xref: /netbsd-src/sys/dev/scsipi/sd.c (revision 4d12bfcd155352508213ace5ccc59ce930ea2974)
1 /*	$NetBSD: sd.c,v 1.301 2013/06/13 00:55:01 tls Exp $	*/
2 
3 /*-
4  * Copyright (c) 1998, 2003, 2004 The NetBSD Foundation, Inc.
5  * All rights reserved.
6  *
7  * This code is derived from software contributed to The NetBSD Foundation
8  * by Charles M. Hannum.
9  *
10  * Redistribution and use in source and binary forms, with or without
11  * modification, are permitted provided that the following conditions
12  * are met:
13  * 1. Redistributions of source code must retain the above copyright
14  *    notice, this list of conditions and the following disclaimer.
15  * 2. Redistributions in binary form must reproduce the above copyright
16  *    notice, this list of conditions and the following disclaimer in the
17  *    documentation and/or other materials provided with the distribution.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
20  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
21  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
22  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
23  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29  * POSSIBILITY OF SUCH DAMAGE.
30  */
31 
32 /*
33  * Originally written by Julian Elischer (julian@dialix.oz.au)
34  * for TRW Financial Systems for use under the MACH(2.5) operating system.
35  *
36  * TRW Financial Systems, in accordance with their agreement with Carnegie
37  * Mellon University, makes this software available to CMU to distribute
38  * or use in any manner that they see fit as long as this message is kept with
39  * the software. For this reason TFS also grants any other persons or
40  * organisations permission to use or modify this software.
41  *
42  * TFS supplies this software to be publicly redistributed
43  * on the understanding that TFS is not responsible for the correct
44  * functioning of this software in any circumstances.
45  *
46  * Ported to run under 386BSD by Julian Elischer (julian@dialix.oz.au) Sept 1992
47  */
48 
49 #include <sys/cdefs.h>
50 __KERNEL_RCSID(0, "$NetBSD: sd.c,v 1.301 2013/06/13 00:55:01 tls Exp $");
51 
52 #include "opt_scsi.h"
53 
54 #include <sys/param.h>
55 #include <sys/systm.h>
56 #include <sys/kernel.h>
57 #include <sys/file.h>
58 #include <sys/stat.h>
59 #include <sys/ioctl.h>
60 #include <sys/scsiio.h>
61 #include <sys/buf.h>
62 #include <sys/bufq.h>
63 #include <sys/uio.h>
64 #include <sys/malloc.h>
65 #include <sys/errno.h>
66 #include <sys/device.h>
67 #include <sys/disklabel.h>
68 #include <sys/disk.h>
69 #include <sys/proc.h>
70 #include <sys/conf.h>
71 #include <sys/vnode.h>
72 #include <sys/rnd.h>
73 #include <sys/cprng.h>
74 
75 #include <dev/scsipi/scsi_spc.h>
76 #include <dev/scsipi/scsipi_all.h>
77 #include <dev/scsipi/scsi_all.h>
78 #include <dev/scsipi/scsipi_disk.h>
79 #include <dev/scsipi/scsi_disk.h>
80 #include <dev/scsipi/scsiconf.h>
81 #include <dev/scsipi/scsipi_base.h>
82 #include <dev/scsipi/sdvar.h>
83 
84 #include <prop/proplib.h>
85 
86 #define	SDUNIT(dev)			DISKUNIT(dev)
87 #define	SDPART(dev)			DISKPART(dev)
88 #define	SDMINOR(unit, part)		DISKMINOR(unit, part)
89 #define	MAKESDDEV(maj, unit, part)	MAKEDISKDEV(maj, unit, part)
90 
91 #define	SDLABELDEV(dev)	(MAKESDDEV(major(dev), SDUNIT(dev), RAW_PART))
92 
93 #define	SD_DEFAULT_BLKSIZE	512
94 
95 static void	sdminphys(struct buf *);
96 static void	sdgetdefaultlabel(struct sd_softc *, struct disklabel *);
97 static int	sdgetdisklabel(struct sd_softc *);
98 static void	sdstart(struct scsipi_periph *);
99 static void	sdrestart(void *);
100 static void	sddone(struct scsipi_xfer *, int);
101 static bool	sd_suspend(device_t, const pmf_qual_t *);
102 static bool	sd_shutdown(device_t, int);
103 static int	sd_interpret_sense(struct scsipi_xfer *);
104 static int	sdlastclose(device_t);
105 
106 static int	sd_mode_sense(struct sd_softc *, u_int8_t, void *, size_t, int,
107 		    int, int *);
108 static int	sd_mode_select(struct sd_softc *, u_int8_t, void *, size_t, int,
109 		    int);
110 static int	sd_validate_blksize(struct scsipi_periph *, int);
111 static u_int64_t sd_read_capacity(struct scsipi_periph *, int *, int flags);
112 static int	sd_get_simplifiedparms(struct sd_softc *, struct disk_parms *,
113 		    int);
114 static int	sd_get_capacity(struct sd_softc *, struct disk_parms *, int);
115 static int	sd_get_parms(struct sd_softc *, struct disk_parms *, int);
116 static int	sd_get_parms_page4(struct sd_softc *, struct disk_parms *,
117 		    int);
118 static int	sd_get_parms_page5(struct sd_softc *, struct disk_parms *,
119 		    int);
120 
121 static int	sd_flush(struct sd_softc *, int);
122 static int	sd_getcache(struct sd_softc *, int *);
123 static int	sd_setcache(struct sd_softc *, int);
124 
125 static int	sdmatch(device_t, cfdata_t, void *);
126 static void	sdattach(device_t, device_t, void *);
127 static int	sddetach(device_t, int);
128 static void	sd_set_geometry(struct sd_softc *);
129 
130 CFATTACH_DECL3_NEW(sd, sizeof(struct sd_softc), sdmatch, sdattach, sddetach,
131     NULL, NULL, NULL, DVF_DETACH_SHUTDOWN);
132 
133 extern struct cfdriver sd_cd;
134 
135 static const struct scsipi_inquiry_pattern sd_patterns[] = {
136 	{T_DIRECT, T_FIXED,
137 	 "",         "",                 ""},
138 	{T_DIRECT, T_REMOV,
139 	 "",         "",                 ""},
140 	{T_OPTICAL, T_FIXED,
141 	 "",         "",                 ""},
142 	{T_OPTICAL, T_REMOV,
143 	 "",         "",                 ""},
144 	{T_SIMPLE_DIRECT, T_FIXED,
145 	 "",         "",                 ""},
146 	{T_SIMPLE_DIRECT, T_REMOV,
147 	 "",         "",                 ""},
148 };
149 
150 static dev_type_open(sdopen);
151 static dev_type_close(sdclose);
152 static dev_type_read(sdread);
153 static dev_type_write(sdwrite);
154 static dev_type_ioctl(sdioctl);
155 static dev_type_strategy(sdstrategy);
156 static dev_type_dump(sddump);
157 static dev_type_size(sdsize);
158 
159 const struct bdevsw sd_bdevsw = {
160 	sdopen, sdclose, sdstrategy, sdioctl, sddump, sdsize, D_DISK
161 };
162 
163 const struct cdevsw sd_cdevsw = {
164 	sdopen, sdclose, sdread, sdwrite, sdioctl,
165 	nostop, notty, nopoll, nommap, nokqfilter, D_DISK
166 };
167 
168 static struct dkdriver sddkdriver = { sdstrategy, sdminphys };
169 
170 static const struct scsipi_periphsw sd_switch = {
171 	sd_interpret_sense,	/* check our error handler first */
172 	sdstart,		/* have a queue, served by this */
173 	NULL,			/* have no async handler */
174 	sddone,			/* deal with stats at interrupt time */
175 };
176 
177 struct sd_mode_sense_data {
178 	/*
179 	 * XXX
180 	 * We are not going to parse this as-is -- it just has to be large
181 	 * enough.
182 	 */
183 	union {
184 		struct scsi_mode_parameter_header_6 small;
185 		struct scsi_mode_parameter_header_10 big;
186 	} header;
187 	struct scsi_general_block_descriptor blk_desc;
188 	union scsi_disk_pages pages;
189 };
190 
191 /*
192  * The routine called by the low level scsi routine when it discovers
193  * A device suitable for this driver
194  */
195 static int
196 sdmatch(device_t parent, cfdata_t match,
197     void *aux)
198 {
199 	struct scsipibus_attach_args *sa = aux;
200 	int priority;
201 
202 	(void)scsipi_inqmatch(&sa->sa_inqbuf,
203 	    sd_patterns, sizeof(sd_patterns) / sizeof(sd_patterns[0]),
204 	    sizeof(sd_patterns[0]), &priority);
205 
206 	return (priority);
207 }
208 
209 /*
210  * Attach routine common to atapi & scsi.
211  */
212 static void
213 sdattach(device_t parent, device_t self, void *aux)
214 {
215 	struct sd_softc *sd = device_private(self);
216 	struct scsipibus_attach_args *sa = aux;
217 	struct scsipi_periph *periph = sa->sa_periph;
218 	int error, result;
219 	struct disk_parms *dp = &sd->params;
220 	char pbuf[9];
221 
222 	SC_DEBUG(periph, SCSIPI_DB2, ("sdattach: "));
223 
224 	sd->sc_dev = self;
225 	sd->type = (sa->sa_inqbuf.type & SID_TYPE);
226 	strncpy(sd->name, sa->sa_inqbuf.product, sizeof(sd->name));
227 	if (sd->type == T_SIMPLE_DIRECT)
228 		periph->periph_quirks |= PQUIRK_ONLYBIG | PQUIRK_NOBIGMODESENSE;
229 
230 	if (SCSIPI_BUSTYPE_TYPE(scsipi_periph_bustype(sa->sa_periph)) ==
231 	    SCSIPI_BUSTYPE_SCSI && periph->periph_version == 0)
232 		sd->flags |= SDF_ANCIENT;
233 
234 	bufq_alloc(&sd->buf_queue, BUFQ_DISK_DEFAULT_STRAT, BUFQ_SORT_RAWBLOCK);
235 
236 	callout_init(&sd->sc_callout, 0);
237 
238 	/*
239 	 * Store information needed to contact our base driver
240 	 */
241 	sd->sc_periph = periph;
242 
243 	periph->periph_dev = sd->sc_dev;
244 	periph->periph_switch = &sd_switch;
245 
246         /*
247          * Increase our openings to the maximum-per-periph
248          * supported by the adapter.  This will either be
249          * clamped down or grown by the adapter if necessary.
250          */
251 	periph->periph_openings =
252 	    SCSIPI_CHAN_MAX_PERIPH(periph->periph_channel);
253 	periph->periph_flags |= PERIPH_GROW_OPENINGS;
254 
255 	/*
256 	 * Initialize and attach the disk structure.
257 	 */
258 	disk_init(&sd->sc_dk, device_xname(sd->sc_dev), &sddkdriver);
259 	disk_attach(&sd->sc_dk);
260 
261 	/*
262 	 * Use the subdriver to request information regarding the drive.
