hpsa.c 130.7 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
/*
 *    Disk Array driver for HP Smart Array SAS controllers
 *    Copyright 2000, 2009 Hewlett-Packard Development Company, L.P.
 *
 *    This program is free software; you can redistribute it and/or modify
 *    it under the terms of the GNU General Public License as published by
 *    the Free Software Foundation; version 2 of the License.
 *
 *    This program is distributed in the hope that it will be useful,
 *    but WITHOUT ANY WARRANTY; without even the implied warranty of
 *    MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
 *    NON INFRINGEMENT.  See the GNU General Public License for more details.
 *
 *    You should have received a copy of the GNU General Public License
 *    along with this program; if not, write to the Free Software
 *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 *
 *    Questions/Comments/Bugfixes to iss_storagedev@hp.com
 *
 */

#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/types.h>
#include <linux/pci.h>
M
Matthew Garrett 已提交
26
#include <linux/pci-aspm.h>
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/delay.h>
#include <linux/fs.h>
#include <linux/timer.h>
#include <linux/seq_file.h>
#include <linux/init.h>
#include <linux/spinlock.h>
#include <linux/compat.h>
#include <linux/blktrace_api.h>
#include <linux/uaccess.h>
#include <linux/io.h>
#include <linux/dma-mapping.h>
#include <linux/completion.h>
#include <linux/moduleparam.h>
#include <scsi/scsi.h>
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_host.h>
46
#include <scsi/scsi_tcq.h>
47 48 49
#include <linux/cciss_ioctl.h>
#include <linux/string.h>
#include <linux/bitmap.h>
A
Arun Sharma 已提交
50
#include <linux/atomic.h>
51
#include <linux/kthread.h>
52
#include <linux/jiffies.h>
53 54 55 56
#include "hpsa_cmd.h"
#include "hpsa.h"

/* HPSA_DRIVER_VERSION must be 3 byte values (0-255) separated by '.' */
M
Mike Miller 已提交
57
#define HPSA_DRIVER_VERSION "2.0.2-1"
58
#define DRIVER_NAME "HP HPSA Driver (v " HPSA_DRIVER_VERSION ")"
59
#define HPSA "hpsa"
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

/* How long to wait (in milliseconds) for board to go into simple mode */
#define MAX_CONFIG_WAIT 30000
#define MAX_IOCTL_CONFIG_WAIT 1000

/*define how many times we will try a command because of bus resets */
#define MAX_CMD_RETRIES 3

/* Embedded module documentation macros - see modules.h */
MODULE_AUTHOR("Hewlett-Packard Company");
MODULE_DESCRIPTION("Driver for HP Smart Array Controller version " \
	HPSA_DRIVER_VERSION);
MODULE_SUPPORTED_DEVICE("HP Smart Array Controllers");
MODULE_VERSION(HPSA_DRIVER_VERSION);
MODULE_LICENSE("GPL");

static int hpsa_allow_any;
module_param(hpsa_allow_any, int, S_IRUGO|S_IWUSR);
MODULE_PARM_DESC(hpsa_allow_any,
		"Allow hpsa driver to access unknown HP Smart Array hardware");
80 81 82 83
static int hpsa_simple_mode;
module_param(hpsa_simple_mode, int, S_IRUGO|S_IWUSR);
MODULE_PARM_DESC(hpsa_simple_mode,
	"Use 'simple mode' rather than 'performant mode'");
84 85 86 87 88 89 90 91 92 93

/* define the PCI info for the cards we can control */
static const struct pci_device_id hpsa_pci_device_id[] = {
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x3241},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x3243},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x3245},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x3247},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x3249},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x324a},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x324b},
94
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSE,     0x103C, 0x3233},
95 96 97 98 99 100 101
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3350},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3351},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3352},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3353},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3354},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3355},
	{PCI_VENDOR_ID_HP,     PCI_DEVICE_ID_HP_CISSF,     0x103C, 0x3356},
102
	{PCI_VENDOR_ID_HP,     PCI_ANY_ID,	PCI_ANY_ID, PCI_ANY_ID,
103
		PCI_CLASS_STORAGE_RAID << 8, 0xffff << 8, 0},
104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
	{0,}
};

MODULE_DEVICE_TABLE(pci, hpsa_pci_device_id);

/*  board_id = Subsystem Device ID & Vendor ID
 *  product = Marketing Name for the board
 *  access = Address of the struct of function pointers
 */
static struct board_type products[] = {
	{0x3241103C, "Smart Array P212", &SA5_access},
	{0x3243103C, "Smart Array P410", &SA5_access},
	{0x3245103C, "Smart Array P410i", &SA5_access},
	{0x3247103C, "Smart Array P411", &SA5_access},
	{0x3249103C, "Smart Array P812", &SA5_access},
	{0x324a103C, "Smart Array P712m", &SA5_access},
	{0x324b103C, "Smart Array P711m", &SA5_access},
121 122 123 124 125 126 127
	{0x3350103C, "Smart Array", &SA5_access},
	{0x3351103C, "Smart Array", &SA5_access},
	{0x3352103C, "Smart Array", &SA5_access},
	{0x3353103C, "Smart Array", &SA5_access},
	{0x3354103C, "Smart Array", &SA5_access},
	{0x3355103C, "Smart Array", &SA5_access},
	{0x3356103C, "Smart Array", &SA5_access},
128 129 130 131 132
	{0xFFFF103C, "Unknown Smart Array", &SA5_access},
};

static int number_of_controllers;

133 134 135 136
static struct list_head hpsa_ctlr_list = LIST_HEAD_INIT(hpsa_ctlr_list);
static spinlock_t lockup_detector_lock;
static struct task_struct *hpsa_lockup_detector;

137 138
static irqreturn_t do_hpsa_intr_intx(int irq, void *dev_id);
static irqreturn_t do_hpsa_intr_msi(int irq, void *dev_id);
139 140 141 142 143 144 145 146 147 148 149
static int hpsa_ioctl(struct scsi_device *dev, int cmd, void *arg);
static void start_io(struct ctlr_info *h);

#ifdef CONFIG_COMPAT
static int hpsa_compat_ioctl(struct scsi_device *dev, int cmd, void *arg);
#endif

static void cmd_free(struct ctlr_info *h, struct CommandList *c);
static void cmd_special_free(struct ctlr_info *h, struct CommandList *c);
static struct CommandList *cmd_alloc(struct ctlr_info *h);
static struct CommandList *cmd_special_alloc(struct ctlr_info *h);
150 151
static void fill_cmd(struct CommandList *c, u8 cmd, struct ctlr_info *h,
	void *buff, size_t size, u8 page_code, unsigned char *scsi3addr,
152 153
	int cmd_type);

J
Jeff Garzik 已提交
154
static int hpsa_scsi_queue_command(struct Scsi_Host *h, struct scsi_cmnd *cmd);
155 156 157
static void hpsa_scan_start(struct Scsi_Host *);
static int hpsa_scan_finished(struct Scsi_Host *sh,
	unsigned long elapsed_time);
158 159
static int hpsa_change_queue_depth(struct scsi_device *sdev,
	int qdepth, int reason);
160 161 162 163 164 165 166 167 168 169

static int hpsa_eh_device_reset_handler(struct scsi_cmnd *scsicmd);
static int hpsa_slave_alloc(struct scsi_device *sdev);
static void hpsa_slave_destroy(struct scsi_device *sdev);

static void hpsa_update_scsi_devices(struct ctlr_info *h, int hostno);
static int check_for_unit_attention(struct ctlr_info *h,
	struct CommandList *c);
static void check_ioctl_unit_attention(struct ctlr_info *h,
	struct CommandList *c);
170 171 172
/* performant mode helper functions */
static void calc_bucket_map(int *bucket, int num_buckets,
	int nsgs, int *bucket_map);
173
static __devinit void hpsa_put_ctlr_into_performant_mode(struct ctlr_info *h);
174
static inline u32 next_command(struct ctlr_info *h);
175 176 177 178 179
static int __devinit hpsa_find_cfg_addrs(struct pci_dev *pdev,
	void __iomem *vaddr, u32 *cfg_base_addr, u64 *cfg_base_addr_index,
	u64 *cfg_offset);
static int __devinit hpsa_pci_find_memory_BAR(struct pci_dev *pdev,
	unsigned long *memory_bar);
180
static int __devinit hpsa_lookup_board_id(struct pci_dev *pdev, u32 *board_id);
181 182 183 184
static int __devinit hpsa_wait_for_board_state(struct pci_dev *pdev,
	void __iomem *vaddr, int wait_for_ready);
#define BOARD_NOT_READY 0
#define BOARD_READY 1
185 186 187 188 189 190 191

static inline struct ctlr_info *sdev_to_hba(struct scsi_device *sdev)
{
	unsigned long *priv = shost_priv(sdev->host);
	return (struct ctlr_info *) *priv;
}

192 193 194 195 196 197
static inline struct ctlr_info *shost_to_hba(struct Scsi_Host *sh)
{
	unsigned long *priv = shost_priv(sh);
	return (struct ctlr_info *) *priv;
}

