT, 0);
939 940 941
		if (rtn == SUCCESS)
			return 0;
	}
L
Linus Torvalds 已提交
942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958

	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)
{
959
	struct scsi_cmnd *scmd, *stu_scmd, *next;
L
Linus Torvalds 已提交
960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979
	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)) {
980 981
				list_for_each_entry_safe(scmd, next,
							  work_q, eh_entry) {
L
Linus Torvalds 已提交
982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011
					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)
{
1012
	struct scsi_cmnd *scmd, *bdr_scmd, *next;
L
Linus Torvalds 已提交
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
	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)) {
1034 1035
				list_for_each_entry_safe(scmd, next,
							 work_q, eh_entry) {
L
Linus Torvalds 已提交
1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
					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_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)
{
1062
	struct scsi_cmnd *scmd, *chan_scmd, *next;
L
Linus Torvalds 已提交
1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
	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) {
1076
			if (channel == scmd_channel(scmd)) {
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
				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) {
1093
			list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
1094
				if (channel == scmd_channel(scmd))
L
Linus Torvalds 已提交
1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117
					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,
							  channel));
		}
	}
	return list_empty(work_q);
}

/**
 * scsi_eh_host_reset - send a host reset 
 * @work_q:	list_head for processed commands.
 * @done_q:	list_head for processed commands.
 **/
static int scsi_eh_host_reset(struct list_head *work_q,
			      struct list_head *done_q)
{
1118
	struct scsi_cmnd *scmd, *next;
L
Linus Torvalds 已提交
1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129
	int rtn;

	if (!list_empty(work_q)) {
		scmd = list_entry(work_q->next,
				  struct scsi_cmnd, eh_entry);

		SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Sending HRST\n"
						  , current->comm));

		rtn = scsi_try_host_reset(scmd);
		if (rtn == SUCCESS) {
1130
			list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
L
Linus Torvalds 已提交
1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153
				if (!scsi_device_online(scmd->device) ||
				    (!scsi_eh_try_stu(scmd) && !scsi_eh_tur(scmd)) ||
				    !scsi_eh_tur(scmd))
					scsi_eh_finish_cmd(scmd, done_q);
			}
		} else {
			SCSI_LOG_ERROR_RECOVERY(3, printk("%s: HRST"
							  " failed\n",
							  current->comm));
		}
	}
	return list_empty(work_q);
}

/**
 * scsi_eh_offline_sdevs - offline scsi devices that fail to recover
 * @work_q:	list_head for processed commands.
 * @done_q:	list_head for processed commands.
 *
 **/
static void scsi_eh_offline_sdevs(struct list_head *work_q,
				  struct list_head *done_q)
{
1154
	struct scsi_cmnd *scmd, *next;
L
Linus Torvalds 已提交
1155

1156
	list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
1157 1158 1159
		sdev_printk(KERN_INFO, scmd->device,
			    "scsi: Device offlined - not"
			    " ready after error recovery\n");
L
Linus Torvalds 已提交
1160
		scsi_device_set_state(scmd->device, SDEV_OFFLINE);
1161
		if (scmd->eh_eflags & SCSI_EH_CANCEL_CMD) {
L
Linus Torvalds 已提交
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
			/*
			 * FIXME: Handle lost cmds.
			 */
		}
		scsi_eh_finish_cmd(scmd, done_q);
	}
	return;
}

/**
 * scsi_decide_disposition - Disposition a cmd on return from LLD.
 * @scmd:	SCSI cmd to examine.
 *
 * Notes:
 *    This is *only* called when we are examining the status after sending
 *    out the actual data command.  any commands that are queued for error
 *    recovery (e.g. test_unit_ready) do *not* come through here.
 *
 *    When this routine returns failed, it means the error handler thread
 *    is woken.  In cases where the error code indicates an error that
 *    doesn't require the error handler read (i.e. we don't need to
 *    abort/reset), this function should return SUCCESS.
 **/
int scsi_decide_disposition(struct scsi_cmnd *scmd)
{
	int rtn;

	/*
	 * if the device is offline, then we clearly just pass the result back
	 * up to the top level.
	 */
	if (!scsi_device_online(scmd->device)) {
		SCSI_LOG_ERROR_RECOVERY(5, printk("%s: device offline - report"
						  " as SUCCESS\n",
						  __FUNCTION__));
		return SUCCESS;
	}