263 	 */
264 	aprint_naive("\n");
265 	aprint_normal("\n");
266 
267 	if (periph->periph_quirks & PQUIRK_START)
268 		(void)scsipi_start(periph, SSS_START, XS_CTL_SILENT);
269 
270 	error = scsipi_test_unit_ready(periph,
271 	    XS_CTL_DISCOVERY | XS_CTL_IGNORE_ILLEGAL_REQUEST |
272 	    XS_CTL_IGNORE_MEDIA_CHANGE | XS_CTL_SILENT_NODEV);
273 
274 	if (error)
275 		result = SDGP_RESULT_OFFLINE;
276 	else
277 		result = sd_get_parms(sd, &sd->params, XS_CTL_DISCOVERY);
278 	aprint_normal_dev(sd->sc_dev, "");
279 	switch (result) {
280 	case SDGP_RESULT_OK:
281 		format_bytes(pbuf, sizeof(pbuf),
282 		    (u_int64_t)dp->disksize * dp->blksize);
283 	        aprint_normal(
284 		"%s, %ld cyl, %ld head, %ld sec, %ld bytes/sect x %llu sectors",
285 		    pbuf, dp->cyls, dp->heads, dp->sectors, dp->blksize,
286 		    (unsigned long long)dp->disksize);
287 		break;
288 
289 	case SDGP_RESULT_OFFLINE:
290 		aprint_normal("drive offline");
291 		break;
292 
293 	case SDGP_RESULT_UNFORMATTED:
294 		aprint_normal("unformatted media");
295 		break;
296 
297 #ifdef DIAGNOSTIC
298 	default:
299 		panic("sdattach: unknown result from get_parms");
300 		break;
301 #endif
302 	}
303 	aprint_normal("\n");
304 
305 	/*
306 	 * Establish a shutdown hook so that we can ensure that
307 	 * our data has actually made it onto the platter at
308 	 * shutdown time.  Note that this relies on the fact
309 	 * that the shutdown hooks at the "leaves" of the device tree
310 	 * are run, first (thus guaranteeing that our hook runs before
311 	 * our ancestors').
312 	 */
313 	if (!pmf_device_register1(self, sd_suspend, NULL, sd_shutdown))
314 		aprint_error_dev(self, "couldn't establish power handler\n");
315 
316 	/*
317 	 * attach the device into the random source list
318 	 */
319 	rnd_attach_source(&sd->rnd_source, device_xname(sd->sc_dev),
320 			  RND_TYPE_DISK, 0);
321 
322 	/* Discover wedges on this disk. */
323 	dkwedge_discover(&sd->sc_dk);
324 
325 	/*
326 	 * Disk insertion and removal times can be a useful source
327 	 * of entropy, though the estimator should never _count_
328 	 * these bits, on insertion, because the deltas to the
329 	 * nonexistent) previous event should never allow it.
330 	 */
331 	rnd_add_uint32(&sd->rnd_source, 0);
332 }
333 
334 static int
335 sddetach(device_t self, int flags)
336 {
337 	struct sd_softc *sd = device_private(self);
338 	int s, bmaj, cmaj, i, mn, rc;
339 
340 	rnd_add_uint32(&sd->rnd_source, 0);
341 
342 	if ((rc = disk_begindetach(&sd->sc_dk, sdlastclose, self, flags)) != 0)
343 		return rc;
344 
345 	/* locate the major number */
346 	bmaj = bdevsw_lookup_major(&sd_bdevsw);
347 	cmaj = cdevsw_lookup_major(&sd_cdevsw);
348 
349 	/* Nuke the vnodes for any open instances */
350 	for (i = 0; i < MAXPARTITIONS; i++) {
351 		mn = SDMINOR(device_unit(self), i);
352 		vdevgone(bmaj, mn, mn, VBLK);
353 		vdevgone(cmaj, mn, mn, VCHR);
354 	}
355 
356 	/* kill any pending restart */
357 	callout_stop(&sd->sc_callout);
358 
359 	/* Delete all of our wedges. */
360 	dkwedge_delall(&sd->sc_dk);
361 
362 	s = splbio();
363 
364 	/* Kill off any queued buffers. */
365 	bufq_drain(sd->buf_queue);
366 
367 	bufq_free(sd->buf_queue);
368 
369 	/* Kill off any pending commands. */
370 	scsipi_kill_pending(sd->sc_periph);
371 
372 	splx(s);
373 
374 	/* Detach from the disk list. */
375 	disk_detach(&sd->sc_dk);
376 	disk_destroy(&sd->sc_dk);
377 
378 	callout_destroy(&sd->sc_callout);
379 
380 	pmf_device_deregister(self);
381 
382 	/* Unhook the entropy source. */
383 	rnd_detach_source(&sd->rnd_source);
384 
385 	return (0);
386 }
387 
388 /*
389  * open the device. Make sure the partition info is a up-to-date as can be.
390  */
391 static int
392 sdopen(dev_t dev, int flag, int fmt, struct lwp *l)
393 {
394 	struct sd_softc *sd;
395 	struct scsipi_periph *periph;
396 	struct scsipi_adapter *adapt;
397 	int unit, part;
398 	int error;
399 
400 	unit = SDUNIT(dev);
401 	sd = device_lookup_private(&sd_cd, unit);
402 	if (sd == NULL)
403 		return (ENXIO);
404 
405 	if (!device_is_active(sd->sc_dev))
406 		return (ENODEV);
407 
408 	part = SDPART(dev);
409 
410 	mutex_enter(&sd->sc_dk.dk_openlock);
411 
412 	/*
413 	 * If there are wedges, and this is not RAW_PART, then we
414 	 * need to fail.
415 	 */
416 	if (sd->sc_dk.dk_nwedges != 0 && part != RAW_PART) {
417 		error = EBUSY;
418 		goto bad1;
419 	}
420 
421 	periph = sd->sc_periph;
422 	adapt = periph->periph_channel->chan_adapter;
423 
424 	SC_DEBUG(periph, SCSIPI_DB1,
425 	    ("sdopen: dev=0x%"PRIx64" (unit %d (of %d), partition %d)\n", dev, unit,
426 	    sd_cd.cd_ndevs, part));
427 
428 	/*
429 	 * If this is the first open of this device, add a reference
430 	 * to the adapter.
431 	 */
432 	if (sd->sc_dk.dk_openmask == 0 &&
433 	    (error = scsipi_adapter_addref(adapt)) != 0)
434 		goto bad1;
435 
436 	if ((periph->periph_flags & PERIPH_OPEN) != 0) {
437 		/*
438 		 * If any partition is open, but the disk has been invalidated,
439 		 * disallow further opens of non-raw partition
440 		 */
441 		if ((periph->periph_flags & PERIPH_MEDIA_LOADED) == 0 &&
442 		    (part != RAW_PART || fmt != S_IFCHR)) {
443 			error = EIO;
444 			goto bad2;
445 		}
446 	} else {
447 		int silent;
448 
449 		if ((part == RAW_PART && fmt == S_IFCHR) || (flag & FSILENT))
450 			silent = XS_CTL_SILENT;
451 		else
452 			silent = 0;
453 
454 		/* Check that it is still responding and ok. */
455 		error = scsipi_test_unit_ready(periph,
456 		    XS_CTL_IGNORE_ILLEGAL_REQUEST | XS_CTL_IGNORE_MEDIA_CHANGE |
457 		    silent);
458 
459 		/*
460 		 * Start the pack spinning if necessary. Always allow the
461 		 * raw parition to be opened, for raw IOCTLs. Data transfers
462 		 * will check for SDEV_MEDIA_LOADED.
463 		 */
464 		if (error == EIO) {
465 			int error2;
466 
467 			error2 = scsipi_start(periph, SSS_START, silent);
468 			switch (error2) {
469 			case 0:
470 				error = 0;
471 				break;
472 			case EIO:
473 			case EINVAL:
474 				break;
475 			default:
476 				error = error2;
477 				break;
478 			}
479 		}
480 		if (error) {
481 			if (silent && (flag & FSILENT) == 0)
482 				goto out;
483 			goto bad2;
484 		}
485 
486 		periph->periph_flags |= PERIPH_OPEN;
487 
488 		if (periph->periph_flags & PERIPH_REMOVABLE) {
489 			/* Lock the pack in. */
490 			error = scsipi_prevent(periph, SPAMR_PREVENT_DT,
491 			    XS_CTL_IGNORE_ILLEGAL_REQUEST |
492 			    XS_CTL_IGNORE_MEDIA_CHANGE |
493 			    XS_CTL_SILENT);
494 			if (error)
495 				goto bad3;
496 		}
497 
498 		if ((periph->periph_flags & PERIPH_MEDIA_LOADED) == 0) {
499 			int param_error;
500 			periph->periph_flags |= PERIPH_MEDIA_LOADED;
501 
502 			/*
503 			 * Load the physical device parameters.
504 			 *
505 			 * Note that if media is present but unformatted,
506 			 * we allow the open (so that it can be formatted!).
507 			 * The drive should refuse real I/O, if the media is
508 			 * unformatted.
509 			 */
510 			if ((param_error = sd_get_parms(sd, &sd->params, 0))
511 			     == SDGP_RESULT_OFFLINE) {
512 				error = ENXIO;
513 				periph->periph_flags &= ~PERIPH_MEDIA_LOADED;
514 				goto bad3;
515 			}
516 			SC_DEBUG(periph, SCSIPI_DB3, ("Params loaded "));
517 
518 			/* Load the partition info if not already loaded. */
519 			if (param_error == 0) {
520 				if ((sdgetdisklabel(sd) != 0) && (part != RAW_PART)) {
521 					error = EIO;
522 					goto bad3;
523 				}
524 				SC_DEBUG(periph, SCSIPI_DB3,
525 				     ("Disklabel loaded "));
526 			}
527 		}
528 	}
529 
530 	/* Check that the partition exists. */
531 	if (part != RAW_PART &&
532 	    (part >= sd->sc_dk.dk_label->d_npartitions ||
533 	     sd->sc_dk.dk_label->d_partitions[part].p_fstype == FS_UNUSED)) {
534 		error = ENXIO;
535 		goto bad3;
536 	}
537 
538  out:	/* Insure only one open at a time. */
539 	switch (fmt) {
540 	case S_IFCHR:
541 		sd->sc_dk.dk_copenmask |= (1 << part);
542 		break;
543 	case S_IFBLK:
544 		sd->sc_dk.dk_bopenmask |= (1 << part);
545 		break;
546 	}
547 	sd->sc_dk.dk_openmask =
548 	    sd->sc_dk.dk_copenmask | sd->sc_dk.dk_bopenmask;
549 
550 	SC_DEBUG(periph, SCSIPI_DB3, ("open complete\n"));
551 	mutex_exit(&sd->sc_dk.dk_openlock);
552 	return (0);
553 
554  bad3:
555 	if (sd->sc_dk.dk_openmask == 0) {
556 		if (periph->periph_flags & PERIPH_REMOVABLE)
557 			scsipi_prevent(periph, SPAMR_ALLOW,
558 			    XS_CTL_IGNORE_ILLEGAL_REQUEST |
559 			    XS_CTL_IGNORE_MEDIA_CHANGE |
560 			    XS_CTL_SILENT);
561 		periph->periph_flags &= ~PERIPH_OPEN;
562 	}
563 
564  bad2:
565 	if (sd->sc_dk.dk_openmask == 0)
566 		scsipi_adapter_delref(adapt);
567 
568  bad1:
569 	mutex_exit(&sd->sc_dk.dk_openlock);
570 	return (error);
571 }
572 
573 /*
574  * Caller must hold sd->sc_dk.dk_openlock.
575  */
576 static int
577 sdlastclose(device_t self)
578 {
579 	struct sd_softc *sd = device_private(self);
580 	struct scsipi_periph *periph = sd->sc_periph;
581 	struct scsipi_adapter *adapt = periph->periph_channel->chan_adapter;
582 
583 	/*
584 	 * If the disk cache needs flushing, and the disk supports
585 	 * it, do it now.
586 	 */
587 	if ((sd->flags & SDF_DIRTY) != 0) {
588 		if (sd_flush(sd, 0)) {
589 			aprint_error_dev(sd->sc_dev,
590 				"cache synchronization failed\n");
591 			sd->flags &= ~SDF_FLUSHING;
592 		} else
593 			sd->flags &= ~(SDF_FLUSHING|SDF_DIRTY);
594 	}
595 
596 	scsipi_wait_drain(periph);
597 
598 	if (periph->periph_flags & PERIPH_REMOVABLE)
599 		scsipi_prevent(periph, SPAMR_ALLOW,
600 		    XS_CTL_IGNORE_ILLEGAL_REQUEST |
601 		    XS_CTL_IGNORE_NOT_READY |
602 		    XS_CTL_SILENT);
603 	periph->periph_flags &= ~PERIPH_OPEN;
604 
605 	scsipi_wait_drain(periph);
606 
607 	scsipi_adapter_delref(adapt);
608 
609 	return 0;
610 }
611 
612 /*
613  * close the device.. only called if we are the LAST occurence of an open
614  * device.  Convenient now but usually a pain.