198 199 200 201 202 203 204 205
static int check_for_unit_attention(struct ctlr_info *h,
	struct CommandList *c)
{
	if (c->err_info->SenseInfo[2] != UNIT_ATTENTION)
		return 0;

	switch (c->err_info->SenseInfo[12]) {
	case STATE_CHANGED:
206
		dev_warn(&h->pdev->dev, HPSA "%d: a state change "
207 208 209
			"detected, command retried\n", h->ctlr);
		break;
	case LUN_FAILED:
210
		dev_warn(&h->pdev->dev, HPSA "%d: LUN failure "
211 212 213
			"detected, action required\n", h->ctlr);
		break;
	case REPORT_LUNS_CHANGED:
214
		dev_warn(&h->pdev->dev, HPSA "%d: report LUN data "
M
Mike Miller 已提交
215
			"changed, action required\n", h->ctlr);
216
	/*
217 218
	 * Note: this REPORT_LUNS_CHANGED condition only occurs on the external
	 * target (array) devices.
219 220 221
	 */
		break;
	case POWER_OR_RESET:
222
		dev_warn(&h->pdev->dev, HPSA "%d: a power on "
223 224 225
			"or device reset detected\n", h->ctlr);
		break;
	case UNIT_ATTENTION_CLEARED:
226
		dev_warn(&h->pdev->dev, HPSA "%d: unit attention "
227 228 229
		    "cleared by another initiator\n", h->ctlr);
		break;
	default:
230
		dev_warn(&h->pdev->dev, HPSA "%d: unknown "
231 232 233 234 235 236
			"unit attention detected\n", h->ctlr);
		break;
	}
	return 1;
}

237 238 239 240 241 242 243 244 245 246
static int check_for_busy(struct ctlr_info *h, struct CommandList *c)
{
	if (c->err_info->CommandStatus != CMD_TARGET_STATUS ||
		(c->err_info->ScsiStatus != SAM_STAT_BUSY &&
		 c->err_info->ScsiStatus != SAM_STAT_TASK_SET_FULL))
		return 0;
	dev_warn(&h->pdev->dev, HPSA "device busy");
	return 1;
}

247 248 249 250 251 252
static ssize_t host_store_rescan(struct device *dev,
				 struct device_attribute *attr,
				 const char *buf, size_t count)
{
	struct ctlr_info *h;
	struct Scsi_Host *shost = class_to_shost(dev);
253
	h = shost_to_hba(shost);
M
Mike Miller 已提交
254
	hpsa_scan_start(h->scsi_host);
255 256 257
	return count;
}

258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
static ssize_t host_show_firmware_revision(struct device *dev,
	     struct device_attribute *attr, char *buf)
{
	struct ctlr_info *h;
	struct Scsi_Host *shost = class_to_shost(dev);
	unsigned char *fwrev;

	h = shost_to_hba(shost);
	if (!h->hba_inquiry_data)
		return 0;
	fwrev = &h->hba_inquiry_data[32];
	return snprintf(buf, 20, "%c%c%c%c\n",
		fwrev[0], fwrev[1], fwrev[2], fwrev[3]);
}

273 274 275 276 277 278 279 280 281
static ssize_t host_show_commands_outstanding(struct device *dev,
	     struct device_attribute *attr, char *buf)
{
	struct Scsi_Host *shost = class_to_shost(dev);
	struct ctlr_info *h = shost_to_hba(shost);

	return snprintf(buf, 20, "%d\n", h->commands_outstanding);
}

282 283 284 285 286 287 288 289
static ssize_t host_show_transport_mode(struct device *dev,
	struct device_attribute *attr, char *buf)
{
	struct ctlr_info *h;
	struct Scsi_Host *shost = class_to_shost(dev);

	h = shost_to_hba(shost);
	return snprintf(buf, 20, "%s\n",
290
		h->transMethod & CFGTBL_Trans_Performant ?
291 292 293
			"performant" : "simple");
}

294
/* List of controllers which cannot be hard reset on kexec with reset_devices */
295 296 297 298 299 300 301 302 303 304 305 306 307
static u32 unresettable_controller[] = {
	0x324a103C, /* Smart Array P712m */
	0x324b103C, /* SmartArray P711m */
	0x3223103C, /* Smart Array P800 */
	0x3234103C, /* Smart Array P400 */
	0x3235103C, /* Smart Array P400i */
	0x3211103C, /* Smart Array E200i */
	0x3212103C, /* Smart Array E200 */
	0x3213103C, /* Smart Array E200i */
	0x3214103C, /* Smart Array E200i */
	0x3215103C, /* Smart Array E200i */
	0x3237103C, /* Smart Array E500 */
	0x323D103C, /* Smart Array P700m */
308
	0x40800E11, /* Smart Array 5i */
309 310
	0x409C0E11, /* Smart Array 6400 */
	0x409D0E11, /* Smart Array 6400 EM */
311 312 313 314 315 316
	0x40700E11, /* Smart Array 5300 */
	0x40820E11, /* Smart Array 532 */
	0x40830E11, /* Smart Array 5312 */
	0x409A0E11, /* Smart Array 641 */
	0x409B0E11, /* Smart Array 642 */
	0x40910E11, /* Smart Array 6i */
317 318
};

319 320
/* List of controllers which cannot even be soft reset */
static u32 soft_unresettable_controller[] = {
321
	0x40800E11, /* Smart Array 5i */
322 323 324 325 326 327
	0x40700E11, /* Smart Array 5300 */
	0x40820E11, /* Smart Array 532 */
	0x40830E11, /* Smart Array 5312 */
	0x409A0E11, /* Smart Array 641 */
	0x409B0E11, /* Smart Array 642 */
	0x40910E11, /* Smart Array 6i */
328 329 330 331 332 333 334 335 336 337 338 339
	/* Exclude 640x boards.  These are two pci devices in one slot
	 * which share a battery backed cache module.  One controls the
	 * cache, the other accesses the cache through the one that controls
	 * it.  If we reset the one controlling the cache, the other will
	 * likely not be happy.  Just forbid resetting this conjoined mess.
	 * The 640x isn't really supported by hpsa anyway.
	 */
	0x409C0E11, /* Smart Array 6400 */
	0x409D0E11, /* Smart Array 6400 EM */
};

static int ctlr_is_hard_resettable(u32 board_id)
340 341 342 343
{
	int i;

	for (i = 0; i < ARRAY_SIZE(unresettable_controller); i++)
344 345 346 347 348 349 350 351 352 353 354
		if (unresettable_controller[i] == board_id)
			return 0;
	return 1;
}

static int ctlr_is_soft_resettable(u32 board_id)
{
	int i;

	for (i = 0; i < ARRAY_SIZE(soft_unresettable_controller); i++)
		if (soft_unresettable_controller[i] == board_id)
355 356 357 358
			return 0;
	return 1;
}

359 360 361 362 363 364
static int ctlr_is_resettable(u32 board_id)
{
	return ctlr_is_hard_resettable(board_id) ||
		ctlr_is_soft_resettable(board_id);
}

365 366 367 368 369 370 371
static ssize_t host_show_resettable(struct device *dev,
	struct device_attribute *attr, char *buf)
{
	struct ctlr_info *h;
	struct Scsi_Host *shost = class_to_shost(dev);

	h = shost_to_hba(shost);
372
	return snprintf(buf, 20, "%d\n", ctlr_is_resettable(h->board_id));
373 374
}

375 376 377 378 379 380 381 382 383 384 385 386 387 388
static inline int is_logical_dev_addr_mode(unsigned char scsi3addr[])
{
	return (scsi3addr[3] & 0xC0) == 0x40;
}

static const char *raid_label[] = { "0", "4", "1(1+0)", "5", "5+1", "ADG",
	"UNKNOWN"
};
#define RAID_UNKNOWN (ARRAY_SIZE(raid_label) - 1)

static ssize_t raid_level_show(struct device *dev,
	     struct device_attribute *attr, char *buf)
{
	ssize_t l = 0;
389
	unsigned char rlevel;
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
	struct ctlr_info *h;
	struct scsi_device *sdev;
	struct hpsa_scsi_dev_t *hdev;
	unsigned long flags;

	sdev = to_scsi_device(dev);
	h = sdev_to_hba(sdev);
	spin_lock_irqsave(&h->lock, flags);
	hdev = sdev->hostdata;
	if (!hdev) {
		spin_unlock_irqrestore(&h->lock, flags);
		return -ENODEV;
	}