	/*
	 * first check the host byte, to see if there is anything in there
	 * that would indicate what we need to do.
	 */
	switch (host_byte(scmd->result)) {
	case DID_PASSTHROUGH:
		/*
		 * no matter what, pass this through to the upper layer.
		 * nuke this special code so that it looks like we are saying
		 * did_ok.
		 */
		scmd->result &= 0xff00ffff;
		return SUCCESS;
	case DID_OK:
		/*
		 * looks good.  drop through, and check the next byte.
		 */
		break;
	case DID_NO_CONNECT:
	case DID_BAD_TARGET:
	case DID_ABORT:
		/*
		 * note - this means that we just report the status back
		 * to the top level driver, not that we actually think
		 * that it indicates SUCCESS.
		 */
		return SUCCESS;
		/*
		 * when the low level driver returns did_soft_error,
		 * it is responsible for keeping an internal retry counter 
		 * in order to avoid endless loops (db)
		 *
		 * actually this is a bug in this function here.  we should
		 * be mindful of the maximum number of retries specified
		 * and not get stuck in a loop.
		 */
	case DID_SOFT_ERROR:
		goto maybe_retry;
	case DID_IMM_RETRY:
		return NEEDS_RETRY;

已提交
1241 1242 1243
	case DID_REQUEUE:
		return ADD_TO_MLQUEUE;

L
Linus Torvalds 已提交
1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
	case DID_ERROR:
		if (msg_byte(scmd->result) == COMMAND_COMPLETE &&
		    status_byte(scmd->result) == RESERVATION_CONFLICT)
			/*
			 * execute reservation conflict processing code
			 * lower down
			 */
			break;
		/* fallthrough */

	case DID_BUS_BUSY:
	case DID_PARITY:
		goto maybe_retry;
	case DID_TIME_OUT:
		/*
		 * when we scan the bus, we get timeout messages for
		 * these commands if there is no device available.
		 * other hosts report did_no_connect for the same thing.
		 */
		if ((scmd->cmnd[0] == TEST_UNIT_READY ||
		     scmd->cmnd[0] == INQUIRY)) {
			return SUCCESS;
		} else {
			return FAILED;
		}
	case DID_RESET:
		return SUCCESS;
	default:
		return FAILED;
	}

	/*
	 * next, check the message byte.
	 */
	if (msg_byte(scmd->result) != COMMAND_COMPLETE)
		return FAILED;

	/*
	 * check the status byte to see if this indicates anything special.
	 */
	switch (status_byte(scmd->result)) {
	case QUEUE_FULL:
		/*
		 * the case of trying to send too many commands to a
		 * tagged queueing device.
		 */
	case BUSY:
		/*
		 * device can't talk to us at the moment.  Should only
		 * occur (SAM-3) when the task queue is empty, so will cause
		 * the empty queue handling to trigger a stall in the
		 * device.
		 */
		return ADD_TO_MLQUEUE;
	case GOOD:
	case COMMAND_TERMINATED:
	case TASK_ABORTED:
		return SUCCESS;
	case CHECK_CONDITION:
		rtn = scsi_check_sense(scmd);
		if (rtn == NEEDS_RETRY)
			goto maybe_retry;
		/* if rtn == FAILED, we have no sense information;
		 * returning FAILED will wake the error handler thread
		 * to collect the sense and redo the decide
		 * disposition */
		return rtn;
	case CONDITION_GOOD:
	case INTERMEDIATE_GOOD:
	case INTERMEDIATE_C_GOOD:
	case ACA_ACTIVE:
		/*
		 * who knows?  FIXME(eric)
		 */
		return SUCCESS;

	case RESERVATION_CONFLICT:
1321 1322
		sdev_printk(KERN_INFO, scmd->device,
			    "reservation conflict\n");
L
Linus Torvalds 已提交
1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334
		return SUCCESS; /* causes immediate i/o error */
	default:
		return FAILED;
	}
	return FAILED;

      maybe_retry:

	/* we requeue for retry because the error was retryable, and
	 * the request was not marked fast fail.  Note that above,
	 * even if the request is marked fast fail, we still requeue
	 * for queue congestion conditions (QUEUE_FULL or BUSY) */
1335
	if ((++scmd->retries) <= scmd->allowed
L
Linus Torvalds 已提交
1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
	    && !blk_noretry_request(scmd->request)) {
		return NEEDS_RETRY;
	} else {
		/*
		 * no more retries - report this one back to upper level.
		 */
		return SUCCESS;
	}
}