615  */
616 static int
617 sdclose(dev_t dev, int flag, int fmt, struct lwp *l)
618 {
619 	struct sd_softc *sd = device_lookup_private(&sd_cd, SDUNIT(dev));
620 	int part = SDPART(dev);
621 
622 	mutex_enter(&sd->sc_dk.dk_openlock);
623 	switch (fmt) {
624 	case S_IFCHR:
625 		sd->sc_dk.dk_copenmask &= ~(1 << part);
626 		break;
627 	case S_IFBLK:
628 		sd->sc_dk.dk_bopenmask &= ~(1 << part);
629 		break;
630 	}
631 	sd->sc_dk.dk_openmask =
632 	    sd->sc_dk.dk_copenmask | sd->sc_dk.dk_bopenmask;
633 
634 	if (sd->sc_dk.dk_openmask == 0)
635 		sdlastclose(sd->sc_dev);
636 
637 	mutex_exit(&sd->sc_dk.dk_openlock);
638 	return (0);
639 }
640 
641 /*
642  * Actually translate the requested transfer into one the physical driver
643  * can understand.  The transfer is described by a buf and will include
644  * only one physical transfer.
645  */
646 static void
647 sdstrategy(struct buf *bp)
648 {
649 	struct sd_softc *sd = device_lookup_private(&sd_cd, SDUNIT(bp->b_dev));
650 	struct scsipi_periph *periph = sd->sc_periph;
651 	struct disklabel *lp;
652 	daddr_t blkno;
653 	int s;
654 	bool sector_aligned;
655 
656 	SC_DEBUG(sd->sc_periph, SCSIPI_DB2, ("sdstrategy "));
657 	SC_DEBUG(sd->sc_periph, SCSIPI_DB1,
658 	    ("%d bytes @ blk %" PRId64 "\n", bp->b_bcount, bp->b_blkno));
659 	/*
660 	 * If the device has been made invalid, error out
661 	 */
662 	if ((periph->periph_flags & PERIPH_MEDIA_LOADED) == 0 ||
663 	    !device_is_active(sd->sc_dev)) {
664 		if (periph->periph_flags & PERIPH_OPEN)
665 			bp->b_error = EIO;
666 		else
667 			bp->b_error = ENODEV;
668 		goto done;
669 	}
670 
671 	lp = sd->sc_dk.dk_label;
672 
673 	/*
674 	 * The transfer must be a whole number of blocks, offset must not be
675 	 * negative.
676 	 */
677 	if (lp->d_secsize == DEV_BSIZE) {
678 		sector_aligned = (bp->b_bcount & (DEV_BSIZE - 1)) == 0;
679 	} else {
680 		sector_aligned = (bp->b_bcount % lp->d_secsize) == 0;
681 	}
682 	if (!sector_aligned || bp->b_blkno < 0) {
683 		bp->b_error = EINVAL;
684 		goto done;
685 	}
686 	/*
687 	 * If it's a null transfer, return immediatly
688 	 */
689 	if (bp->b_bcount == 0)
690 		goto done;
691 
692 	/*
693 	 * Do bounds checking, adjust transfer. if error, process.
694 	 * If end of partition, just return.
695 	 */
696 	if (SDPART(bp->b_dev) == RAW_PART) {
697 		if (bounds_check_with_mediasize(bp, DEV_BSIZE,
698 		    sd->params.disksize512) <= 0)
699 			goto done;
700 	} else {
701 		if (bounds_check_with_label(&sd->sc_dk, bp,
702 		    (sd->flags & (SDF_WLABEL|SDF_LABELLING)) != 0) <= 0)
703 			goto done;
704 	}
705 
706 	/*
707 	 * Now convert the block number to absolute and put it in
708 	 * terms of the device's logical block size.
709 	 */
710 	if (lp->d_secsize == DEV_BSIZE)
711 		blkno = bp->b_blkno;
712 	else if (lp->d_secsize > DEV_BSIZE)
713 		blkno = bp->b_blkno / (lp->d_secsize / DEV_BSIZE);
714 	else
715 		blkno = bp->b_blkno * (DEV_BSIZE / lp->d_secsize);
716 
717 	if (SDPART(bp->b_dev) != RAW_PART)
718 		blkno += lp->d_partitions[SDPART(bp->b_dev)].p_offset;
719 
720 	bp->b_rawblkno = blkno;
721 
722 	s = splbio();
723 
724 	/*
725 	 * Place it in the queue of disk activities for this disk.
726 	 *
727 	 * XXX Only do disksort() if the current operating mode does not
728 	 * XXX include tagged queueing.
729 	 */
730 	bufq_put(sd->buf_queue, bp);
731 
732 	/*
733 	 * Tell the device to get going on the transfer if it's
734 	 * not doing anything, otherwise just wait for completion
735 	 */
736 	sdstart(sd->sc_periph);
737 
738 	splx(s);
739 	return;
740 
741 done:
742 	/*
743 	 * Correctly set the buf to indicate a completed xfer
744 	 */
745 	bp->b_resid = bp->b_bcount;
746 	biodone(bp);
747 }
748 
749 /*
750  * sdstart looks to see if there is a buf waiting for the device
751  * and that the device is not already busy. If both are true,
752  * It dequeues the buf and creates a scsi command to perform the
753  * transfer in the buf. The transfer request will call scsipi_done
754  * on completion, which will in turn call this routine again
755  * so that the next queued transfer is performed.
756  * The bufs are queued by the strategy routine (sdstrategy)
757  *
758  * This routine is also called after other non-queued requests
759  * have been made of the scsi driver, to ensure that the queue
760  * continues to be drained.
761  *
762  * must be called at the correct (highish) spl level
763  * sdstart() is called at splbio from sdstrategy, sdrestart and scsipi_done
764  */
765 static void
766 sdstart(struct scsipi_periph *periph)
767 {
768 	struct sd_softc *sd = device_private(periph->periph_dev);
769 	struct disklabel *lp = sd->sc_dk.dk_label;
770 	struct buf *bp = 0;
771 	struct scsipi_rw_16 cmd16;
772 	struct scsipi_rw_10 cmd_big;
773 	struct scsi_rw_6 cmd_small;
774 	struct scsipi_generic *cmdp;
775 	struct scsipi_xfer *xs;
776 	int nblks, cmdlen, error, flags;
777 
778 	SC_DEBUG(periph, SCSIPI_DB2, ("sdstart "));
779 	/*
780 	 * Check if the device has room for another command
781 	 */
782 	while (periph->periph_active < periph->periph_openings) {
783 		/*
784 		 * there is excess capacity, but a special waits
785 		 * It'll need the adapter as soon as we clear out of the
786 		 * way and let it run (user level wait).
787 		 */
788 		if (periph->periph_flags & PERIPH_WAITING) {
789 			periph->periph_flags &= ~PERIPH_WAITING;
790 			wakeup((void *)periph);
791 			return;
792 		}
793 
794 		/*
795 		 * If the device has become invalid, abort all the
796 		 * reads and writes until all files have been closed and
797 		 * re-opened
798 		 */
799 		if (__predict_false(
800 		    (periph->periph_flags & PERIPH_MEDIA_LOADED) == 0)) {
801 			if ((bp = bufq_get(sd->buf_queue)) != NULL) {
802 				bp->b_error = EIO;
803 				bp->b_resid = bp->b_bcount;
804 				biodone(bp);
805 				continue;
806 			} else {
807 				return;
808 			}
809 		}
810 
811 		/*
812 		 * See if there is a buf with work for us to do..
813 		 */
814 		if ((bp = bufq_peek(sd->buf_queue)) == NULL)
815 			return;
816 
817 		/*
818 		 * We have a buf, now we should make a command.
819 		 */
820 
821 		if (lp->d_secsize == DEV_BSIZE)
822 			nblks = bp->b_bcount >> DEV_BSHIFT;
823 		else
824 			nblks = howmany(bp->b_bcount, lp->d_secsize);
825 
826 		/*
827 		 * Fill out the scsi command.  Use the smallest CDB possible
828 		 * (6-byte, 10-byte, or 16-byte).
829 		 */
830 		if (((bp->b_rawblkno & 0x1fffff) == bp->b_rawblkno) &&
831 		    ((nblks & 0xff) == nblks) &&
832 		    !(periph->periph_quirks & PQUIRK_ONLYBIG)) {
833 			/* 6-byte CDB */
834 			memset(&cmd_small, 0, sizeof(cmd_small));
835 			cmd_small.opcode = (bp->b_flags & B_READ) ?
836 			    SCSI_READ_6_COMMAND : SCSI_WRITE_6_COMMAND;
837 			_lto3b(bp->b_rawblkno, cmd_small.addr);
838 			cmd_small.length = nblks & 0xff;
839 			cmdlen = sizeof(cmd_small);
840 			cmdp = (struct scsipi_generic *)&cmd_small;
841 		} else if ((bp->b_rawblkno & 0xffffffff) == bp->b_rawblkno) {
842 			/* 10-byte CDB */
843 			memset(&cmd_big, 0, sizeof(cmd_big));
844 			cmd_big.opcode = (bp->b_flags & B_READ) ?
845 			    READ_10 : WRITE_10;
846 			_lto4b(bp->b_rawblkno, cmd_big.addr);
847 			_lto2b(nblks, cmd_big.length);
848 			cmdlen = sizeof(cmd_big);
849 			cmdp = (struct scsipi_generic *)&cmd_big;
850 		} else {
851 			/* 16-byte CDB */
852 			memset(&cmd16, 0, sizeof(cmd16));
853 			cmd16.opcode = (bp->b_flags & B_READ) ?
854 			    READ_16 : WRITE_16;
855 			_lto8b(bp->b_rawblkno, cmd16.addr);
856 			_lto4b(nblks, cmd16.length);
857 			cmdlen = sizeof(cmd16);
858 			cmdp = (struct scsipi_generic *)&cmd16;
859 		}
860 
861 		/* Instrumentation. */
862 		disk_busy(&sd->sc_dk);
863 
864 		/*
865 		 * Mark the disk dirty so that the cache will be
866 		 * flushed on close.
867 		 */
868 		if ((bp->b_flags & B_READ) == 0)
869 			sd->flags |= SDF_DIRTY;
870 
871 		/*
872 		 * Figure out what flags to use.
873 		 */
874 		flags = XS_CTL_NOSLEEP|XS_CTL_ASYNC|XS_CTL_SIMPLE_TAG;
875 		if (bp->b_flags & B_READ)
876 			flags |= XS_CTL_DATA_IN;
877 		else
878 			flags |= XS_CTL_DATA_OUT;
879 
880 		/*
881 		 * Call the routine that chats with the adapter.
882 		 * Note: we cannot sleep as we may be an interrupt
883 		 */
884 		xs = scsipi_make_xs(periph, cmdp, cmdlen,
885 		    (u_char *)bp->b_data, bp->b_bcount,
886 		    SDRETRIES, SD_IO_TIMEOUT, bp, flags);
887 		if (__predict_false(xs == NULL)) {
888 			/*
889 			 * out of memory. Keep this buffer in the queue, and
890 			 * retry later.