	/* Is this even a logical drive? */
	if (!is_logical_dev_addr_mode(hdev->scsi3addr)) {
		spin_unlock_irqrestore(&h->lock, flags);
		l = snprintf(buf, PAGE_SIZE, "N/A\n");
		return l;
	}

	rlevel = hdev->raid_level;
	spin_unlock_irqrestore(&h->lock, flags);
413
	if (rlevel > RAID_UNKNOWN)
414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470
		rlevel = RAID_UNKNOWN;
	l = snprintf(buf, PAGE_SIZE, "RAID %s\n", raid_label[rlevel]);
	return l;
}

static ssize_t lunid_show(struct device *dev,
	     struct device_attribute *attr, char *buf)
{
	struct ctlr_info *h;
	struct scsi_device *sdev;
	struct hpsa_scsi_dev_t *hdev;
	unsigned long flags;
	unsigned char lunid[8];

	sdev = to_scsi_device(dev);
	h = sdev_to_hba(sdev);
	spin_lock_irqsave(&h->lock, flags);
	hdev = sdev->hostdata;
	if (!hdev) {
		spin_unlock_irqrestore(&h->lock, flags);
		return -ENODEV;
	}
	memcpy(lunid, hdev->scsi3addr, sizeof(lunid));
	spin_unlock_irqrestore(&h->lock, flags);
	return snprintf(buf, 20, "0x%02x%02x%02x%02x%02x%02x%02x%02x\n",
		lunid[0], lunid[1], lunid[2], lunid[3],
		lunid[4], lunid[5], lunid[6], lunid[7]);
}

static ssize_t unique_id_show(struct device *dev,
	     struct device_attribute *attr, char *buf)
{
	struct ctlr_info *h;
	struct scsi_device *sdev;
	struct hpsa_scsi_dev_t *hdev;
	unsigned long flags;
	unsigned char sn[16];

	sdev = to_scsi_device(dev);
	h = sdev_to_hba(sdev);
	spin_lock_irqsave(&h->lock, flags);
	hdev = sdev->hostdata;
	if (!hdev) {
		spin_unlock_irqrestore(&h->lock, flags);
		return -ENODEV;
	}
	memcpy(sn, hdev->device_id, sizeof(sn));
	spin_unlock_irqrestore(&h->lock, flags);
	return snprintf(buf, 16 * 2 + 2,
			"%02X%02X%02X%02X%02X%02X%02X%02X"
			"%02X%02X%02X%02X%02X%02X%02X%02X\n",
			sn[0], sn[1], sn[2], sn[3],
			sn[4], sn[5], sn[6], sn[7],
			sn[8], sn[9], sn[10], sn[11],
			sn[12], sn[13], sn[14], sn[15]);
}

471 472 473 474 475 476 477 478 479 480
static DEVICE_ATTR(raid_level, S_IRUGO, raid_level_show, NULL);
static DEVICE_ATTR(lunid, S_IRUGO, lunid_show, NULL);
static DEVICE_ATTR(unique_id, S_IRUGO, unique_id_show, NULL);
static DEVICE_ATTR(rescan, S_IWUSR, NULL, host_store_rescan);
static DEVICE_ATTR(firmware_revision, S_IRUGO,
	host_show_firmware_revision, NULL);
static DEVICE_ATTR(commands_outstanding, S_IRUGO,
	host_show_commands_outstanding, NULL);
static DEVICE_ATTR(transport_mode, S_IRUGO,
	host_show_transport_mode, NULL);
481 482
static DEVICE_ATTR(resettable, S_IRUGO,
	host_show_resettable, NULL);
483 484 485 486 487 488 489 490 491 492 493 494 495

static struct device_attribute *hpsa_sdev_attrs[] = {
	&dev_attr_raid_level,
	&dev_attr_lunid,
	&dev_attr_unique_id,
	NULL,
};

static struct device_attribute *hpsa_shost_attrs[] = {
	&dev_attr_rescan,
	&dev_attr_firmware_revision,
	&dev_attr_commands_outstanding,
	&dev_attr_transport_mode,
496
	&dev_attr_resettable,
497 498 499 500 501
	NULL,
};

static struct scsi_host_template hpsa_driver_template = {
	.module			= THIS_MODULE,
502 503
	.name			= HPSA,
	.proc_name		= HPSA,
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518
	.queuecommand		= hpsa_scsi_queue_command,
	.scan_start		= hpsa_scan_start,
	.scan_finished		= hpsa_scan_finished,
	.change_queue_depth	= hpsa_change_queue_depth,
	.this_id		= -1,
	.use_clustering		= ENABLE_CLUSTERING,
	.eh_device_reset_handler = hpsa_eh_device_reset_handler,
	.ioctl			= hpsa_ioctl,
	.slave_alloc		= hpsa_slave_alloc,
	.slave_destroy		= hpsa_slave_destroy,
#ifdef CONFIG_COMPAT
	.compat_ioctl		= hpsa_compat_ioctl,
#endif
	.sdev_attrs = hpsa_sdev_attrs,
	.shost_attrs = hpsa_shost_attrs,
519
	.max_sectors = 8192,
520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594
};


/* Enqueuing and dequeuing functions for cmdlists. */
static inline void addQ(struct list_head *list, struct CommandList *c)
{
	list_add_tail(&c->list, list);
}

static inline u32 next_command(struct ctlr_info *h)
{
	u32 a;

	if (unlikely(!(h->transMethod & CFGTBL_Trans_Performant)))
		return h->access.command_completed(h);

	if ((*(h->reply_pool_head) & 1) == (h->reply_pool_wraparound)) {
		a = *(h->reply_pool_head); /* Next cmd in ring buffer */
		(h->reply_pool_head)++;
		h->commands_outstanding--;
	} else {
		a = FIFO_EMPTY;
	}
	/* Check for wraparound */
	if (h->reply_pool_head == (h->reply_pool + h->max_commands)) {
		h->reply_pool_head = h->reply_pool;
		h->reply_pool_wraparound ^= 1;
	}
	return a;
}

/* set_performant_mode: Modify the tag for cciss performant
 * set bit 0 for pull model, bits 3-1 for block fetch
 * register number
 */
static void set_performant_mode(struct ctlr_info *h, struct CommandList *c)
{
	if (likely(h->transMethod & CFGTBL_Trans_Performant))
		c->busaddr |= 1 | (h->blockFetchTable[c->Header.SGList] << 1);
}

static void enqueue_cmd_and_start_io(struct ctlr_info *h,
	struct CommandList *c)
{
	unsigned long flags;

	set_performant_mode(h, c);
	spin_lock_irqsave(&h->lock, flags);
	addQ(&h->reqQ, c);
	h->Qdepth++;
	start_io(h);
	spin_unlock_irqrestore(&h->lock, flags);
}

static inline void removeQ(struct CommandList *c)
{
	if (WARN_ON(list_empty(&c->list)))
		return;
	list_del_init(&c->list);
}

static inline int is_hba_lunid(unsigned char scsi3addr[])
{
	return memcmp(scsi3addr, RAID_CTLR_LUNID, 8) == 0;
}

static inline int is_scsi_rev_5(struct ctlr_info *h)
{
	if (!h->hba_inquiry_data)
		return 0;
	if ((h->hba_inquiry_data[2] & 0x07) == 5)
		return 1;
	return 0;
}

595 596 597 598 599 600 601
static int hpsa_find_target_lun(struct ctlr_info *h,
	unsigned char scsi3addr[], int bus, int *target, int *lun)
{
	/* finds an unused bus, target, lun for a new physical device
	 * assumes h->devlock is held
	 */
	int i, found = 0;
602
	DECLARE_BITMAP(lun_taken, HPSA_MAX_DEVICES);
603

604
	bitmap_zero(lun_taken, HPSA_MAX_DEVICES);
605 606 607

	for (i = 0; i < h->ndevices; i++) {
		if (h->dev[i]->bus == bus && h->dev[i]->target != -1)
608
			__set_bit(h->dev[i]->target, lun_taken);
609 610
	}

611 612 613 614 615 616
	i = find_first_zero_bit(lun_taken, HPSA_MAX_DEVICES);
	if (i < HPSA_MAX_DEVICES) {
		/* *bus = 1; */
		*target = i;
		*lun = 0;
		found = 1;
617 618 619 620 621 622 623 624 625 626 627 628 629 630 631
	}
	return !found;
}

/* Add an entry into h->dev[] array. */
static int hpsa_scsi_add_entry(struct ctlr_info *h, int hostno,
		struct hpsa_scsi_dev_t *device,
		struct hpsa_scsi_dev_t *added[], int *nadded)
{
	/* assumes h->devlock is held */
	int n = h->ndevices;
	int i;
	unsigned char addr1[8], addr2[8];
	struct hpsa_scsi_dev_t *sd;

632
	if (n >= HPSA_MAX_DEVICES) {
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699
		dev_err(&h->pdev->dev, "too many devices, some will be "
			"inaccessible.\n");
		return -1;
	}

	/* physical devices do not have lun or target assigned until now. */
	if (device->lun != -1)
		/* Logical device, lun is already assigned. */
		goto lun_assigned;

	/* If this device a non-zero lun of a multi-lun device
	 * byte 4 of the 8-byte LUN addr will contain the logical
	 * unit no, zero otherise.
	 */
	if (device->scsi3addr[4] == 0) {
		/* This is not a non-zero lun of a multi-lun device */
		if (hpsa_find_target_lun(h, device->scsi3addr,
			device->bus, &device->target, &device->lun) != 0)
			return -1;
		goto lun_assigned;
	}