/**
 * scsi_eh_lock_door - Prevent medium removal for the specified device
 * @sdev:	SCSI device to prevent medium removal
 *
 * Locking:
 * 	We must be called from process context; scsi_allocate_request()
 * 	may sleep.
 *
 * Notes:
 * 	We queue up an asynchronous "ALLOW MEDIUM REMOVAL" request on the
 * 	head of the devices request queue, and continue.
 *
 * Bugs:
 * 	scsi_allocate_request() may sleep waiting for existing requests to
 * 	be processed.  However, since we haven't kicked off any request
 * 	processing for this host, this may deadlock.
 *
 *	If scsi_allocate_request() fails for what ever reason, we
 *	completely forget to lock the door.
 **/
static void scsi_eh_lock_door(struct scsi_device *sdev)
{
1368
	unsigned char cmnd[MAX_COMMAND_SIZE];
L
Linus Torvalds 已提交
1369

1370 1371 1372 1373 1374 1375
	cmnd[0] = ALLOW_MEDIUM_REMOVAL;
	cmnd[1] = 0;
	cmnd[2] = 0;
	cmnd[3] = 0;
	cmnd[4] = SCSI_REMOVAL_PREVENT;
	cmnd[5] = 0;
L
Linus Torvalds 已提交
1376

1377
	scsi_execute_async(sdev, cmnd, 6, DMA_NONE, NULL, 0, 0, 10 * HZ,
1378
			   5, NULL, NULL, GFP_KERNEL);
L
Linus Torvalds 已提交
1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
}


/**
 * scsi_restart_operations - restart io operations to the specified host.
 * @shost:	Host we are restarting.
 *
 * Notes:
 *    When we entered the error handler, we blocked all further i/o to
 *    this device.  we need to 'reverse' this process.
 **/
static void scsi_restart_operations(struct Scsi_Host *shost)
{
	struct scsi_device *sdev;
1393
	unsigned long flags;
L
Linus Torvalds 已提交
1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412

	/*
	 * If the door was locked, we need to insert a door lock request
	 * onto the head of the SCSI request queue for the device.  There
	 * is no point trying to lock the door of an off-line device.
	 */
	shost_for_each_device(sdev, shost) {
		if (scsi_device_online(sdev) && sdev->locked)
			scsi_eh_lock_door(sdev);
	}

	/*
	 * next free up anything directly waiting upon the host.  this
	 * will be requests for character device operations, and also for
	 * ioctls to queued block devices.
	 */
	SCSI_LOG_ERROR_RECOVERY(3, printk("%s: waking up host to restart\n",
					  __FUNCTION__));

1413 1414 1415 1416 1417
	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);
L
Linus Torvalds 已提交
1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435

	wake_up(&shost->host_wait);

	/*
	 * finally we need to re-initiate requests that may be pending.  we will
	 * have had everything blocked while error handling is taking place, and
	 * now that error recovery is done, we will need to ensure that these
	 * requests are started.
	 */
	scsi_run_host_queues(shost);
}

/**
 * scsi_eh_ready_devs - check device ready state and recover if not.
 * @shost: 	host to be recovered.
 * @eh_done_q:	list_head for processed commands.
 *
 **/
1436 1437 1438
void scsi_eh_ready_devs(struct Scsi_Host *shost,
			struct list_head *work_q,
			struct list_head *done_q)
L
Linus Torvalds 已提交
1439 1440 1441 1442 1443 1444 1445
{
	if (!scsi_eh_stu(shost, work_q, done_q))
		if (!scsi_eh_bus_device_reset(shost, work_q, done_q))
			if (!scsi_eh_bus_reset(shost, work_q, done_q))
				if (!scsi_eh_host_reset(work_q, done_q))
					scsi_eh_offline_sdevs(work_q, done_q);
}
1446
EXPORT_SYMBOL_GPL(scsi_eh_ready_devs);
L
Linus Torvalds 已提交
1447 1448 1449 1450 1451 1452

/**
 * scsi_eh_flush_done_q - finish processed commands or retry them.
 * @done_q:	list_head of processed commands.
 *
 **/
1453
void scsi_eh_flush_done_q(struct list_head *done_q)
L
Linus Torvalds 已提交
1454
{
1455
	struct scsi_cmnd *scmd, *next;
L
Linus Torvalds 已提交
1456