891 			 */
892 			callout_reset(&sd->sc_callout, hz / 2, sdrestart,
893 			    periph);
894 			return;
895 		}
896 		/*
897 		 * need to dequeue the buffer before queuing the command,
898 		 * because cdstart may be called recursively from the
899 		 * HBA driver
900 		 */
901 #ifdef DIAGNOSTIC
902 		if (bufq_get(sd->buf_queue) != bp)
903 			panic("sdstart(): dequeued wrong buf");
904 #else
905 		bufq_get(sd->buf_queue);
906 #endif
907 		error = scsipi_execute_xs(xs);
908 		/* with a scsipi_xfer preallocated, scsipi_command can't fail */
909 		KASSERT(error == 0);
910 	}
911 }
912 
913 static void
914 sdrestart(void *v)
915 {
916 	int s = splbio();
917 	sdstart((struct scsipi_periph *)v);
918 	splx(s);
919 }
920 
921 static void
922 sddone(struct scsipi_xfer *xs, int error)
923 {
924 	struct sd_softc *sd = device_private(xs->xs_periph->periph_dev);
925 	struct buf *bp = xs->bp;
926 
927 	if (sd->flags & SDF_FLUSHING) {
928 		/* Flush completed, no longer dirty. */
929 		sd->flags &= ~(SDF_FLUSHING|SDF_DIRTY);
930 	}
931 
932 	if (bp) {
933 		bp->b_error = error;
934 		bp->b_resid = xs->resid;
935 		if (error) {
936 			/* on a read/write error bp->b_resid is zero, so fix */
937 			bp->b_resid = bp->b_bcount;
938 		}
939 
940 		disk_unbusy(&sd->sc_dk, bp->b_bcount - bp->b_resid,
941 		    (bp->b_flags & B_READ));
942 		rnd_add_uint32(&sd->rnd_source, bp->b_rawblkno);
943 
944 		biodone(bp);
945 	}
946 }
947 
948 static void
949 sdminphys(struct buf *bp)
950 {
951 	struct sd_softc *sd = device_lookup_private(&sd_cd, SDUNIT(bp->b_dev));
952 	long xmax;
953 
954 	/*
955 	 * If the device is ancient, we want to make sure that
956 	 * the transfer fits into a 6-byte cdb.
957 	 *
958 	 * XXX Note that the SCSI-I spec says that 256-block transfers
959 	 * are allowed in a 6-byte read/write, and are specified
960 	 * by settng the "length" to 0.  However, we're conservative
961 	 * here, allowing only 255-block transfers in case an
962 	 * ancient device gets confused by length == 0.  A length of 0
963 	 * in a 10-byte read/write actually means 0 blocks.
964 	 */
965 	if ((sd->flags & SDF_ANCIENT) &&
966 	    ((sd->sc_periph->periph_flags &
967 	    (PERIPH_REMOVABLE | PERIPH_MEDIA_LOADED)) != PERIPH_REMOVABLE)) {
968 		xmax = sd->sc_dk.dk_label->d_secsize * 0xff;
969 
970 		if (bp->b_bcount > xmax)
971 			bp->b_bcount = xmax;
972 	}
973 
974 	scsipi_adapter_minphys(sd->sc_periph->periph_channel, bp);
975 }
976 
977 static int
978 sdread(dev_t dev, struct uio *uio, int ioflag)
979 {
980 
981 	return (physio(sdstrategy, NULL, dev, B_READ, sdminphys, uio));
982 }
983 
984 static int
985 sdwrite(dev_t dev, struct uio *uio, int ioflag)
986 {
987 
988 	return (physio(sdstrategy, NULL, dev, B_WRITE, sdminphys, uio));
989 }
990 
991 /*
992  * Perform special action on behalf of the user
993  * Knows about the internals of this device
994  */
995 static int
996 sdioctl(dev_t dev, u_long cmd, void *addr, int flag, struct lwp *l)
997 {
998 	struct sd_softc *sd = device_lookup_private(&sd_cd, SDUNIT(dev));
999 	struct scsipi_periph *periph = sd->sc_periph;
1000 	int part = SDPART(dev);
1001 	int error = 0;
1002 	int s;
1003 #ifdef __HAVE_OLD_DISKLABEL
1004 	struct disklabel *newlabel = NULL;
1005 #endif
1006 
1007 	SC_DEBUG(sd->sc_periph, SCSIPI_DB2, ("sdioctl 0x%lx ", cmd));
1008 
1009 	/*
1010 	 * If the device is not valid, some IOCTLs can still be
1011 	 * handled on the raw partition. Check this here.
1012 	 */
1013 	if ((periph->periph_flags & PERIPH_MEDIA_LOADED) == 0) {
1014 		switch (cmd) {
1015 		case DIOCKLABEL:
1016 		case DIOCWLABEL:
1017 		case DIOCLOCK:
1018 		case DIOCEJECT:
1019 		case ODIOCEJECT:
1020 		case DIOCGCACHE:
1021 		case DIOCSCACHE:
1022 		case DIOCGSTRATEGY:
1023 		case DIOCSSTRATEGY:
1024 		case SCIOCIDENTIFY:
1025 		case OSCIOCIDENTIFY:
1026 		case SCIOCCOMMAND:
1027 		case SCIOCDEBUG:
1028 			if (part == RAW_PART)
1029 				break;
1030 		/* FALLTHROUGH */
1031 		default:
1032 			if ((periph->periph_flags & PERIPH_OPEN) == 0)
1033 				return (ENODEV);
1034 			else
1035 				return (EIO);
1036 		}
1037 	}
1038 
1039 	error = disk_ioctl(&sd->sc_dk, cmd, addr, flag, l);
1040 	if (error != EPASSTHROUGH)
1041 		return (error);
1042 
1043 	switch (cmd) {
1044 	case DIOCGDINFO:
1045 		*(struct disklabel *)addr = *(sd->sc_dk.dk_label);
1046 		return (0);
1047 
1048 #ifdef __HAVE_OLD_DISKLABEL
1049 	case ODIOCGDINFO:
1050 		newlabel = malloc(sizeof *newlabel, M_TEMP, M_WAITOK);
1051 		if (newlabel == NULL)
1052 			return EIO;
1053 		memcpy(newlabel, sd->sc_dk.dk_label, sizeof (*newlabel));
1054 		if (newlabel->d_npartitions <= OLDMAXPARTITIONS)
1055 			memcpy(addr, newlabel, sizeof (struct olddisklabel));
1056 		else
1057 			error = ENOTTY;
1058 		free(newlabel, M_TEMP);
1059 		return error;
1060 #endif
1061 
1062 	case DIOCGPART:
1063 		((struct partinfo *)addr)->disklab = sd->sc_dk.dk_label;
1064 		((struct partinfo *)addr)->part =
1065 		    &sd->sc_dk.dk_label->d_partitions[part];
1066 		return (0);
1067 
1068 	case DIOCWDINFO:
1069 	case DIOCSDINFO:
1070 #ifdef __HAVE_OLD_DISKLABEL
1071 	case ODIOCWDINFO:
1072 	case ODIOCSDINFO:
1073 #endif
1074 	{
1075 		struct disklabel *lp;
1076 
1077 		if ((flag & FWRITE) == 0)
1078 			return (EBADF);
1079 
1080 #ifdef __HAVE_OLD_DISKLABEL
1081  		if (cmd == ODIOCSDINFO || cmd == ODIOCWDINFO) {
1082 			newlabel = malloc(sizeof *newlabel, M_TEMP,
1083 			    M_WAITOK | M_ZERO);
1084 			if (newlabel == NULL)
1085 				return EIO;
1086 			memcpy(newlabel, addr, sizeof (struct olddisklabel));
1087 			lp = newlabel;
1088 		} else
1089 #endif
1090 		lp = (struct disklabel *)addr;
1091 
1092 		mutex_enter(&sd->sc_dk.dk_openlock);
1093 		sd->flags |= SDF_LABELLING;
1094 
1095 		error = setdisklabel(sd->sc_dk.dk_label,
1096 		    lp, /*sd->sc_dk.dk_openmask : */0,
1097 		    sd->sc_dk.dk_cpulabel);
1098 		if (error == 0) {
1099 			if (cmd == DIOCWDINFO
1100 #ifdef __HAVE_OLD_DISKLABEL
1101 			    || cmd == ODIOCWDINFO
1102 #endif
1103 			   )
1104 				error = writedisklabel(SDLABELDEV(dev),
1105 				    sdstrategy, sd->sc_dk.dk_label,
1106 				    sd->sc_dk.dk_cpulabel);
1107 		}
1108 
1109 		sd->flags &= ~SDF_LABELLING;
1110 		mutex_exit(&sd->sc_dk.dk_openlock);
1111 #ifdef __HAVE_OLD_DISKLABEL
1112 		if (newlabel != NULL)
1113 			free(newlabel, M_TEMP);
1114 #endif
1115 		return (error);
1116 	}
1117 
1118 	case DIOCKLABEL:
1119 		if (*(int *)addr)
1120 			periph->periph_flags |= PERIPH_KEEP_LABEL;
1121 		else
1122 			periph->periph_flags &= ~PERIPH_KEEP_LABEL;
1123 		return (0);
1124 
1125 	case DIOCWLABEL:
1126 		if ((flag & FWRITE) == 0)
1127 			return (EBADF);
1128 		if (*(int *)addr)
1129 			sd->flags |= SDF_WLABEL;
1130 		else
1131 			sd->flags &= ~SDF_WLABEL;
1132 		return (0);
1133 
1134 	case DIOCLOCK:
1135 		if (periph->periph_flags & PERIPH_REMOVABLE)
1136 			return (scsipi_prevent(periph,
1137 			    (*(int *)addr) ?
1138 			    SPAMR_PREVENT_DT : SPAMR_ALLOW, 0));
1139 		else
1140 			return (ENOTTY);
1141 
1142 	case DIOCEJECT:
1143 		if ((periph->periph_flags & PERIPH_REMOVABLE) == 0)
1144 			return (ENOTTY);
1145 		if (*(int *)addr == 0) {
1146 			/*
1147 			 * Don't force eject: check that we are the only
1148 			 * partition open. If so, unlock it.
1149 			 */
1150 			if ((sd->sc_dk.dk_openmask & ~(1 << part)) == 0 &&
1151 			    sd->sc_dk.dk_bopenmask + sd->sc_dk.dk_copenmask ==
1152 			    sd->sc_dk.dk_openmask) {
1153 				error = scsipi_prevent(periph, SPAMR_ALLOW,
1154 				    XS_CTL_IGNORE_NOT_READY);
1155 				if (error)
1156 					return (error);
1157 			} else {
1158 				return (EBUSY);
1159 			}
1160 		}
1161 		/* FALLTHROUGH */
1162 	case ODIOCEJECT:
1163 		return ((periph->periph_flags & PERIPH_REMOVABLE) == 0 ?
1164 		    ENOTTY : scsipi_start(periph, SSS_STOP|SSS_LOEJ, 0));
1165 
1166 	case DIOCGDEFLABEL:
1167 		sdgetdefaultlabel(sd, (struct disklabel *)addr);
1168 		return (0);
1169 
1170 #ifdef __HAVE_OLD_DISKLABEL
1171 	case ODIOCGDEFLABEL:
1172 		newlabel = malloc(sizeof *newlabel, M_TEMP, M_WAITOK);
1173 		if (newlabel == NULL)
1174 			return EIO;
1175 		sdgetdefaultlabel(sd, newlabel);
1176 		if (newlabel->d_npartitions <= OLDMAXPARTITIONS)
1177 			memcpy(addr, newlabel, sizeof (struct olddisklabel));
1178 		else
1179 			error = ENOTTY;
1180 		free(newlabel, M_TEMP);
1181 		return error;
1182 #endif
1183 
1184 	case DIOCGCACHE:
1185 		return (sd_getcache(sd, (int *) addr));
1186 
1187 	case DIOCSCACHE:
1188 		if ((flag & FWRITE) == 0)
1189 			return (EBADF);
1190 		return (sd_setcache(sd, *(int *) addr));
1191 
1192 	case DIOCCACHESYNC:
1193 		/*
1194 		 * XXX Do we really need to care about having a writable
1195 		 * file descriptor here?