	/* This is a non-zero lun of a multi-lun device.
	 * Search through our list and find the device which
	 * has the same 8 byte LUN address, excepting byte 4.
	 * Assign the same bus and target for this new LUN.
	 * Use the logical unit number from the firmware.
	 */
	memcpy(addr1, device->scsi3addr, 8);
	addr1[4] = 0;
	for (i = 0; i < n; i++) {
		sd = h->dev[i];
		memcpy(addr2, sd->scsi3addr, 8);
		addr2[4] = 0;
		/* differ only in byte 4? */
		if (memcmp(addr1, addr2, 8) == 0) {
			device->bus = sd->bus;
			device->target = sd->target;
			device->lun = device->scsi3addr[4];
			break;
		}
	}
	if (device->lun == -1) {
		dev_warn(&h->pdev->dev, "physical device with no LUN=0,"
			" suspect firmware bug or unsupported hardware "
			"configuration.\n");
			return -1;
	}

lun_assigned:

	h->dev[n] = device;
	h->ndevices++;
	added[*nadded] = device;
	(*nadded)++;

	/* initially, (before registering with scsi layer) we don't
	 * know our hostno and we don't want to print anything first
	 * time anyway (the scsi layer's inquiries will show that info)
	 */
	/* if (hostno != -1) */
		dev_info(&h->pdev->dev, "%s device c%db%dt%dl%d added.\n",
			scsi_device_type(device->devtype), hostno,
			device->bus, device->target, device->lun);
	return 0;
}

700 701 702 703 704 705 706 707 708 709 710 711 712 713
/* Update an entry in h->dev[] array. */
static void hpsa_scsi_update_entry(struct ctlr_info *h, int hostno,
	int entry, struct hpsa_scsi_dev_t *new_entry)
{
	/* assumes h->devlock is held */
	BUG_ON(entry < 0 || entry >= HPSA_MAX_DEVICES);

	/* Raid level changed. */
	h->dev[entry]->raid_level = new_entry->raid_level;
	dev_info(&h->pdev->dev, "%s device c%db%dt%dl%d updated.\n",
		scsi_device_type(new_entry->devtype), hostno, new_entry->bus,
		new_entry->target, new_entry->lun);
}

714 715 716 717 718 719 720
/* Replace an entry from h->dev[] array. */
static void hpsa_scsi_replace_entry(struct ctlr_info *h, int hostno,
	int entry, struct hpsa_scsi_dev_t *new_entry,
	struct hpsa_scsi_dev_t *added[], int *nadded,
	struct hpsa_scsi_dev_t *removed[], int *nremoved)
{
	/* assumes h->devlock is held */
721
	BUG_ON(entry < 0 || entry >= HPSA_MAX_DEVICES);
722 723
	removed[*nremoved] = h->dev[entry];
	(*nremoved)++;
724 725 726 727 728 729 730 731 732 733

	/*
	 * New physical devices won't have target/lun assigned yet
	 * so we need to preserve the values in the slot we are replacing.
	 */
	if (new_entry->target == -1) {
		new_entry->target = h->dev[entry]->target;
		new_entry->lun = h->dev[entry]->lun;
	}

734 735 736 737 738 739 740 741
	h->dev[entry] = new_entry;
	added[*nadded] = new_entry;
	(*nadded)++;
	dev_info(&h->pdev->dev, "%s device c%db%dt%dl%d changed.\n",
		scsi_device_type(new_entry->devtype), hostno, new_entry->bus,
			new_entry->target, new_entry->lun);
}

742 743 744 745 746 747 748 749
/* Remove an entry from h->dev[] array. */
static void hpsa_scsi_remove_entry(struct ctlr_info *h, int hostno, int entry,
	struct hpsa_scsi_dev_t *removed[], int *nremoved)
{
	/* assumes h->devlock is held */
	int i;
	struct hpsa_scsi_dev_t *sd;

750
	BUG_ON(entry < 0 || entry >= HPSA_MAX_DEVICES);
751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819

	sd = h->dev[entry];
	removed[*nremoved] = h->dev[entry];
	(*nremoved)++;

	for (i = entry; i < h->ndevices-1; i++)
		h->dev[i] = h->dev[i+1];
	h->ndevices--;
	dev_info(&h->pdev->dev, "%s device c%db%dt%dl%d removed.\n",
		scsi_device_type(sd->devtype), hostno, sd->bus, sd->target,
		sd->lun);
}

#define SCSI3ADDR_EQ(a, b) ( \
	(a)[7] == (b)[7] && \
	(a)[6] == (b)[6] && \
	(a)[5] == (b)[5] && \
	(a)[4] == (b)[4] && \
	(a)[3] == (b)[3] && \
	(a)[2] == (b)[2] && \
	(a)[1] == (b)[1] && \
	(a)[0] == (b)[0])

static void fixup_botched_add(struct ctlr_info *h,
	struct hpsa_scsi_dev_t *added)
{
	/* called when scsi_add_device fails in order to re-adjust
	 * h->dev[] to match the mid layer's view.
	 */
	unsigned long flags;
	int i, j;

	spin_lock_irqsave(&h->lock, flags);
	for (i = 0; i < h->ndevices; i++) {
		if (h->dev[i] == added) {
			for (j = i; j < h->ndevices-1; j++)
				h->dev[j] = h->dev[j+1];
			h->ndevices--;
			break;
		}
	}
	spin_unlock_irqrestore(&h->lock, flags);
	kfree(added);
}

static inline int device_is_the_same(struct hpsa_scsi_dev_t *dev1,
	struct hpsa_scsi_dev_t *dev2)
{
	/* we compare everything except lun and target as these
	 * are not yet assigned.  Compare parts likely
	 * to differ first
	 */
	if (memcmp(dev1->scsi3addr, dev2->scsi3addr,
		sizeof(dev1->scsi3addr)) != 0)
		return 0;
	if (memcmp(dev1->device_id, dev2->device_id,
		sizeof(dev1->device_id)) != 0)
		return 0;
	if (memcmp(dev1->model, dev2->model, sizeof(dev1->model)) != 0)
		return 0;
	if (memcmp(dev1->vendor, dev2->vendor, sizeof(dev1->vendor)) != 0)
		return 0;
	if (dev1->devtype != dev2->devtype)
		return 0;
	if (dev1->bus != dev2->bus)
		return 0;
	return 1;
}

820 821 822 823 824 825 826 827 828 829 830 831
static inline int device_updated(struct hpsa_scsi_dev_t *dev1,
	struct hpsa_scsi_dev_t *dev2)
{
	/* Device attributes that can change, but don't mean
	 * that the device is a different device, nor that the OS
	 * needs to be told anything about the change.
	 */
	if (dev1->raid_level != dev2->raid_level)
		return 1;
	return 0;
}

832 833 834
/* Find needle in haystack.  If exact match found, return DEVICE_SAME,
 * and return needle location in *index.  If scsi3addr matches, but not
 * vendor, model, serial num, etc. return DEVICE_CHANGED, and return needle
835 836 837 838
 * location in *index.
 * In the case of a minor device attribute change, such as RAID level, just
 * return DEVICE_UPDATED, along with the updated device's location in index.
 * If needle not found, return DEVICE_NOT_FOUND.
839 840 841 842 843 844 845 846 847
 */
static int hpsa_scsi_find_entry(struct hpsa_scsi_dev_t *needle,
	struct hpsa_scsi_dev_t *haystack[], int haystack_size,
	int *index)
{
	int i;
#define DEVICE_NOT_FOUND 0
#define DEVICE_CHANGED 1
#define DEVICE_SAME 2
848
#define DEVICE_UPDATED 3
849
	for (i = 0; i < haystack_size; i++) {
850 851
		if (haystack[i] == NULL) /* previously removed. */
			continue;
852 853
		if (SCSI3ADDR_EQ(needle->scsi3addr, haystack[i]->scsi3addr)) {
			*index = i;
854 855 856
			if (device_is_the_same(needle, haystack[i])) {
				if (device_updated(needle, haystack[i]))
					return DEVICE_UPDATED;
857
				return DEVICE_SAME;
858
			} else {
859
				return DEVICE_CHANGED;
860
			}
861 862 863 864 865 866
		}
	}
	*index = -1;
	return DEVICE_NOT_FOUND;
}

867
static void adjust_hpsa_scsi_table(struct ctlr_info *h, int hostno,
868 869 870 871 872 873 874 875 876 877 878 879 880
	struct hpsa_scsi_dev_t *sd[], int nsds)
{
	/* sd contains scsi3 addresses and devtypes, and inquiry
	 * data.  This function takes what's in sd to be the current
	 * reality and updates h->dev[] to reflect that reality.
	 */
	int i, entry, device_change, changes = 0;
	struct hpsa_scsi_dev_t *csd;
	unsigned long flags;
	struct hpsa_scsi_dev_t **added, **removed;
	int nadded, nremoved;
	struct Scsi_Host *sh = NULL;

881 882
	added = kzalloc(sizeof(*added) * HPSA_MAX_DEVICES, GFP_KERNEL);
	removed = kzalloc(sizeof(*removed) * HPSA_MAX_DEVICES, GFP_KERNEL);
883 884 885 886 887 888 889 890 891 892 893 894 895

	if (!added || !removed) {
		dev_warn(&h->pdev->dev, "out of memory in "
			"adjust_hpsa_scsi_table\n");
		goto free_and_out;
	}

	spin_lock_irqsave(&h->devlock, flags);