1457 1458
	list_for_each_entry_safe(scmd, next, done_q, eh_entry) {
		list_del_init(&scmd->eh_entry);
L
Linus Torvalds 已提交
1459 1460
		if (scsi_device_online(scmd->device) &&
		    !blk_noretry_request(scmd->request) &&
1461
		    (++scmd->retries <= scmd->allowed)) {
L
Linus Torvalds 已提交
1462 1463 1464 1465 1466 1467
			SCSI_LOG_ERROR_RECOVERY(3, printk("%s: flush"
							  " retry cmd: %p\n",
							  current->comm,
							  scmd));
				scsi_queue_insert(scmd, SCSI_MLQUEUE_EH_RETRY);
		} else {
1468 1469 1470 1471 1472
			/*
			 * If just we got sense for the device (called
			 * scsi_eh_get_sense), scmd->result is already
			 * set, do not set DRIVER_TIMEOUT.
			 */
L
Linus Torvalds 已提交
1473 1474 1475 1476 1477 1478 1479 1480 1481
			if (!scmd->result)
				scmd->result |= (DRIVER_TIMEOUT << 24);
			SCSI_LOG_ERROR_RECOVERY(3, printk("%s: flush finish"
							" cmd: %p\n",
							current->comm, scmd));
			scsi_finish_command(scmd);
		}
	}
}
1482
EXPORT_SYMBOL(scsi_eh_flush_done_q);
L
Linus Torvalds 已提交
1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526

/**
 * scsi_unjam_host - Attempt to fix a host which has a cmd that failed.
 * @shost:	Host to unjam.
 *
 * Notes:
 *    When we come in here, we *know* that all commands on the bus have
 *    either completed, failed or timed out.  we also know that no further
 *    commands are being sent to the host, so things are relatively quiet
 *    and we have freedom to fiddle with things as we wish.
 *
 *    This is only the *default* implementation.  it is possible for
 *    individual drivers to supply their own version of this function, and
 *    if the maintainer wishes to do this, it is strongly suggested that
 *    this function be taken as a template and modified.  this function
 *    was designed to correctly handle problems for about 95% of the
 *    different cases out there, and it should always provide at least a
 *    reasonable amount of error recovery.
 *
 *    Any command marked 'failed' or 'timeout' must eventually have
 *    scsi_finish_cmd() called for it.  we do all of the retry stuff
 *    here, so when we restart the host after we return it should have an
 *    empty queue.
 **/
static void scsi_unjam_host(struct Scsi_Host *shost)
{
	unsigned long flags;
	LIST_HEAD(eh_work_q);
	LIST_HEAD(eh_done_q);

	spin_lock_irqsave(shost->host_lock, flags);
	list_splice_init(&shost->eh_cmd_q, &eh_work_q);
	spin_unlock_irqrestore(shost->host_lock, flags);

	SCSI_LOG_ERROR_RECOVERY(1, scsi_eh_prt_fail_stats(shost, &eh_work_q));

	if (!scsi_eh_get_sense(&eh_work_q, &eh_done_q))
		if (!scsi_eh_abort_cmds(&eh_work_q, &eh_done_q))
			scsi_eh_ready_devs(shost, &eh_work_q, &eh_done_q);

	scsi_eh_flush_done_q(&eh_done_q);
}

/**
1527
 * scsi_error_handler - SCSI error handler thread
L
Linus Torvalds 已提交
1528 1529 1530
 * @data:	Host for which we are running.
 *
 * Notes:
1531 1532
 *    This is the main error handling loop.  This is run as a kernel thread
 *    for every SCSI host and handles all error handling activity.
L
Linus Torvalds 已提交
1533 1534 1535
 **/
int scsi_error_handler(void *data)
{
1536
	struct Scsi_Host *shost = data;
L
Linus Torvalds 已提交
1537 1538 1539 1540

	current->flags |= PF_NOFREEZE;