1196 		 */
1197 		if ((flag & FWRITE) == 0)
1198 			return (EBADF);
1199 		if (((sd->flags & SDF_DIRTY) != 0 || *(int *)addr != 0)) {
1200 			error = sd_flush(sd, 0);
1201 			if (error)
1202 				sd->flags &= ~SDF_FLUSHING;
1203 			else
1204 				sd->flags &= ~(SDF_FLUSHING|SDF_DIRTY);
1205 		} else
1206 			error = 0;
1207 		return (error);
1208 
1209 	case DIOCAWEDGE:
1210 	    {
1211 	    	struct dkwedge_info *dkw = (void *) addr;
1212 
1213 		if ((flag & FWRITE) == 0)
1214 			return (EBADF);
1215 
1216 		/* If the ioctl happens here, the parent is us. */
1217 		strlcpy(dkw->dkw_parent, device_xname(sd->sc_dev),
1218 			sizeof(dkw->dkw_parent));
1219 		return (dkwedge_add(dkw));
1220 	    }
1221 
1222 	case DIOCDWEDGE:
1223 	    {
1224 	    	struct dkwedge_info *dkw = (void *) addr;
1225 
1226 		if ((flag & FWRITE) == 0)
1227 			return (EBADF);
1228 
1229 		/* If the ioctl happens here, the parent is us. */
1230 		strlcpy(dkw->dkw_parent, device_xname(sd->sc_dev),
1231 			sizeof(dkw->dkw_parent));
1232 		return (dkwedge_del(dkw));
1233 	    }
1234 
1235 	case DIOCLWEDGES:
1236 	    {
1237 	    	struct dkwedge_list *dkwl = (void *) addr;
1238 
1239 		return (dkwedge_list(&sd->sc_dk, dkwl, l));
1240 	    }
1241 
1242 	case DIOCGSTRATEGY:
1243 	    {
1244 		struct disk_strategy *dks = addr;
1245 
1246 		s = splbio();
1247 		strlcpy(dks->dks_name, bufq_getstrategyname(sd->buf_queue),
1248 		    sizeof(dks->dks_name));
1249 		splx(s);
1250 		dks->dks_paramlen = 0;
1251 
1252 		return 0;
1253 	    }
1254 
1255 	case DIOCSSTRATEGY:
1256 	    {
1257 		struct disk_strategy *dks = addr;
1258 		struct bufq_state *new;
1259 		struct bufq_state *old;
1260 
1261 		if ((flag & FWRITE) == 0) {
1262 			return EBADF;
1263 		}
1264 
1265 		if (dks->dks_param != NULL) {
1266 			return EINVAL;
1267 		}
1268 		dks->dks_name[sizeof(dks->dks_name) - 1] = 0; /* ensure term */
1269 		error = bufq_alloc(&new, dks->dks_name,
1270 		    BUFQ_EXACT|BUFQ_SORT_RAWBLOCK);
1271 		if (error) {
1272 			return error;
1273 		}
1274 		s = splbio();
1275 		old = sd->buf_queue;
1276 		bufq_move(new, old);
1277 		sd->buf_queue = new;
1278 		splx(s);
1279 		bufq_free(old);
1280 
1281 		return 0;
1282 	    }
1283 
1284 	default:
1285 		if (part != RAW_PART)
1286 			return (ENOTTY);
1287 		return (scsipi_do_ioctl(periph, dev, cmd, addr, flag, l));
1288 	}
1289 
1290 #ifdef DIAGNOSTIC
1291 	panic("sdioctl: impossible");
1292 #endif
1293 }
1294 
1295 static void
1296 sdgetdefaultlabel(struct sd_softc *sd, struct disklabel *lp)
1297 {
1298 
1299 	memset(lp, 0, sizeof(struct disklabel));
1300 
1301 	lp->d_secsize = sd->params.blksize;
1302 	lp->d_ntracks = sd->params.heads;
1303 	lp->d_nsectors = sd->params.sectors;
1304 	lp->d_ncylinders = sd->params.cyls;
1305 	lp->d_secpercyl = lp->d_ntracks * lp->d_nsectors;
1306 
1307 	switch (SCSIPI_BUSTYPE_TYPE(scsipi_periph_bustype(sd->sc_periph))) {
1308 	case SCSIPI_BUSTYPE_SCSI:
1309 		lp->d_type = DTYPE_SCSI;
1310 		break;
1311 	case SCSIPI_BUSTYPE_ATAPI:
1312 		lp->d_type = DTYPE_ATAPI;
1313 		break;
1314 	}
1315 	/*
1316 	 * XXX
1317 	 * We could probe the mode pages to figure out what kind of disc it is.
1318 	 * Is this worthwhile?
1319 	 */
1320 	strncpy(lp->d_typename, sd->name, 16);
1321 	strncpy(lp->d_packname, "fictitious", 16);
1322 	if (sd->params.disksize > UINT32_MAX)
1323 		lp->d_secperunit = UINT32_MAX;
1324 	else
1325 		lp->d_secperunit = sd->params.disksize;
1326 	lp->d_rpm = sd->params.rot_rate;
1327 	lp->d_interleave = 1;
1328 	lp->d_flags = sd->sc_periph->periph_flags & PERIPH_REMOVABLE ?
1329 	    D_REMOVABLE : 0;
1330 
1331 	lp->d_partitions[RAW_PART].p_offset = 0;
1332 	lp->d_partitions[RAW_PART].p_size = lp->d_secperunit;
1333 	lp->d_partitions[RAW_PART].p_fstype = FS_UNUSED;
1334 	lp->d_npartitions = RAW_PART + 1;
1335 
1336 	lp->d_magic = DISKMAGIC;
1337 	lp->d_magic2 = DISKMAGIC;
1338 	lp->d_checksum = dkcksum(lp);
1339 }
1340 
1341 
1342 /*
1343  * Load the label information on the named device
1344  */
1345 static int
1346 sdgetdisklabel(struct sd_softc *sd)
1347 {
1348 	struct disklabel *lp = sd->sc_dk.dk_label;
1349 	const char *errstring;
1350 
1351 	memset(sd->sc_dk.dk_cpulabel, 0, sizeof(struct cpu_disklabel));
1352 
1353 	sdgetdefaultlabel(sd, lp);
1354 
1355 	if (lp->d_secpercyl == 0) {
1356 		lp->d_secpercyl = 100;
1357 		/* as long as it's not 0 - readdisklabel divides by it (?) */
1358 	}
1359 
1360 	/*
1361 	 * Call the generic disklabel extraction routine
1362 	 */
1363 	errstring = readdisklabel(MAKESDDEV(0, device_unit(sd->sc_dev),
1364 	    RAW_PART), sdstrategy, lp, sd->sc_dk.dk_cpulabel);
1365 	if (errstring) {
1366 		aprint_error_dev(sd->sc_dev, "%s\n", errstring);
1367 		return EIO;
1368 	}
1369 	return 0;
1370 }
1371 
1372 static bool
1373 sd_shutdown(device_t self, int how)
1374 {
1375 	struct sd_softc *sd = device_private(self);
1376 
1377 	/*
1378 	 * If the disk cache needs to be flushed, and the disk supports
1379 	 * it, flush it.  We're cold at this point, so we poll for
1380 	 * completion.
1381 	 */
1382 	if ((sd->flags & SDF_DIRTY) != 0) {
1383 		if (sd_flush(sd, XS_CTL_NOSLEEP|XS_CTL_POLL)) {
1384 			aprint_error_dev(sd->sc_dev,
1385 				"cache synchronization failed\n");
1386 			sd->flags &= ~SDF_FLUSHING;
1387 		} else
1388 			sd->flags &= ~(SDF_FLUSHING|SDF_DIRTY);
1389 	}
1390 	return true;
1391 }
1392 
1393 static bool
1394 sd_suspend(device_t dv, const pmf_qual_t *qual)
1395 {
1396 	return sd_shutdown(dv, boothowto); /* XXX no need to poll */
1397 }
1398 
1399 /*
1400  * Check Errors
1401  */
1402 static int
1403 sd_interpret_sense(struct scsipi_xfer *xs)
1404 {
1405 	struct scsipi_periph *periph = xs->xs_periph;
1406 	struct scsi_sense_data *sense = &xs->sense.scsi_sense;
1407 	struct sd_softc *sd = device_private(periph->periph_dev);
1408 	int s, error, retval = EJUSTRETURN;
1409 
1410 	/*
1411 	 * If the periph is already recovering, just do the normal
1412 	 * error processing.
1413 	 */
1414 	if (periph->periph_flags & PERIPH_RECOVERING)
1415 		return (retval);
1416 
1417 	/*
1418 	 * Ignore errors from accessing illegal fields (e.g. trying to
1419 	 * lock the door of a digicam, which doesn't have a door that
1420 	 * can be locked) for the SCSI_PREVENT_ALLOW_MEDIUM_REMOVAL command.
1421 	 */
1422 	if (xs->cmd->opcode == SCSI_PREVENT_ALLOW_MEDIUM_REMOVAL &&
1423 	    SSD_SENSE_KEY(sense->flags) == SKEY_ILLEGAL_REQUEST &&
1424 	    sense->asc == 0x24 &&
1425 	    sense->ascq == 0x00) { /* Illegal field in CDB */
1426 		if (!(xs->xs_control & XS_CTL_SILENT)) {
1427 			scsipi_printaddr(periph);
1428 			printf("no door lock\n");
1429 		}
1430 		xs->xs_control |= XS_CTL_IGNORE_ILLEGAL_REQUEST;
1431 		return (retval);
1432 	}
1433 
1434 
1435 
1436 	/*
1437 	 * If the device is not open yet, let the generic code handle it.
1438 	 */
1439 	if ((periph->periph_flags & PERIPH_MEDIA_LOADED) == 0)
1440 		return (retval);
1441 
1442 	/*
1443 	 * If it isn't a extended or extended/deferred error, let
1444 	 * the generic code handle it.
1445 	 */
1446 	if (SSD_RCODE(sense->response_code) != SSD_RCODE_CURRENT &&
1447 	    SSD_RCODE(sense->response_code) != SSD_RCODE_DEFERRED)
1448 		return (retval);
1449 
1450 	if (SSD_SENSE_KEY(sense->flags) == SKEY_NOT_READY &&
1451 	    sense->asc == 0x4) {
1452 		if (sense->ascq == 0x01)	{
1453 			/*
1454 			 * Unit In The Process Of Becoming Ready.