	/* find any devices in h->dev[] that are not in
	 * sd[] and remove them from h->dev[], and for any
	 * devices which have changed, remove the old device
	 * info and add the new device info.
896 897
	 * If minor device attributes change, just update
	 * the existing device structure.
898 899 900 901 902 903 904 905 906 907 908 909 910 911
	 */
	i = 0;
	nremoved = 0;
	nadded = 0;
	while (i < h->ndevices) {
		csd = h->dev[i];
		device_change = hpsa_scsi_find_entry(csd, sd, nsds, &entry);
		if (device_change == DEVICE_NOT_FOUND) {
			changes++;
			hpsa_scsi_remove_entry(h, hostno, i,
				removed, &nremoved);
			continue; /* remove ^^^, hence i not incremented */
		} else if (device_change == DEVICE_CHANGED) {
			changes++;
912 913
			hpsa_scsi_replace_entry(h, hostno, i, sd[entry],
				added, &nadded, removed, &nremoved);
914 915 916 917
			/* Set it to NULL to prevent it from being freed
			 * at the bottom of hpsa_update_scsi_devices()
			 */
			sd[entry] = NULL;
918 919
		} else if (device_change == DEVICE_UPDATED) {
			hpsa_scsi_update_entry(h, hostno, i, sd[entry]);
920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
		}
		i++;
	}

	/* Now, make sure every device listed in sd[] is also
	 * listed in h->dev[], adding them if they aren't found
	 */

	for (i = 0; i < nsds; i++) {
		if (!sd[i]) /* if already added above. */
			continue;
		device_change = hpsa_scsi_find_entry(sd[i], h->dev,
					h->ndevices, &entry);
		if (device_change == DEVICE_NOT_FOUND) {
			changes++;
			if (hpsa_scsi_add_entry(h, hostno, sd[i],
				added, &nadded) != 0)
				break;
			sd[i] = NULL; /* prevent from being freed later. */
		} else if (device_change == DEVICE_CHANGED) {
			/* should never happen... */
			changes++;
			dev_warn(&h->pdev->dev,
				"device unexpectedly changed.\n");
			/* but if it does happen, we just ignore that device */
		}
	}
	spin_unlock_irqrestore(&h->devlock, flags);

	/* Don't notify scsi mid layer of any changes the first time through
	 * (or if there are no changes) scsi_scan_host will do it later the
	 * first time through.
	 */
	if (hostno == -1 || !changes)
		goto free_and_out;

	sh = h->scsi_host;
	/* Notify scsi mid layer of any removed devices */
	for (i = 0; i < nremoved; i++) {
		struct scsi_device *sdev =
			scsi_device_lookup(sh, removed[i]->bus,
				removed[i]->target, removed[i]->lun);
		if (sdev != NULL) {
			scsi_remove_device(sdev);
			scsi_device_put(sdev);
		} else {
			/* We don't expect to get here.
			 * future cmds to this device will get selection
			 * timeout as if the device was gone.
			 */
			dev_warn(&h->pdev->dev, "didn't find c%db%dt%dl%d "
				" for removal.", hostno, removed[i]->bus,
				removed[i]->target, removed[i]->lun);
		}
		kfree(removed[i]);
		removed[i] = NULL;
	}

	/* Notify scsi mid layer of any added devices */
	for (i = 0; i < nadded; i++) {
		if (scsi_add_device(sh, added[i]->bus,
			added[i]->target, added[i]->lun) == 0)
			continue;
		dev_warn(&h->pdev->dev, "scsi_add_device c%db%dt%dl%d failed, "
			"device not added.\n", hostno, added[i]->bus,
			added[i]->target, added[i]->lun);
		/* now we have to remove it from h->dev,
		 * since it didn't get added to scsi mid layer
		 */
		fixup_botched_add(h, added[i]);
	}

free_and_out:
	kfree(added);
	kfree(removed);
}

/*
 * Lookup bus/target/lun and retrun corresponding struct hpsa_scsi_dev_t *
 * Assume's h->devlock is held.
 */
static struct hpsa_scsi_dev_t *lookup_hpsa_scsi_dev(struct ctlr_info *h,
	int bus, int target, int lun)
{
	int i;
	struct hpsa_scsi_dev_t *sd;

	for (i = 0; i < h->ndevices; i++) {
		sd = h->dev[i];
		if (sd->bus == bus && sd->target == target && sd->lun == lun)
			return sd;
	}
	return NULL;
}

/* link sdev->hostdata to our per-device structure. */
static int hpsa_slave_alloc(struct scsi_device *sdev)
{
	struct hpsa_scsi_dev_t *sd;
	unsigned long flags;
	struct ctlr_info *h;

	h = sdev_to_hba(sdev);
	spin_lock_irqsave(&h->devlock, flags);
	sd = lookup_hpsa_scsi_dev(h, sdev_channel(sdev),
		sdev_id(sdev), sdev->lun);
	if (sd != NULL)
		sdev->hostdata = sd;
	spin_unlock_irqrestore(&h->devlock, flags);
	return 0;
}

static void hpsa_slave_destroy(struct scsi_device *sdev)
{
1034
	/* nothing to do. */
1035 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 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106
static void hpsa_free_sg_chain_blocks(struct ctlr_info *h)
{
	int i;

	if (!h->cmd_sg_list)
		return;
	for (i = 0; i < h->nr_cmds; i++) {
		kfree(h->cmd_sg_list[i]);
		h->cmd_sg_list[i] = NULL;
	}
	kfree(h->cmd_sg_list);
	h->cmd_sg_list = NULL;
}

static int hpsa_allocate_sg_chain_blocks(struct ctlr_info *h)
{
	int i;

	if (h->chainsize <= 0)
		return 0;

	h->cmd_sg_list = kzalloc(sizeof(*h->cmd_sg_list) * h->nr_cmds,
				GFP_KERNEL);
	if (!h->cmd_sg_list)
		return -ENOMEM;
	for (i = 0; i < h->nr_cmds; i++) {
		h->cmd_sg_list[i] = kmalloc(sizeof(*h->cmd_sg_list[i]) *
						h->chainsize, GFP_KERNEL);
		if (!h->cmd_sg_list[i])
			goto clean;
	}
	return 0;

clean:
	hpsa_free_sg_chain_blocks(h);
	return -ENOMEM;
}

static void hpsa_map_sg_chain_block(struct ctlr_info *h,
	struct CommandList *c)
{
	struct SGDescriptor *chain_sg, *chain_block;
	u64 temp64;

	chain_sg = &c->SG[h->max_cmd_sg_entries - 1];
	chain_block = h->cmd_sg_list[c->cmdindex];
	chain_sg->Ext = HPSA_SG_CHAIN;
	chain_sg->Len = sizeof(*chain_sg) *
		(c->Header.SGTotal - h->max_cmd_sg_entries);
	temp64 = pci_map_single(h->pdev, chain_block, chain_sg->Len,
				PCI_DMA_TODEVICE);
	chain_sg->Addr.lower = (u32) (temp64 & 0x0FFFFFFFFULL);
	chain_sg->Addr.upper = (u32) ((temp64 >> 32) & 0x0FFFFFFFFULL);
}

static void hpsa_unmap_sg_chain_block(struct ctlr_info *h,
	struct CommandList *c)
{
	struct SGDescriptor *chain_sg;
	union u64bit temp64;

	if (c->Header.SGTotal <= h->max_cmd_sg_entries)
		return;

	chain_sg = &c->SG[h->max_cmd_sg_entries - 1];
	temp64.val32.lower = chain_sg->Addr.lower;
	temp64.val32.upper = chain_sg->Addr.upper;
	pci_unmap_single(h->pdev, temp64.val, chain_sg->Len, PCI_DMA_TODEVICE);
}

1107
static void complete_scsi_command(struct CommandList *cp)
1108 1109 1110 1111 1112 1113 1114 1115
{
	struct scsi_cmnd *cmd;
	struct ctlr_info *h;
	struct ErrorInfo *ei;

	unsigned char sense_key;
	unsigned char asc;      /* additional sense code */
	unsigned char ascq;     /* additional sense code qualifier */
1116
	unsigned long sense_data_size;
1117 1118 1119 1120 1121 1122

	ei = cp->err_info;
	cmd = (struct scsi_cmnd *) cp->scsi_cmd;
	h = cp->h;

	scsi_dma_unmap(cmd); /* undo the DMA mappings */
1123 1124
	if (cp->Header.SGTotal > h->max_cmd_sg_entries)
		hpsa_unmap_sg_chain_block(h, cp);
1125 1126 1127

	cmd->result = (DID_OK << 16); 		/* host byte */
	cmd->result |= (COMMAND_COMPLETE << 8);	/* msg byte */
1128
	cmd->result |= ei->ScsiStatus;
1129 1130