	/*
1541 1542 1543 1544
	 * We use TASK_INTERRUPTIBLE so that the thread is not
	 * counted against the load average as a running process.
	 * We never actually get interrupted because kthread_run
	 * disables singal delivery for the created thread.
L
Linus Torvalds 已提交
1545
	 */
1546 1547
	set_current_state(TASK_INTERRUPTIBLE);
	while (!kthread_should_stop()) {
1548
		if ((shost->host_failed == 0 && shost->host_eh_scheduled == 0) ||
1549
		    shost->host_failed != shost->host_busy) {
1550 1551 1552
			SCSI_LOG_ERROR_RECOVERY(1,
				printk("Error handler scsi_eh_%d sleeping\n",
					shost->host_no));
1553 1554 1555 1556
			schedule();
			set_current_state(TASK_INTERRUPTIBLE);
			continue;
		}
L
Linus Torvalds 已提交
1557

1558
		__set_current_state(TASK_RUNNING);
1559 1560 1561
		SCSI_LOG_ERROR_RECOVERY(1,
			printk("Error handler scsi_eh_%d waking up\n",
				shost->host_no));
L
Linus Torvalds 已提交
1562 1563 1564 1565 1566 1567

		/*
		 * We have a host that is failing for some reason.  Figure out
		 * what we need to do to get it up and online again (if we can).
		 * If we fail, we end up taking the thing offline.
		 */
1568 1569
		if (shost->transportt->eh_strategy_handler)
			shost->transportt->eh_strategy_handler(shost);
L
Linus Torvalds 已提交
1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580
		else
			scsi_unjam_host(shost);

		/*
		 * Note - if the above fails completely, the action is to take
		 * individual devices offline and flush the queue of any
		 * outstanding requests that may have been pending.  When we
		 * restart, we restart any I/O to any other devices on the bus
		 * which are still online.
		 */
		scsi_restart_operations(shost);
1581
		set_current_state(TASK_INTERRUPTIBLE);
L
Linus Torvalds 已提交
1582
	}
1583 1584
	__set_current_state(TASK_RUNNING);

1585 1586
	SCSI_LOG_ERROR_RECOVERY(1,
		printk("Error handler scsi_eh_%d exiting\n", shost->host_no));
1587
	shost->ehandler = NULL;
L
Linus Torvalds 已提交
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
	return 0;
}

/*
 * Function:    scsi_report_bus_reset()
 *
 * Purpose:     Utility function used by low-level drivers to report that
 *		they have observed a bus reset on the bus being handled.
 *
 * Arguments:   shost       - Host in question
 *		channel     - channel on which reset was observed.
 *
 * Returns:     Nothing
 *
 * Lock status: Host lock must be held.
 *
 * Notes:       This only needs to be called if the reset is one which
 *		originates from an unknown location.  Resets originated
 *		by the mid-level itself don't need to call this, but there
 *		should be no harm.
 *
 *		The main purpose of this is to make sure that a CHECK_CONDITION
 *		is properly treated.
 */
void scsi_report_bus_reset(struct Scsi_Host *shost, int channel)
{
	struct scsi_device *sdev;

	__shost_for_each_device(sdev, shost) {
1617
		if (channel == sdev_channel(sdev)) {
L
Linus Torvalds 已提交
1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651
			sdev->was_reset = 1;
			sdev->expecting_cc_ua = 1;
		}
	}
}
EXPORT_SYMBOL(scsi_report_bus_reset);

/*
 * Function:    scsi_report_device_reset()
 *
 * Purpose:     Utility function used by low-level drivers to report that
 *		they have observed a device reset on the device being handled.
 *
 * Arguments:   shost       - Host in question
 *		channel     - channel on which reset was observed
 *		target	    - target on which reset was observed
 *
 * Returns:     Nothing
 *
 * Lock status: Host lock must be held
 *
 * Notes:       This only needs to be called if the reset is one which
 *		originates from an unknown location.  Resets originated
 *		by the mid-level itself don't need to call this, but there
 *		should be no harm.
 *
 *		The main purpose of this is to make sure that a CHECK_CONDITION
 *		is properly treated.
 */
void scsi_report_device_reset(struct Scsi_Host *shost, int channel, int target)
{
	struct scsi_device *sdev;