1455 			 */
1456 			printf("%s: waiting for pack to spin up...\n",
1457 			    device_xname(sd->sc_dev));
1458 			if (!callout_pending(&periph->periph_callout))
1459 				scsipi_periph_freeze(periph, 1);
1460 			callout_reset(&periph->periph_callout,
1461 			    5 * hz, scsipi_periph_timed_thaw, periph);
1462 			retval = ERESTART;
1463 		} else if (sense->ascq == 0x02) {
1464 			printf("%s: pack is stopped, restarting...\n",
1465 			    device_xname(sd->sc_dev));
1466 			s = splbio();
1467 			periph->periph_flags |= PERIPH_RECOVERING;
1468 			splx(s);
1469 			error = scsipi_start(periph, SSS_START,
1470 			    XS_CTL_URGENT|XS_CTL_HEAD_TAG|
1471 			    XS_CTL_THAW_PERIPH|XS_CTL_FREEZE_PERIPH);
1472 			if (error) {
1473 				aprint_error_dev(sd->sc_dev,
1474 					"unable to restart pack\n");
1475 				retval = error;
1476 			} else
1477 				retval = ERESTART;
1478 			s = splbio();
1479 			periph->periph_flags &= ~PERIPH_RECOVERING;
1480 			splx(s);
1481 		}
1482 	}
1483 	if (SSD_SENSE_KEY(sense->flags) == SKEY_MEDIUM_ERROR &&
1484 	    sense->asc == 0x31 &&
1485 	    sense->ascq == 0x00)	{ /* maybe for any asq ? */
1486 		/* Medium Format Corrupted */
1487 		retval = EFTYPE;
1488 	}
1489 	return (retval);
1490 }
1491 
1492 
1493 static int
1494 sdsize(dev_t dev)
1495 {
1496 	struct sd_softc *sd;
1497 	int part, unit, omask;
1498 	int size;
1499 
1500 	unit = SDUNIT(dev);
1501 	sd = device_lookup_private(&sd_cd, unit);
1502 	if (sd == NULL)
1503 		return (-1);
1504 
1505 	if (!device_is_active(sd->sc_dev))
1506 		return (-1);
1507 
1508 	part = SDPART(dev);
1509 	omask = sd->sc_dk.dk_openmask & (1 << part);
1510 
1511 	if (omask == 0 && sdopen(dev, 0, S_IFBLK, NULL) != 0)
1512 		return (-1);
1513 	if ((sd->sc_periph->periph_flags & PERIPH_MEDIA_LOADED) == 0)
1514 		size = -1;
1515 	else if (sd->sc_dk.dk_label->d_partitions[part].p_fstype != FS_SWAP)
1516 		size = -1;
1517 	else
1518 		size = sd->sc_dk.dk_label->d_partitions[part].p_size *
1519 		    (sd->sc_dk.dk_label->d_secsize / DEV_BSIZE);
1520 	if (omask == 0 && sdclose(dev, 0, S_IFBLK, NULL) != 0)
1521 		return (-1);
1522 	return (size);
1523 }
1524 
1525 /* #define SD_DUMP_NOT_TRUSTED if you just want to watch */
1526 static struct scsipi_xfer sx;
1527 static int sddoingadump;
1528 
1529 /*
1530  * dump all of physical memory into the partition specified, starting
1531  * at offset 'dumplo' into the partition.
1532  */
1533 static int
1534 sddump(dev_t dev, daddr_t blkno, void *va, size_t size)
1535 {
1536 	struct sd_softc *sd;	/* disk unit to do the I/O */
1537 	struct disklabel *lp;	/* disk's disklabel */
1538 	int	unit, part;
1539 	int	sectorsize;	/* size of a disk sector */
1540 	int	nsects;		/* number of sectors in partition */
1541 	int	sectoff;	/* sector offset of partition */
1542 	int	totwrt;		/* total number of sectors left to write */
1543 	int	nwrt;		/* current number of sectors to write */
1544 	struct scsipi_rw_10 cmd;	/* write command */
1545 	struct scsipi_xfer *xs;	/* ... convenience */
1546 	struct scsipi_periph *periph;
1547 	struct scsipi_channel *chan;
1548 
1549 	/* Check if recursive dump; if so, punt. */
1550 	if (sddoingadump)
1551 		return (EFAULT);
1552 
1553 	/* Mark as active early. */
1554 	sddoingadump = 1;
1555 
1556 	unit = SDUNIT(dev);	/* Decompose unit & partition. */
1557 	part = SDPART(dev);
1558 
1559 	/* Check for acceptable drive number. */
1560 	sd = device_lookup_private(&sd_cd, unit);
1561 	if (sd == NULL)
1562 		return (ENXIO);
1563 
1564 	if (!device_is_active(sd->sc_dev))
1565 		return (ENODEV);
1566 
1567 	periph = sd->sc_periph;
1568 	chan = periph->periph_channel;
1569 
1570 	/* Make sure it was initialized. */
1571 	if ((periph->periph_flags & PERIPH_MEDIA_LOADED) == 0)
1572 		return (ENXIO);
1573 
1574 	/* Convert to disk sectors.  Request must be a multiple of size. */
1575 	lp = sd->sc_dk.dk_label;
1576 	sectorsize = lp->d_secsize;
1577 	if ((size % sectorsize) != 0)
1578 		return (EFAULT);
1579 	totwrt = size / sectorsize;
1580 	blkno = dbtob(blkno) / sectorsize;	/* blkno in DEV_BSIZE units */
1581 
1582 	nsects = lp->d_partitions[part].p_size;
1583 	sectoff = lp->d_partitions[part].p_offset;
1584 
1585 	/* Check transfer bounds against partition size. */
1586 	if ((blkno < 0) || ((blkno + totwrt) > nsects))
1587 		return (EINVAL);
1588 
1589 	/* Offset block number to start of partition. */
1590 	blkno += sectoff;
1591 
1592 	xs = &sx;
1593 
1594 	while (totwrt > 0) {
1595 		nwrt = totwrt;		/* XXX */
1596 #ifndef	SD_DUMP_NOT_TRUSTED
1597 		/*
1598 		 *  Fill out the scsi command
1599 		 */
1600 		memset(&cmd, 0, sizeof(cmd));
1601 		cmd.opcode = WRITE_10;
1602 		_lto4b(blkno, cmd.addr);
1603 		_lto2b(nwrt, cmd.length);
1604 		/*
1605 		 * Fill out the scsipi_xfer structure
1606 		 *    Note: we cannot sleep as we may be an interrupt
1607 		 * don't use scsipi_command() as it may want to wait
1608 		 * for an xs.
1609 		 */
1610 		memset(xs, 0, sizeof(sx));
1611 		xs->xs_control |= XS_CTL_NOSLEEP | XS_CTL_POLL |
1612 		    XS_CTL_DATA_OUT;
1613 		xs->xs_status = 0;
1614 		xs->xs_periph = periph;
1615 		xs->xs_retries = SDRETRIES;
1616 		xs->timeout = 10000;	/* 10000 millisecs for a disk ! */
1617 		xs->cmd = (struct scsipi_generic *)&cmd;
1618 		xs->cmdlen = sizeof(cmd);
1619 		xs->resid = nwrt * sectorsize;
1620 		xs->error = XS_NOERROR;
1621 		xs->bp = 0;
1622 		xs->data = va;
1623 		xs->datalen = nwrt * sectorsize;
1624 		callout_init(&xs->xs_callout, 0);
1625 
1626 		/*
1627 		 * Pass all this info to the scsi driver.
1628 		 */
1629 		scsipi_adapter_request(chan, ADAPTER_REQ_RUN_XFER, xs);
1630 		if ((xs->xs_status & XS_STS_DONE) == 0 ||
1631 		    xs->error != XS_NOERROR)
1632 			return (EIO);
1633 #else	/* SD_DUMP_NOT_TRUSTED */
1634 		/* Let's just talk about this first... */
1635 		printf("sd%d: dump addr 0x%x, blk %d\n", unit, va, blkno);
1636 		delay(500 * 1000);	/* half a second */
1637 #endif	/* SD_DUMP_NOT_TRUSTED */
1638 
1639 		/* update block count */
1640 		totwrt -= nwrt;
1641 		blkno += nwrt;
1642 		va = (char *)va + sectorsize * nwrt;
1643 	}
1644 	sddoingadump = 0;
1645 	return (0);
1646 }
1647 
1648 static int
1649 sd_mode_sense(struct sd_softc *sd, u_int8_t byte2, void *sense, size_t size,
1650     int page, int flags, int *big)
1651 {
1652 
1653 	if ((sd->sc_periph->periph_quirks & PQUIRK_ONLYBIG) &&
1654 	    !(sd->sc_periph->periph_quirks & PQUIRK_NOBIGMODESENSE)) {
1655 		*big = 1;
1656 		return scsipi_mode_sense_big(sd->sc_periph, byte2, page, sense,
1657 		    size + sizeof(struct scsi_mode_parameter_header_10),
1658 		    flags, SDRETRIES, 6000);
1659 	} else {
1660 		*big = 0;
1661 		return scsipi_mode_sense(sd->sc_periph, byte2, page, sense,
1662 		    size + sizeof(struct scsi_mode_parameter_header_6),
1663 		    flags, SDRETRIES, 6000);
1664 	}
1665 }
1666 
1667 static int
1668 sd_mode_select(struct sd_softc *sd, u_int8_t byte2, void *sense, size_t size,
1669     int flags, int big)
1670 {
1671 
1672 	if (big) {
1673 		struct scsi_mode_parameter_header_10 *header = sense;
1674 
1675 		_lto2b(0, header->data_length);
1676 		return scsipi_mode_select_big(sd->sc_periph, byte2, sense,
1677 		    size + sizeof(struct scsi_mode_parameter_header_10),
1678 		    flags, SDRETRIES, 6000);
1679 	} else {
1680 		struct scsi_mode_parameter_header_6 *header = sense;
1681 
1682 		header->data_length = 0;
1683 		return scsipi_mode_select(sd->sc_periph, byte2, sense,
1684 		    size + sizeof(struct scsi_mode_parameter_header_6),
1685 		    flags, SDRETRIES, 6000);
1686 	}
1687 }
1688 
1689 /*
1690  * sd_validate_blksize:
1691  *
1692  *	Validate the block size.  Print error if periph is specified,
1693  */
1694 static int
1695 sd_validate_blksize(struct scsipi_periph *periph, int len)
1696 {
1697 
1698 	switch (len) {
1699 	case 256:
1700 	case 512:
1701 	case 1024:
1702 	case 2048:
1703 	case 4096:
1704 		return 1;
1705 	}
1706 
1707 	if (periph) {
1708 		scsipi_printaddr(periph);
1709 		printf("%s sector size: 0x%x.  Defaulting to %d bytes.\n",
1710 		    (len ^ (1 << (ffs(len) - 1))) ?
1711 		    "preposterous" : "unsupported",
1712 		    len, SD_DEFAULT_BLKSIZE);
1713 	}
1714 
1715 	return 0;
1716 }
1717 
1718 /*
1719  * sd_read_capacity:
1720  *
1721  *	Find out from the device what its capacity is.
1722  */
1723 static u_int64_t
1724 sd_read_capacity(struct scsipi_periph *periph, int *blksize, int flags)
1725 {
1726 	union {
1727 		struct scsipi_read_capacity_10 cmd;
1728 		struct scsipi_read_capacity_16 cmd16;
1729 	} cmd;
1730 	union {
1731 		struct scsipi_read_capacity_10_data data;
1732 		struct scsipi_read_capacity_16_data data16;
1733 	} *datap;
1734 	uint64_t rv;
1735 
1736 	memset(&cmd, 0, sizeof(cmd));
1737 	cmd.cmd.opcode = READ_CAPACITY_10;
1738 
1739 	/*
1740 	 * Don't allocate data buffer on stack;
1741 	 * The lower driver layer might use the same stack and
1742 	 * if it uses region which is in the same cacheline,
1743 	 * cache flush ops against the data buffer won't work properly.
1744 	 */
1745 	datap = malloc(sizeof(*datap), M_TEMP, M_WAITOK);
1746 	if (datap == NULL)
1747 		return 0;
1748 
1749 	/*
1750 	 * If the command works, interpret the result as a 4 byte
1751 	 * number of blocks
1752 	 */
1753 	rv = 0;
1754 	memset(datap, 0, sizeof(datap->data));
1755 	if (scsipi_command(periph, (void *)&cmd.cmd, sizeof(cmd.cmd),
1756 	    (void *)datap, sizeof(datap->data), SCSIPIRETRIES, 20000, NULL,
1757 	    flags | XS_CTL_DATA_IN | XS_CTL_SILENT) != 0)
1758 		goto out;
1759 
1760 	if (_4btol(datap->data.addr) != 0xffffffff) {
1761 		*blksize = _4btol(datap->data.length);
1762 		rv = _4btol(datap->data.addr) + 1;
1763 		goto out;
1764 	}
1765 
1766 	/*
1767 	 * Device is larger than can be reflected by READ CAPACITY (10).
1768 	 * Try READ CAPACITY (16).