	/* copy the sense data whether we need to or not. */
1131 1132 1133 1134 1135 1136 1137 1138
	if (SCSI_SENSE_BUFFERSIZE < sizeof(ei->SenseInfo))
		sense_data_size = SCSI_SENSE_BUFFERSIZE;
	else
		sense_data_size = sizeof(ei->SenseInfo);
	if (ei->SenseLen < sense_data_size)
		sense_data_size = ei->SenseLen;

	memcpy(cmd->sense_buffer, ei->SenseInfo, sense_data_size);
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 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
	scsi_set_resid(cmd, ei->ResidualCnt);

	if (ei->CommandStatus == 0) {
		cmd->scsi_done(cmd);
		cmd_free(h, cp);
		return;
	}

	/* an error has occurred */
	switch (ei->CommandStatus) {

	case CMD_TARGET_STATUS:
		if (ei->ScsiStatus) {
			/* Get sense key */
			sense_key = 0xf & ei->SenseInfo[2];
			/* Get additional sense code */
			asc = ei->SenseInfo[12];
			/* Get addition sense code qualifier */
			ascq = ei->SenseInfo[13];
		}

		if (ei->ScsiStatus == SAM_STAT_CHECK_CONDITION) {
			if (check_for_unit_attention(h, cp)) {
				cmd->result = DID_SOFT_ERROR << 16;
				break;
			}
			if (sense_key == ILLEGAL_REQUEST) {
				/*
				 * SCSI REPORT_LUNS is commonly unsupported on
				 * Smart Array.  Suppress noisy complaint.
				 */
				if (cp->Request.CDB[0] == REPORT_LUNS)
					break;

				/* If ASC/ASCQ indicate Logical Unit
				 * Not Supported condition,
				 */
				if ((asc == 0x25) && (ascq == 0x0)) {
					dev_warn(&h->pdev->dev, "cp %p "
						"has check condition\n", cp);
					break;
				}
			}

			if (sense_key == NOT_READY) {
				/* If Sense is Not Ready, Logical Unit
				 * Not ready, Manual Intervention
				 * required
				 */
				if ((asc == 0x04) && (ascq == 0x03)) {
					dev_warn(&h->pdev->dev, "cp %p "
						"has check condition: unit "
						"not ready, manual "
						"intervention required\n", cp);
					break;
				}
			}
1196 1197 1198 1199 1200 1201 1202 1203 1204
			if (sense_key == ABORTED_COMMAND) {
				/* Aborted command is retryable */
				dev_warn(&h->pdev->dev, "cp %p "
					"has check condition: aborted command: "
					"ASC: 0x%x, ASCQ: 0x%x\n",
					cp, asc, ascq);
				cmd->result = DID_SOFT_ERROR << 16;
				break;
			}
1205
			/* Must be some other type of check condition */
1206
			dev_dbg(&h->pdev->dev, "cp %p has check condition: "
1207 1208 1209 1210
					"unknown type: "
					"Sense: 0x%x, ASC: 0x%x, ASCQ: 0x%x, "
					"Returning result: 0x%x, "
					"cmd=[%02x %02x %02x %02x %02x "
1211
					"%02x %02x %02x %02x %02x %02x "
1212 1213 1214 1215 1216 1217 1218
					"%02x %02x %02x %02x %02x]\n",
					cp, sense_key, asc, ascq,
					cmd->result,
					cmd->cmnd[0], cmd->cmnd[1],
					cmd->cmnd[2], cmd->cmnd[3],
					cmd->cmnd[4], cmd->cmnd[5],
					cmd->cmnd[6], cmd->cmnd[7],
1219 1220 1221 1222
					cmd->cmnd[8], cmd->cmnd[9],
					cmd->cmnd[10], cmd->cmnd[11],
					cmd->cmnd[12], cmd->cmnd[13],
					cmd->cmnd[14], cmd->cmnd[15]);
1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 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
			break;
		}


		/* Problem was not a check condition
		 * Pass it up to the upper layers...
		 */
		if (ei->ScsiStatus) {
			dev_warn(&h->pdev->dev, "cp %p has status 0x%x "
				"Sense: 0x%x, ASC: 0x%x, ASCQ: 0x%x, "
				"Returning result: 0x%x\n",
				cp, ei->ScsiStatus,
				sense_key, asc, ascq,
				cmd->result);
		} else {  /* scsi status is zero??? How??? */
			dev_warn(&h->pdev->dev, "cp %p SCSI status was 0. "
				"Returning no connection.\n", cp),

			/* Ordinarily, this case should never happen,
			 * but there is a bug in some released firmware
			 * revisions that allows it to happen if, for
			 * example, a 4100 backplane loses power and
			 * the tape drive is in it.  We assume that
			 * it's a fatal error of some kind because we
			 * can't show that it wasn't. We will make it
			 * look like selection timeout since that is
			 * the most common reason for this to occur,
			 * and it's severe enough.
			 */

			cmd->result = DID_NO_CONNECT << 16;
		}
		break;

	case CMD_DATA_UNDERRUN: /* let mid layer handle it. */
		break;
	case CMD_DATA_OVERRUN:
		dev_warn(&h->pdev->dev, "cp %p has"
			" completed with data overrun "
			"reported\n", cp);
		break;
	case CMD_INVALID: {
		/* print_bytes(cp, sizeof(*cp), 1, 0);
		print_cmd(cp); */
		/* We get CMD_INVALID if you address a non-existent device
		 * instead of a selection timeout (no response).  You will
		 * see this if you yank out a drive, then try to access it.
		 * This is kind of a shame because it means that any other
		 * CMD_INVALID (e.g. driver bug) will get interpreted as a
		 * missing target. */
		cmd->result = DID_NO_CONNECT << 16;
	}
		break;
	case CMD_PROTOCOL_ERR:
		dev_warn(&h->pdev->dev, "cp %p has "
			"protocol error \n", cp);
		break;
	case CMD_HARDWARE_ERR:
		cmd->result = DID_ERROR << 16;
		dev_warn(&h->pdev->dev, "cp %p had  hardware error\n", cp);
		break;
	case CMD_CONNECTION_LOST:
		cmd->result = DID_ERROR << 16;
		dev_warn(&h->pdev->dev, "cp %p had connection lost\n", cp);
		break;
	case CMD_ABORTED:
		cmd->result = DID_ABORT << 16;
		dev_warn(&h->pdev->dev, "cp %p was aborted with status 0x%x\n",
				cp, ei->ScsiStatus);
		break;
	case CMD_ABORT_FAILED:
		cmd->result = DID_ERROR << 16;
		dev_warn(&h->pdev->dev, "cp %p reports abort failed\n", cp);
		break;
	case CMD_UNSOLICITED_ABORT:
1298 1299
		cmd->result = DID_SOFT_ERROR << 16; /* retry the command */
		dev_warn(&h->pdev->dev, "cp %p aborted due to an unsolicited "
1300 1301 1302 1303 1304 1305
			"abort\n", cp);
		break;
	case CMD_TIMEOUT:
		cmd->result = DID_TIME_OUT << 16;
		dev_warn(&h->pdev->dev, "cp %p timedout\n", cp);
		break;
1306 1307 1308 1309
	case CMD_UNABORTABLE:
		cmd->result = DID_ERROR << 16;
		dev_warn(&h->pdev->dev, "Command unabortable\n");
		break;
1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
	default:
		cmd->result = DID_ERROR << 16;
		dev_warn(&h->pdev->dev, "cp %p returned unknown status %x\n",
				cp, ei->CommandStatus);
	}
	cmd->scsi_done(cmd);
	cmd_free(h, cp);
}

static void hpsa_pci_unmap(struct pci_dev *pdev,
	struct CommandList *c, int sg_used, int data_direction)
{
	int i;
	union u64bit addr64;

	for (i = 0; i < sg_used; i++) {
		addr64.val32.lower = c->SG[i].Addr.lower;
		addr64.val32.upper = c->SG[i].Addr.upper;
		pci_unmap_single(pdev, (dma_addr_t) addr64.val, c->SG[i].Len,
			data_direction);
	}
}

static void hpsa_map_one(struct pci_dev *pdev,
		struct CommandList *cp,
		unsigned char *buf,
		size_t buflen,
		int data_direction)
{
1339
	u64 addr64;
1340 1341 1342 1343 1344 1345 1346

	if (buflen == 0 || data_direction == PCI_DMA_NONE) {
		cp->Header.SGList = 0;
		cp->Header.SGTotal = 0;
		return;
	}

1347
	addr64 = (u64) pci_map_single(pdev, buf, buflen, data_direction);
1348
	cp->SG[0].Addr.lower =
1349
	  (u32) (addr64 & (u64) 0x00000000FFFFFFFF);
1350
	cp->SG[0].Addr.upper =
1351
	  (u32) ((addr64 >> 32) & (u64) 0x00000000FFFFFFFF);
1352
	cp->SG[0].Len = buflen;
1353 1354
	cp->Header.SGList = (u8) 1;   /* no. SGs contig in this cmd */
	cp->Header.SGTotal = (u16) 1; /* total sgs in this cmd list */
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
}

static inline void hpsa_scsi_do_simple_cmd_core(struct ctlr_info *h,
	struct CommandList *c)
{
	DECLARE_COMPLETION_ONSTACK(wait);

	c->waiting = &wait;
	enqueue_cmd_and_start_io(h, c);
	wait_for_completion(&wait);
}