	__shost_for_each_device(sdev, shost) {
1652 1653
		if (channel == sdev_channel(sdev) &&
		    target == sdev_id(sdev)) {
L
Linus Torvalds 已提交
1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682
			sdev->was_reset = 1;
			sdev->expecting_cc_ua = 1;
		}
	}
}
EXPORT_SYMBOL(scsi_report_device_reset);

static void
scsi_reset_provider_done_command(struct scsi_cmnd *scmd)
{
}

/*
 * Function:	scsi_reset_provider
 *
 * Purpose:	Send requested reset to a bus or device at any phase.
 *
 * Arguments:	device	- device to send reset to
 *		flag - reset type (see scsi.h)
 *
 * Returns:	SUCCESS/FAILURE.
 *
 * Notes:	This is used by the SCSI Generic driver to provide
 *		Bus/Device reset capability.
 */
int
scsi_reset_provider(struct scsi_device *dev, int flag)
{
	struct scsi_cmnd *scmd = scsi_get_command(dev, GFP_KERNEL);
1683
	struct Scsi_Host *shost = dev->host;
L
Linus Torvalds 已提交
1684
	struct request req;
1685
	unsigned long flags;
L
Linus Torvalds 已提交
1686 1687 1688 1689
	int rtn;

	scmd->request = &req;
	memset(&scmd->eh_timeout, 0, sizeof(scmd->eh_timeout));
1690

L
Linus Torvalds 已提交
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
	memset(&scmd->cmnd, '\0', sizeof(scmd->cmnd));
    
	scmd->scsi_done		= scsi_reset_provider_done_command;
	scmd->done			= NULL;
	scmd->request_buffer		= NULL;
	scmd->request_bufflen		= 0;

	scmd->cmd_len			= 0;

	scmd->sc_data_direction		= DMA_BIDIRECTIONAL;

	init_timer(&scmd->eh_timeout);

	/*
	 * Sometimes the command can get back into the timer chain,
	 * so use the pid as an identifier.
	 */
	scmd->pid			= 0;

1710 1711 1712 1713
	spin_lock_irqsave(shost->host_lock, flags);
	shost->tmf_in_progress = 1;
	spin_unlock_irqrestore(shost->host_lock, flags);

L
Linus Torvalds 已提交
1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731
	switch (flag) {
	case SCSI_TRY_RESET_DEVICE:
		rtn = scsi_try_bus_device_reset(scmd);
		if (rtn == SUCCESS)
			break;
		/* FALLTHROUGH */
	case SCSI_TRY_RESET_BUS:
		rtn = scsi_try_bus_reset(scmd);
		if (rtn == SUCCESS)
			break;
		/* FALLTHROUGH */
	case SCSI_TRY_RESET_HOST:
		rtn = scsi_try_host_reset(scmd);
		break;
	default:
		rtn = FAILED;
	}

1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747
	spin_lock_irqsave(shost->host_lock, flags);
	shost->tmf_in_progress = 0;
	spin_unlock_irqrestore(shost->host_lock, flags);

	/*
	 * be sure to wake up anyone who was sleeping or had their queue
	 * suspended while we performed the TMF.
	 */
	SCSI_LOG_ERROR_RECOVERY(3,
		printk("%s: waking up host to restart after TMF\n",
		__FUNCTION__));

	wake_up(&shost->host_wait);

	scsi_run_host_queues(shost);

L
Linus Torvalds 已提交
1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774
	scsi_next_command(scmd);
	return rtn;
}
EXPORT_SYMBOL(scsi_reset_provider);

/**
 * scsi_normalize_sense - normalize main elements from either fixed or
 *			descriptor sense data format into a common format.
 *
 * @sense_buffer:	byte array containing sense data returned by device
 * @sb_len:		number of valid bytes in sense_buffer
 * @sshdr:		pointer to instance of structure that common
 *			elements are written to.
 *
 * Notes:
 *	The "main elements" from sense data are: response_code, sense_key,
 *	asc, ascq and additional_length (only for descriptor format).
 *
 *	Typically this function can be called after a device has
 *	responded to a SCSI command with the CHECK_CONDITION status.
 *
 * Return value:
 *	1 if valid sense data information found, else 0;
 **/
int scsi_normalize_sense(const u8 *sense_buffer, int sb_len,
                         struct scsi_sense_hdr *sshdr)
{
1775
	if (!sense_buffer || !sb_len)
L
Linus Torvalds 已提交
1776 1777 1778 1779 1780
		return 0;

	memset(sshdr, 0, sizeof(struct scsi_sense_hdr));

	sshdr->response_code = (sense_buffer[0] & 0x7f);
1781 1782 1783 1784

	if (!scsi_sense_valid(sshdr))
		return 0;