1769 	 */
1770 
1771 	memset(&cmd, 0, sizeof(cmd));
1772 	cmd.cmd16.opcode = READ_CAPACITY_16;
1773 	cmd.cmd16.byte2 = SRC16_SERVICE_ACTION;
1774 	_lto4b(sizeof(datap->data16), cmd.cmd16.len);
1775 
1776 	memset(datap, 0, sizeof(datap->data16));
1777 	if (scsipi_command(periph, (void *)&cmd.cmd16, sizeof(cmd.cmd16),
1778 	    (void *)datap, sizeof(datap->data16), SCSIPIRETRIES, 20000, NULL,
1779 	    flags | XS_CTL_DATA_IN | XS_CTL_SILENT) != 0)
1780 		goto out;
1781 
1782 	*blksize = _4btol(datap->data16.length);
1783 	rv = _8btol(datap->data16.addr) + 1;
1784 
1785  out:
1786 	free(datap, M_TEMP);
1787 	return rv;
1788 }
1789 
1790 static int
1791 sd_get_simplifiedparms(struct sd_softc *sd, struct disk_parms *dp, int flags)
1792 {
1793 	struct {
1794 		struct scsi_mode_parameter_header_6 header;
1795 		/* no block descriptor */
1796 		u_int8_t pg_code; /* page code (should be 6) */
1797 		u_int8_t pg_length; /* page length (should be 11) */
1798 		u_int8_t wcd; /* bit0: cache disable */
1799 		u_int8_t lbs[2]; /* logical block size */
1800 		u_int8_t size[5]; /* number of log. blocks */
1801 		u_int8_t pp; /* power/performance */
1802 		u_int8_t flags;
1803 		u_int8_t resvd;
1804 	} scsipi_sense;
1805 	u_int64_t blocks;
1806 	int error, blksize;
1807 
1808 	/*
1809 	 * sd_read_capacity (ie "read capacity") and mode sense page 6
1810 	 * give the same information. Do both for now, and check
1811 	 * for consistency.
1812 	 * XXX probably differs for removable media
1813 	 */
1814 	dp->blksize = SD_DEFAULT_BLKSIZE;
1815 	if ((blocks = sd_read_capacity(sd->sc_periph, &blksize, flags)) == 0)
1816 		return (SDGP_RESULT_OFFLINE);		/* XXX? */
1817 
1818 	error = scsipi_mode_sense(sd->sc_periph, SMS_DBD, 6,
1819 	    &scsipi_sense.header, sizeof(scsipi_sense),
1820 	    flags, SDRETRIES, 6000);
1821 
1822 	if (error != 0)
1823 		return (SDGP_RESULT_OFFLINE);		/* XXX? */
1824 
1825 	dp->blksize = blksize;
1826 	if (!sd_validate_blksize(NULL, dp->blksize))
1827 		dp->blksize = _2btol(scsipi_sense.lbs);
1828 	if (!sd_validate_blksize(sd->sc_periph, dp->blksize))
1829 		dp->blksize = SD_DEFAULT_BLKSIZE;
1830 
1831 	/*
1832 	 * Create a pseudo-geometry.
1833 	 */
1834 	dp->heads = 64;
1835 	dp->sectors = 32;
1836 	dp->cyls = blocks / (dp->heads * dp->sectors);
1837 	dp->disksize = _5btol(scsipi_sense.size);
1838 	if (dp->disksize <= UINT32_MAX && dp->disksize != blocks) {
1839 		printf("RBC size: mode sense=%llu, get cap=%llu\n",
1840 		       (unsigned long long)dp->disksize,
1841 		       (unsigned long long)blocks);
1842 		dp->disksize = blocks;
1843 	}
1844 	dp->disksize512 = (dp->disksize * dp->blksize) / DEV_BSIZE;
1845 
1846 	return (SDGP_RESULT_OK);
1847 }
1848 
1849 /*
1850  * Get the scsi driver to send a full inquiry to the * device and use the
1851  * results to fill out the disk parameter structure.
1852  */
1853 static int
1854 sd_get_capacity(struct sd_softc *sd, struct disk_parms *dp, int flags)
1855 {
1856 	u_int64_t blocks;
1857 	int error, blksize;
1858 #if 0
1859 	int i;
1860 	u_int8_t *p;
1861 #endif
1862 
1863 	dp->disksize = blocks = sd_read_capacity(sd->sc_periph, &blksize,
1864 	    flags);
1865 	if (blocks == 0) {
1866 		struct scsipi_read_format_capacities cmd;
1867 		struct {
1868 			struct scsipi_capacity_list_header header;
1869 			struct scsipi_capacity_descriptor desc;
1870 		} __packed data;
1871 
1872 		memset(&cmd, 0, sizeof(cmd));
1873 		memset(&data, 0, sizeof(data));
1874 		cmd.opcode = READ_FORMAT_CAPACITIES;
1875 		_lto2b(sizeof(data), cmd.length);
1876 
1877 		error = scsipi_command(sd->sc_periph,
1878 		    (void *)&cmd, sizeof(cmd), (void *)&data, sizeof(data),
1879 		    SDRETRIES, 20000, NULL,
1880 		    flags | XS_CTL_DATA_IN);
1881 		if (error == EFTYPE) {
1882 			/* Medium Format Corrupted, handle as not formatted */
1883 			return (SDGP_RESULT_UNFORMATTED);
1884 		}
1885 		if (error || data.header.length == 0)
1886 			return (SDGP_RESULT_OFFLINE);
1887 
1888 #if 0
1889 printf("rfc: length=%d\n", data.header.length);
1890 printf("rfc result:"); for (i = sizeof(struct scsipi_capacity_list_header) + data.header.length, p = (void *)&data; i; i--, p++) printf(" %02x", *p); printf("\n");
1891 #endif
1892 		switch (data.desc.byte5 & SCSIPI_CAP_DESC_CODE_MASK) {
1893 		case SCSIPI_CAP_DESC_CODE_RESERVED:
1894 		case SCSIPI_CAP_DESC_CODE_FORMATTED:
1895 			break;
1896 
1897 		case SCSIPI_CAP_DESC_CODE_UNFORMATTED:
1898 			return (SDGP_RESULT_UNFORMATTED);
1899 
1900 		case SCSIPI_CAP_DESC_CODE_NONE:
1901 			return (SDGP_RESULT_OFFLINE);
1902 		}
1903 
1904 		dp->disksize = blocks = _4btol(data.desc.nblks);
1905 		if (blocks == 0)
1906 			return (SDGP_RESULT_OFFLINE);		/* XXX? */
1907 
1908 		blksize = _3btol(data.desc.blklen);
1909 
1910 	} else if (!sd_validate_blksize(NULL, blksize)) {
1911 		struct sd_mode_sense_data scsipi_sense;
1912 		int big, bsize;
1913 		struct scsi_general_block_descriptor *bdesc;
1914 
1915 		memset(&scsipi_sense, 0, sizeof(scsipi_sense));
1916 		error = sd_mode_sense(sd, 0, &scsipi_sense,
1917 		    sizeof(scsipi_sense.blk_desc), 0, flags | XS_CTL_SILENT, &big);
1918 		if (!error) {
1919 			if (big) {
1920 				bdesc = (void *)(&scsipi_sense.header.big + 1);
1921 				bsize = _2btol(scsipi_sense.header.big.blk_desc_len);
1922 			} else {
1923 				bdesc = (void *)(&scsipi_sense.header.small + 1);
1924 				bsize = scsipi_sense.header.small.blk_desc_len;
1925 			}
1926 
1927 #if 0
1928 printf("page 0 sense:"); for (i = sizeof(scsipi_sense), p = (void *)&scsipi_sense; i; i--, p++) printf(" %02x", *p); printf("\n");
1929 printf("page 0 bsize=%d\n", bsize);
1930 printf("page 0 ok\n");
1931 #endif
1932 
1933 			if (bsize >= 8) {
1934 				blksize = _3btol(bdesc->blklen);
1935 			}
1936 		}
1937 	}
1938 
1939 	if (!sd_validate_blksize(sd->sc_periph, blksize))
1940 		blksize = SD_DEFAULT_BLKSIZE;
1941 
1942 	dp->blksize = blksize;
1943 	dp->disksize512 = (blocks * dp->blksize) / DEV_BSIZE;
1944 	return (0);
1945 }
1946 
1947 static int
1948 sd_get_parms_page4(struct sd_softc *sd, struct disk_parms *dp, int flags)
1949 {
1950 	struct sd_mode_sense_data scsipi_sense;
1951 	int error;
1952 	int big, byte2;
1953 	size_t poffset;
1954 	union scsi_disk_pages *pages;
1955 
1956 	byte2 = SMS_DBD;
1957 again:
1958 	memset(&scsipi_sense, 0, sizeof(scsipi_sense));
1959 	error = sd_mode_sense(sd, byte2, &scsipi_sense,
1960 	    (byte2 ? 0 : sizeof(scsipi_sense.blk_desc)) +
1961 	    sizeof(scsipi_sense.pages.rigid_geometry), 4,
1962 	    flags | XS_CTL_SILENT, &big);
1963 	if (error) {
1964 		if (byte2 == SMS_DBD) {
1965 			/* No result; try once more with DBD off */
1966 			byte2 = 0;
1967 			goto again;
1968 		}
1969 		return (error);
1970 	}
1971 
1972 	if (big) {
1973 		poffset = sizeof scsipi_sense.header.big;
1974 		poffset += _2btol(scsipi_sense.header.big.blk_desc_len);
1975 	} else {
1976 		poffset = sizeof scsipi_sense.header.small;
1977 		poffset += scsipi_sense.header.small.blk_desc_len;
1978 	}
1979 
1980 	if (poffset > sizeof(scsipi_sense) - sizeof(pages->rigid_geometry))
1981 		return ERESTART;
1982 
1983 	pages = (void *)((u_long)&scsipi_sense + poffset);
1984 #if 0
1985 	{
1986 		size_t i;
1987 		u_int8_t *p;
1988 
1989 		printf("page 4 sense:");
1990 		for (i = sizeof(scsipi_sense), p = (void *)&scsipi_sense; i;
1991 		    i--, p++)
1992 			printf(" %02x", *p);
1993 		printf("\n");
1994 		printf("page 4 pg_code=%d sense=%p/%p\n",
1995 		    pages->rigid_geometry.pg_code, &scsipi_sense, pages);
1996 	}
1997 #endif
1998 
1999 	if ((pages->rigid_geometry.pg_code & PGCODE_MASK) != 4)
2000 		return (ERESTART);
2001 
2002 	SC_DEBUG(sd->sc_periph, SCSIPI_DB3,
2003 	    ("%d cyls, %d heads, %d precomp, %d red_write, %d land_zone\n",
2004 	    _3btol(pages->rigid_geometry.ncyl),
2005 	    pages->rigid_geometry.nheads,
2006 	    _2btol(pages->rigid_geometry.st_cyl_wp),
2007 	    _2btol(pages->rigid_geometry.st_cyl_rwc),
2008 	    _2btol(pages->rigid_geometry.land_zone)));
2009 
2010 	/*
2011 	 * KLUDGE!! (for zone recorded disks)
2012 	 * give a number of sectors so that sec * trks * cyls
2013 	 * is <= disk_size
2014 	 * can lead to wasted space! THINK ABOUT THIS !