1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382
static void hpsa_scsi_do_simple_cmd_core_if_no_lockup(struct ctlr_info *h,
	struct CommandList *c)
{
	unsigned long flags;

	/* If controller lockup detected, fake a hardware error. */
	spin_lock_irqsave(&h->lock, flags);
	if (unlikely(h->lockup_detected)) {
		spin_unlock_irqrestore(&h->lock, flags);
		c->err_info->CommandStatus = CMD_HARDWARE_ERR;
	} else {
		spin_unlock_irqrestore(&h->lock, flags);
		hpsa_scsi_do_simple_cmd_core(h, c);
	}
}

1383
#define MAX_DRIVER_CMD_RETRIES 25
1384 1385 1386
static void hpsa_scsi_do_simple_cmd_with_retry(struct ctlr_info *h,
	struct CommandList *c, int data_direction)
{
1387
	int backoff_time = 10, retry_count = 0;
1388 1389

	do {
1390
		memset(c->err_info, 0, sizeof(*c->err_info));
1391 1392
		hpsa_scsi_do_simple_cmd_core(h, c);
		retry_count++;
1393 1394 1395 1396 1397
		if (retry_count > 3) {
			msleep(backoff_time);
			if (backoff_time < 1000)
				backoff_time *= 2;
		}
1398
	} while ((check_for_unit_attention(h, c) ||
1399 1400
			check_for_busy(h, c)) &&
			retry_count <= MAX_DRIVER_CMD_RETRIES);
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458
	hpsa_pci_unmap(h->pdev, c, 1, data_direction);
}

static void hpsa_scsi_interpret_error(struct CommandList *cp)
{
	struct ErrorInfo *ei;
	struct device *d = &cp->h->pdev->dev;

	ei = cp->err_info;
	switch (ei->CommandStatus) {
	case CMD_TARGET_STATUS:
		dev_warn(d, "cmd %p has completed with errors\n", cp);
		dev_warn(d, "cmd %p has SCSI Status = %x\n", cp,
				ei->ScsiStatus);
		if (ei->ScsiStatus == 0)
			dev_warn(d, "SCSI status is abnormally zero.  "
			"(probably indicates selection timeout "
			"reported incorrectly due to a known "
			"firmware bug, circa July, 2001.)\n");
		break;
	case CMD_DATA_UNDERRUN: /* let mid layer handle it. */
			dev_info(d, "UNDERRUN\n");
		break;
	case CMD_DATA_OVERRUN:
		dev_warn(d, "cp %p has completed with data overrun\n", cp);
		break;
	case CMD_INVALID: {
		/* controller unfortunately reports SCSI passthru's
		 * to non-existent targets as invalid commands.
		 */
		dev_warn(d, "cp %p is reported invalid (probably means "
			"target device no longer present)\n", cp);
		/* print_bytes((unsigned char *) cp, sizeof(*cp), 1, 0);
		print_cmd(cp);  */
		}
		break;
	case CMD_PROTOCOL_ERR:
		dev_warn(d, "cp %p has protocol error \n", cp);
		break;
	case CMD_HARDWARE_ERR:
		/* cmd->result = DID_ERROR << 16; */
		dev_warn(d, "cp %p had hardware error\n", cp);
		break;
	case CMD_CONNECTION_LOST:
		dev_warn(d, "cp %p had connection lost\n", cp);
		break;
	case CMD_ABORTED:
		dev_warn(d, "cp %p was aborted\n", cp);
		break;
	case CMD_ABORT_FAILED:
		dev_warn(d, "cp %p reports abort failed\n", cp);
		break;
	case CMD_UNSOLICITED_ABORT:
		dev_warn(d, "cp %p aborted due to an unsolicited abort\n", cp);
		break;
	case CMD_TIMEOUT:
		dev_warn(d, "cp %p timed out\n", cp);
		break;
1459 1460 1461
	case CMD_UNABORTABLE:
		dev_warn(d, "Command unabortable\n");
		break;
1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
	default:
		dev_warn(d, "cp %p returned unknown status %x\n", cp,
				ei->CommandStatus);
	}
}

static int hpsa_scsi_do_inquiry(struct ctlr_info *h, unsigned char *scsi3addr,
			unsigned char page, unsigned char *buf,
			unsigned char bufsize)
{
	int rc = IO_OK;
	struct CommandList *c;
	struct ErrorInfo *ei;

	c = cmd_special_alloc(h);

	if (c == NULL) {			/* trouble... */
		dev_warn(&h->pdev->dev, "cmd_special_alloc returned NULL!\n");
1480
		return -ENOMEM;
1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
	}

	fill_cmd(c, HPSA_INQUIRY, h, buf, bufsize, page, scsi3addr, TYPE_CMD);
	hpsa_scsi_do_simple_cmd_with_retry(h, c, PCI_DMA_FROMDEVICE);
	ei = c->err_info;
	if (ei->CommandStatus != 0 && ei->CommandStatus != CMD_DATA_UNDERRUN) {
		hpsa_scsi_interpret_error(c);
		rc = -1;
	}
	cmd_special_free(h, c);
	return rc;
}

static int hpsa_send_reset(struct ctlr_info *h, unsigned char *scsi3addr)
{
	int rc = IO_OK;
	struct CommandList *c;
	struct ErrorInfo *ei;

	c = cmd_special_alloc(h);

	if (c == NULL) {			/* trouble... */
		dev_warn(&h->pdev->dev, "cmd_special_alloc returned NULL!\n");
1504
		return -ENOMEM;
1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571
	}

	fill_cmd(c, HPSA_DEVICE_RESET_MSG, h, NULL, 0, 0, scsi3addr, TYPE_MSG);
	hpsa_scsi_do_simple_cmd_core(h, c);
	/* no unmap needed here because no data xfer. */

	ei = c->err_info;
	if (ei->CommandStatus != 0) {
		hpsa_scsi_interpret_error(c);
		rc = -1;
	}
	cmd_special_free(h, c);
	return rc;
}

static void hpsa_get_raid_level(struct ctlr_info *h,
	unsigned char *scsi3addr, unsigned char *raid_level)
{
	int rc;
	unsigned char *buf;

	*raid_level = RAID_UNKNOWN;
	buf = kzalloc(64, GFP_KERNEL);
	if (!buf)
		return;
	rc = hpsa_scsi_do_inquiry(h, scsi3addr, 0xC1, buf, 64);
	if (rc == 0)
		*raid_level = buf[8];
	if (*raid_level > RAID_UNKNOWN)
		*raid_level = RAID_UNKNOWN;
	kfree(buf);
	return;
}

/* Get the device id from inquiry page 0x83 */
static int hpsa_get_device_id(struct ctlr_info *h, unsigned char *scsi3addr,
	unsigned char *device_id, int buflen)
{
	int rc;
	unsigned char *buf;

	if (buflen > 16)
		buflen = 16;
	buf = kzalloc(64, GFP_KERNEL);
	if (!buf)
		return -1;
	rc = hpsa_scsi_do_inquiry(h, scsi3addr, 0x83, buf, 64);
	if (rc == 0)
		memcpy(device_id, &buf[8], buflen);
	kfree(buf);
	return rc != 0;
}

static int hpsa_scsi_do_report_luns(struct ctlr_info *h, int logical,
		struct ReportLUNdata *buf, int bufsize,
		int extended_response)
{
	int rc = IO_OK;
	struct CommandList *c;
	unsigned char scsi3addr[8];
	struct ErrorInfo *ei;

	c = cmd_special_alloc(h);
	if (c == NULL) {			/* trouble... */
		dev_err(&h->pdev->dev, "cmd_special_alloc returned NULL!\n");
		return -1;
	}
1572 1573
	/* address the controller */
	memset(scsi3addr, 0, sizeof(scsi3addr));
1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610
	fill_cmd(c, logical ? HPSA_REPORT_LOG : HPSA_REPORT_PHYS, h,
		buf, bufsize, 0, scsi3addr, TYPE_CMD);
	if (extended_response)
		c->Request.CDB[1] = extended_response;
	hpsa_scsi_do_simple_cmd_with_retry(h, c, PCI_DMA_FROMDEVICE);
	ei = c->err_info;
	if (ei->CommandStatus != 0 &&
	    ei->CommandStatus != CMD_DATA_UNDERRUN) {
		hpsa_scsi_interpret_error(c);
		rc = -1;
	}
	cmd_special_free(h, c);
	return rc;
}

static inline int hpsa_scsi_do_report_phys_luns(struct ctlr_info *h,
		struct ReportLUNdata *buf,
		int bufsize, int extended_response)
{
	return hpsa_scsi_do_report_luns(h, 0, buf, bufsize, extended_response);
}

static inline int hpsa_scsi_do_report_log_luns(struct ctlr_info *h,
		struct ReportLUNdata *buf, int bufsize)
{
	return hpsa_scsi_do_report_luns(h, 1, buf, bufsize, 0);
}

static inline void hpsa_set_bus_target_lun(struct hpsa_scsi_dev_t *device,
	int bus, int target, int lun)
{
	device->bus = bus;
	device->target = target;
	device->lun = lun;
}

static int hpsa_update_device_info(struct ctlr_info *h,
1611 1612
	unsigned char scsi3addr[], struct hpsa_scsi_dev_t *this_device,
	unsigned char *is_OBDR_device)
1613
{
1614 1615 1616 1617 1618 1619