L
Linus Torvalds 已提交
1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916
	if (sshdr->response_code >= 0x72) {
		/*
		 * descriptor format
		 */
		if (sb_len > 1)
			sshdr->sense_key = (sense_buffer[1] & 0xf);
		if (sb_len > 2)
			sshdr->asc = sense_buffer[2];
		if (sb_len > 3)
			sshdr->ascq = sense_buffer[3];
		if (sb_len > 7)
			sshdr->additional_length = sense_buffer[7];
	} else {
		/* 
		 * fixed format
		 */
		if (sb_len > 2)
			sshdr->sense_key = (sense_buffer[2] & 0xf);
		if (sb_len > 7) {
			sb_len = (sb_len < (sense_buffer[7] + 8)) ?
					 sb_len : (sense_buffer[7] + 8);
			if (sb_len > 12)
				sshdr->asc = sense_buffer[12];
			if (sb_len > 13)
				sshdr->ascq = sense_buffer[13];
		}
	}

	return 1;
}
EXPORT_SYMBOL(scsi_normalize_sense);

int scsi_command_normalize_sense(struct scsi_cmnd *cmd,
				 struct scsi_sense_hdr *sshdr)
{
	return scsi_normalize_sense(cmd->sense_buffer,
			sizeof(cmd->sense_buffer), sshdr);
}
EXPORT_SYMBOL(scsi_command_normalize_sense);

/**
 * scsi_sense_desc_find - search for a given descriptor type in
 *			descriptor sense data format.
 *
 * @sense_buffer:	byte array of descriptor format sense data
 * @sb_len:		number of valid bytes in sense_buffer
 * @desc_type:		value of descriptor type to find
 *			(e.g. 0 -> information)
 *
 * Notes:
 *	only valid when sense data is in descriptor format
 *
 * Return value:
 *	pointer to start of (first) descriptor if found else NULL
 **/
const u8 * scsi_sense_desc_find(const u8 * sense_buffer, int sb_len,
				int desc_type)
{
	int add_sen_len, add_len, desc_len, k;
	const u8 * descp;

	if ((sb_len < 8) || (0 == (add_sen_len = sense_buffer[7])))
		return NULL;
	if ((sense_buffer[0] < 0x72) || (sense_buffer[0] > 0x73))
		return NULL;
	add_sen_len = (add_sen_len < (sb_len - 8)) ?
			add_sen_len : (sb_len - 8);
	descp = &sense_buffer[8];
	for (desc_len = 0, k = 0; k < add_sen_len; k += desc_len) {
		descp += desc_len;
		add_len = (k < (add_sen_len - 1)) ? descp[1]: -1;
		desc_len = add_len + 2;
		if (descp[0] == desc_type)
			return descp;
		if (add_len < 0) // short descriptor ??
			break;
	}
	return NULL;
}
EXPORT_SYMBOL(scsi_sense_desc_find);

/**
 * scsi_get_sense_info_fld - attempts to get information field from
 *			sense data (either fixed or descriptor format)
 *
 * @sense_buffer:	byte array of sense data
 * @sb_len:		number of valid bytes in sense_buffer
 * @info_out:		pointer to 64 integer where 8 or 4 byte information
 *			field will be placed if found.
 *
 * Return value:
 *	1 if information field found, 0 if not found.
 **/
int scsi_get_sense_info_fld(const u8 * sense_buffer, int sb_len,
			    u64 * info_out)
{
	int j;
	const u8 * ucp;
	u64 ull;

	if (sb_len < 7)
		return 0;
	switch (sense_buffer[0] & 0x7f) {
	case 0x70:
	case 0x71:
		if (sense_buffer[0] & 0x80) {
			*info_out = (sense_buffer[3] << 24) +
				    (sense_buffer[4] << 16) +
				    (sense_buffer[5] << 8) + sense_buffer[6];
			return 1;
		} else
			return 0;
	case 0x72:
	case 0x73:
		ucp = scsi_sense_desc_find(sense_buffer, sb_len,
					   0 /* info desc */);
		if (ucp && (0xa == ucp[1])) {
			ull = 0;
			for (j = 0; j < 8; ++j) {
				if (j > 0)
					ull <<= 8;
				ull |= ucp[4 + j];
			}
			*info_out = ull;
			return 1;
		} else
			return 0;
	default:
		return 0;
	}
}
EXPORT_SYMBOL(scsi_get_sense_info_fld);