2015 	 */
2016 	dp->heads = pages->rigid_geometry.nheads;
2017 	dp->cyls = _3btol(pages->rigid_geometry.ncyl);
2018 	if (dp->heads == 0 || dp->cyls == 0)
2019 		return (ERESTART);
2020 	dp->sectors = dp->disksize / (dp->heads * dp->cyls);	/* XXX */
2021 
2022 	dp->rot_rate = _2btol(pages->rigid_geometry.rpm);
2023 	if (dp->rot_rate == 0)
2024 		dp->rot_rate = 3600;
2025 
2026 #if 0
2027 printf("page 4 ok\n");
2028 #endif
2029 	return (0);
2030 }
2031 
2032 static int
2033 sd_get_parms_page5(struct sd_softc *sd, struct disk_parms *dp, int flags)
2034 {
2035 	struct sd_mode_sense_data scsipi_sense;
2036 	int error;
2037 	int big, byte2;
2038 	size_t poffset;
2039 	union scsi_disk_pages *pages;
2040 
2041 	byte2 = SMS_DBD;
2042 again:
2043 	memset(&scsipi_sense, 0, sizeof(scsipi_sense));
2044 	error = sd_mode_sense(sd, 0, &scsipi_sense,
2045 	    (byte2 ? 0 : sizeof(scsipi_sense.blk_desc)) +
2046 	    sizeof(scsipi_sense.pages.flex_geometry), 5,
2047 	    flags | XS_CTL_SILENT, &big);
2048 	if (error) {
2049 		if (byte2 == SMS_DBD) {
2050 			/* No result; try once more with DBD off */
2051 			byte2 = 0;
2052 			goto again;
2053 		}
2054 		return (error);
2055 	}
2056 
2057 	if (big) {
2058 		poffset = sizeof scsipi_sense.header.big;
2059 		poffset += _2btol(scsipi_sense.header.big.blk_desc_len);
2060 	} else {
2061 		poffset = sizeof scsipi_sense.header.small;
2062 		poffset += scsipi_sense.header.small.blk_desc_len;
2063 	}
2064 
2065 	if (poffset > sizeof(scsipi_sense) - sizeof(pages->flex_geometry))
2066 		return ERESTART;
2067 
2068 	pages = (void *)((u_long)&scsipi_sense + poffset);
2069 #if 0
2070 	{
2071 		size_t i;
2072 		u_int8_t *p;
2073 
2074 		printf("page 5 sense:");
2075 		for (i = sizeof(scsipi_sense), p = (void *)&scsipi_sense; i;
2076 		    i--, p++)
2077 			printf(" %02x", *p);
2078 		printf("\n");
2079 		printf("page 5 pg_code=%d sense=%p/%p\n",
2080 		    pages->flex_geometry.pg_code, &scsipi_sense, pages);
2081 	}
2082 #endif
2083 
2084 	if ((pages->flex_geometry.pg_code & PGCODE_MASK) != 5)
2085 		return (ERESTART);
2086 
2087 	SC_DEBUG(sd->sc_periph, SCSIPI_DB3,
2088 	    ("%d cyls, %d heads, %d sec, %d bytes/sec\n",
2089 	    _3btol(pages->flex_geometry.ncyl),
2090 	    pages->flex_geometry.nheads,
2091 	    pages->flex_geometry.ph_sec_tr,
2092 	    _2btol(pages->flex_geometry.bytes_s)));
2093 
2094 	dp->heads = pages->flex_geometry.nheads;
2095 	dp->cyls = _2btol(pages->flex_geometry.ncyl);
2096 	dp->sectors = pages->flex_geometry.ph_sec_tr;
2097 	if (dp->heads == 0 || dp->cyls == 0 || dp->sectors == 0)
2098 		return (ERESTART);
2099 
2100 	dp->rot_rate = _2btol(pages->rigid_geometry.rpm);
2101 	if (dp->rot_rate == 0)
2102 		dp->rot_rate = 3600;
2103 
2104 #if 0
2105 printf("page 5 ok\n");
2106 #endif
2107 	return (0);
2108 }
2109 
2110 static int
2111 sd_get_parms(struct sd_softc *sd, struct disk_parms *dp, int flags)
2112 {
2113 	int error;
2114 
2115 	/*
2116 	 * If offline, the SDEV_MEDIA_LOADED flag will be
2117 	 * cleared by the caller if necessary.
2118 	 */
2119 	if (sd->type == T_SIMPLE_DIRECT) {
2120 		error = sd_get_simplifiedparms(sd, dp, flags);
2121 		if (!error)
2122 			disk_blocksize(&sd->sc_dk, dp->blksize);
2123 		return (error);
2124 	}
2125 
2126 	error = sd_get_capacity(sd, dp, flags);
2127 	if (error)
2128 		return (error);
2129 
2130 	disk_blocksize(&sd->sc_dk, dp->blksize);
2131 
2132 	if (sd->type == T_OPTICAL)
2133 		goto page0;
2134 
2135 	if (sd->sc_periph->periph_flags & PERIPH_REMOVABLE) {
2136 		if (!sd_get_parms_page5(sd, dp, flags) ||
2137 		    !sd_get_parms_page4(sd, dp, flags))
2138 			goto setprops;
2139 	} else {
2140 		if (!sd_get_parms_page4(sd, dp, flags) ||
2141 		    !sd_get_parms_page5(sd, dp, flags))
2142 			goto setprops;
2143 	}
2144 
2145 page0:
2146 	printf("%s: fabricating a geometry\n", device_xname(sd->sc_dev));
2147 	/* Try calling driver's method for figuring out geometry. */
2148 	if (!sd->sc_periph->periph_channel->chan_adapter->adapt_getgeom ||
2149 	    !(*sd->sc_periph->periph_channel->chan_adapter->adapt_getgeom)
2150 		(sd->sc_periph, dp, dp->disksize)) {
2151 		/*
2152 		 * Use adaptec standard fictitious geometry
2153 		 * this depends on which controller (e.g. 1542C is
2154 		 * different. but we have to put SOMETHING here..)
2155 		 */
2156 		dp->heads = 64;
2157 		dp->sectors = 32;
2158 		dp->cyls = dp->disksize / (64 * 32);
2159 	}
2160 	dp->rot_rate = 3600;
2161 
2162 setprops:
2163 	sd_set_geometry(sd);
2164 
2165 	return (SDGP_RESULT_OK);
2166 }
2167 
2168 static int
2169 sd_flush(struct sd_softc *sd, int flags)
2170 {
2171 	struct scsipi_periph *periph = sd->sc_periph;
2172 	struct scsi_synchronize_cache_10 cmd;
2173 
2174 	/*
2175 	 * If the device is SCSI-2, issue a SYNCHRONIZE CACHE.
2176 	 * We issue with address 0 length 0, which should be
2177 	 * interpreted by the device as "all remaining blocks
2178 	 * starting at address 0".  We ignore ILLEGAL REQUEST
2179 	 * in the event that the command is not supported by
2180 	 * the device, and poll for completion so that we know
2181 	 * that the cache has actually been flushed.
2182 	 *
2183 	 * Unless, that is, the device can't handle the SYNCHRONIZE CACHE
2184 	 * command, as indicated by our quirks flags.
2185 	 *
2186 	 * XXX What about older devices?
2187 	 */
2188 	if (periph->periph_version < 2 ||
2189 	    (periph->periph_quirks & PQUIRK_NOSYNCCACHE))
2190 		return (0);
2191 
2192 	sd->flags |= SDF_FLUSHING;
2193 	memset(&cmd, 0, sizeof(cmd));
2194 	cmd.opcode = SCSI_SYNCHRONIZE_CACHE_10;
2195 
2196 	return (scsipi_command(periph, (void *)&cmd, sizeof(cmd), 0, 0,
2197 	    SDRETRIES, 100000, NULL, flags | XS_CTL_IGNORE_ILLEGAL_REQUEST));
2198 }
2199 
2200 static int
2201 sd_getcache(struct sd_softc *sd, int *bitsp)
2202 {
2203 	struct scsipi_periph *periph = sd->sc_periph;
2204 	struct sd_mode_sense_data scsipi_sense;
2205 	int error, bits = 0;
2206 	int big;
2207 	union scsi_disk_pages *pages;
2208 
2209 	if (periph->periph_version < 2)
2210 		return (EOPNOTSUPP);
2211 
2212 	memset(&scsipi_sense, 0, sizeof(scsipi_sense));
2213 	error = sd_mode_sense(sd, SMS_DBD, &scsipi_sense,
2214 	    sizeof(scsipi_sense.pages.caching_params), 8, 0, &big);
2215 	if (error)
2216 		return (error);
2217 
2218 	if (big)
2219 		pages = (void *)(&scsipi_sense.header.big + 1);
2220 	else
2221 		pages = (void *)(&scsipi_sense.header.small + 1);
2222 
2223 	if ((pages->caching_params.flags & CACHING_RCD) == 0)
2224 		bits |= DKCACHE_READ;
2225 	if (pages->caching_params.flags & CACHING_WCE)
2226 		bits |= DKCACHE_WRITE;
2227 	if (pages->caching_params.pg_code & PGCODE_PS)
2228 		bits |= DKCACHE_SAVE;
2229 
2230 	memset(&scsipi_sense, 0, sizeof(scsipi_sense));
2231 	error = sd_mode_sense(sd, SMS_DBD, &scsipi_sense,
2232 	    sizeof(scsipi_sense.pages.caching_params),
2233 	    SMS_PCTRL_CHANGEABLE|8, 0, &big);
2234 	if (error == 0) {
2235 		if (big)
2236 			pages = (void *)(&scsipi_sense.header.big + 1);
2237 		else
2238 			pages = (void *)(&scsipi_sense.header.small + 1);
2239 
2240 		if (pages->caching_params.flags & CACHING_RCD)
2241 			bits |= DKCACHE_RCHANGE;
2242 		if (pages->caching_params.flags & CACHING_WCE)
2243 			bits |= DKCACHE_WCHANGE;
2244 	}
2245 
2246 	*bitsp = bits;
2247 
2248 	return (0);
2249 }
2250 
2251 static int
2252 sd_setcache(struct sd_softc *sd, int bits)
2253 {
2254 	struct scsipi_periph *periph = sd->sc_periph;
2255 	struct sd_mode_sense_data scsipi_sense;
2256 	int error;
2257 	uint8_t oflags, byte2 = 0;
2258 	int big;
2259 	union scsi_disk_pages *pages;
2260 
2261 	if (periph->periph_version < 2)
2262 		return (EOPNOTSUPP);
2263 
2264 	memset(&scsipi_sense, 0, sizeof(scsipi_sense));
2265 	error = sd_mode_sense(sd, SMS_DBD, &scsipi_sense,
2266 	    sizeof(scsipi_sense.pages.caching_params), 8, 0, &big);
2267 	if (error)
2268 		return (error);
2269 
2270 	if (big)
2271 		pages = (void *)(&scsipi_sense.header.big + 1);
2272 	else
2273 		pages = (void *)(&scsipi_sense.header.small + 1);
2274 
2275 	oflags = pages->caching_params.flags;
2276 
2277 	if (bits & DKCACHE_READ)
2278 		pages->caching_params.flags &= ~CACHING_RCD;
2279 	else
2280 		pages->caching_params.flags |= CACHING_RCD;
2281 
2282 	if (bits & DKCACHE_WRITE)
2283 		pages->caching_params.flags |= CACHING_WCE;
2284 	else
2285 		pages->caching_params.flags &= ~CACHING_WCE;
2286 
2287 	if (oflags == pages->caching_params.flags)
2288 		return (0);
2289 
2290 	pages->caching_params.pg_code &= PGCODE_MASK;
2291 
2292 	if (bits & DKCACHE_SAVE)
2293 		byte2 |= SMS_SP;
2294 
2295 	return (sd_mode_select(sd, byte2|SMS_PF, &scsipi_sense,
2296 	    sizeof(struct scsi_mode_page_header) +
2297 	    pages->caching_params.pg_length, 0, big));
2298 }
2299 
2300 static void
2301 sd_set_geometry(struct sd_softc *sd)
2302 {
2303 	struct disk_geom *dg = &sd->sc_dk.dk_geom;
2304 
2305 	memset(dg, 0, sizeof(*dg));
2306 
2307 	dg->dg_secperunit = sd->params.disksize;
2308 	dg->dg_secsize = sd->params.blksize;
2309 	dg->dg_nsectors = sd->params.sectors;
2310 	dg->dg_ntracks = sd->params.heads;
2311 	dg->dg_ncylinders = sd->params.cyls;
2312 
2313 	disk_set_info(sd->sc_dev, &sd->sc_dk, NULL);
2314 }
2315