#define OBDR_SIG_OFFSET 43
#define OBDR_TAPE_SIG "$DR-10"
#define OBDR_SIG_LEN (sizeof(OBDR_TAPE_SIG) - 1)
#define OBDR_TAPE_INQ_SIZE (OBDR_SIG_OFFSET + OBDR_SIG_LEN)

1620
	unsigned char *inq_buff;
1621
	unsigned char *obdr_sig;
1622

1623
	inq_buff = kzalloc(OBDR_TAPE_INQ_SIZE, GFP_KERNEL);
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 1652
	if (!inq_buff)
		goto bail_out;

	/* Do an inquiry to the device to see what it is. */
	if (hpsa_scsi_do_inquiry(h, scsi3addr, 0, inq_buff,
		(unsigned char) OBDR_TAPE_INQ_SIZE) != 0) {
		/* Inquiry failed (msg printed already) */
		dev_err(&h->pdev->dev,
			"hpsa_update_device_info: inquiry failed\n");
		goto bail_out;
	}

	this_device->devtype = (inq_buff[0] & 0x1f);
	memcpy(this_device->scsi3addr, scsi3addr, 8);
	memcpy(this_device->vendor, &inq_buff[8],
		sizeof(this_device->vendor));
	memcpy(this_device->model, &inq_buff[16],
		sizeof(this_device->model));
	memset(this_device->device_id, 0,
		sizeof(this_device->device_id));
	hpsa_get_device_id(h, scsi3addr, this_device->device_id,
		sizeof(this_device->device_id));

	if (this_device->devtype == TYPE_DISK &&
		is_logical_dev_addr_mode(scsi3addr))
		hpsa_get_raid_level(h, scsi3addr, &this_device->raid_level);
	else
		this_device->raid_level = RAID_UNKNOWN;

1653 1654 1655 1656 1657 1658 1659 1660 1661 1662
	if (is_OBDR_device) {
		/* See if this is a One-Button-Disaster-Recovery device
		 * by looking for "$DR-10" at offset 43 in inquiry data.
		 */
		obdr_sig = &inq_buff[OBDR_SIG_OFFSET];
		*is_OBDR_device = (this_device->devtype == TYPE_ROM &&
					strncmp(obdr_sig, OBDR_TAPE_SIG,
						OBDR_SIG_LEN) == 0);
	}

1663 1664 1665 1666 1667 1668 1669 1670
	kfree(inq_buff);
	return 0;

bail_out:
	kfree(inq_buff);
	return 1;
}

1671
static unsigned char *ext_target_model[] = {
1672 1673 1674 1675
	"MSA2012",
	"MSA2024",
	"MSA2312",
	"MSA2324",
1676
	"P2000 G3 SAS",
1677 1678 1679
	NULL,
};

1680
static int is_ext_target(struct ctlr_info *h, struct hpsa_scsi_dev_t *device)
1681 1682 1683
{
	int i;

1684 1685 1686
	for (i = 0; ext_target_model[i]; i++)
		if (strncmp(device->model, ext_target_model[i],
			strlen(ext_target_model[i])) == 0)
1687 1688 1689 1690 1691
			return 1;
	return 0;
}

/* Helper function to assign bus, target, lun mapping of devices.
1692
 * Puts non-external target logical volumes on bus 0, external target logical
1693 1694 1695 1696 1697 1698
 * volumes on bus 1, physical devices on bus 2. and the hba on bus 3.
 * Logical drive target and lun are assigned at this time, but
 * physical device lun and target assignment are deferred (assigned
 * in hpsa_find_target_lun, called by hpsa_scsi_add_entry.)
 */
static void figure_bus_target_lun(struct ctlr_info *h,
1699
	u8 *lunaddrbytes, struct hpsa_scsi_dev_t *device)
1700
{
1701 1702 1703 1704
	u32 lunid = le32_to_cpu(*((__le32 *) lunaddrbytes));

	if (!is_logical_dev_addr_mode(lunaddrbytes)) {
		/* physical device, target and lun filled in later */
1705
		if (is_hba_lunid(lunaddrbytes))
1706
			hpsa_set_bus_target_lun(device, 3, 0, lunid & 0x3fff);
1707
		else
1708 1709 1710 1711 1712
			/* defer target, lun assignment for physical devices */
			hpsa_set_bus_target_lun(device, 2, -1, -1);
		return;
	}
	/* It's a logical device */
1713 1714
	if (is_ext_target(h, device)) {
		/* external target way, put logicals on bus 1
1715 1716 1717 1718 1719 1720
		 * and match target/lun numbers box
		 * reports, other smart array, bus 0, target 0, match lunid
		 */
		hpsa_set_bus_target_lun(device,
			1, (lunid >> 16) & 0x3fff, lunid & 0x00ff);
		return;
1721
	}
1722
	hpsa_set_bus_target_lun(device, 0, 0, lunid & 0x3fff);
1723 1724 1725 1726
}

/*
 * If there is no lun 0 on a target, linux won't find any devices.
1727
 * For the external targets (arrays), we have to manually detect the enclosure
1728 1729 1730 1731 1732 1733 1734 1735
 * which is at lun zero, as CCISS_REPORT_PHYSICAL_LUNS doesn't report
 * it for some reason.  *tmpdevice is the target we're adding,
 * this_device is a pointer into the current element of currentsd[]
 * that we're building up in update_scsi_devices(), below.
 * lunzerobits is a bitmap that tracks which targets already have a
 * lun 0 assigned.
 * Returns 1 if an enclosure was added, 0 if not.
 */
1736
static int add_ext_target_dev(struct ctlr_info *h,
1737
	struct hpsa_scsi_dev_t *tmpdevice,
1738
	struct hpsa_scsi_dev_t *this_device, u8 *lunaddrbytes,
1739
	unsigned long lunzerobits[], int *n_ext_target_devs)
1740 1741 1742
{
	unsigned char scsi3addr[8];

1743
	if (test_bit(tmpdevice->target, lunzerobits))
1744 1745 1746 1747 1748
		return 0; /* There is already a lun 0 on this target. */

	if (!is_logical_dev_addr_mode(lunaddrbytes))
		return 0; /* It's the logical targets that may lack lun 0. */

1749 1750
	if (!is_ext_target(h, tmpdevice))
		return 0; /* Only external target devices have this problem. */
1751

1752
	if (tmpdevice->lun == 0) /* if lun is 0, then we have a lun 0. */
1753 1754
		return 0;

1755
	memset(scsi3addr, 0, 8);
1756
	scsi3addr[3] = tmpdevice->target;
1757 1758 1759
	if (is_hba_lunid(scsi3addr))
		return 0; /* Don't add the RAID controller here. */

1760 1761 1762
	if (is_scsi_rev_5(h))
		return 0; /* p1210m doesn't need to do this. */

1763
	if (*n_ext_target_devs >= MAX_EXT_TARGETS) {
1764 1765
		dev_warn(&h->pdev->dev, "Maximum number of external "
			"target devices exceeded.  Check your hardware "
1766 1767 1768 1769
			"configuration.");
		return 0;
	}

1770
	if (hpsa_update_device_info(h, scsi3addr, this_device, NULL))
1771
		return 0;
1772
	(*n_ext_target_devs)++;
1773 1774 1775
	hpsa_set_bus_target_lun(this_device,
				tmpdevice->bus, tmpdevice->target, 0);
	set_bit(tmpdevice->target, lunzerobits);
1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786
	return 1;
}

/*
 * Do CISS_REPORT_PHYS and CISS_REPORT_LOG.  Data is returned in physdev,
 * logdev.  The number of luns in physdev and logdev are returned in
 * *nphysicals and *nlogicals, respectively.
 * Returns 0 on success, -1 otherwise.
 */
static int hpsa_gather_lun_info(struct ctlr_info *h,
	int reportlunsize,
1787 1788
	struct ReportLUNdata *physdev, u32 *nphysicals,
	struct ReportLUNdata *logdev, u32 *nlogicals)
1789 1790 1791 1792 1793
{
	if (hpsa_scsi_do_report_phys_luns(h, physdev, reportlunsize, 0)) {
		dev_err(&h->pdev->dev, "report physical LUNs failed.\n");
		return -1;
	}
1794
	*nphysicals = be32_to_cpu(*((__be32 *)physdev->LUNListLength)) / 8;
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804
	if (*nphysicals > HPSA_MAX_PHYS_LUN) {
		dev_warn(&h->pdev->dev, "maximum physical LUNs (%d) exceeded."
			"  %d LUNs ignored.\n", HPSA_MAX_PHYS_LUN,
			*nphysicals - HPSA_MAX_PHYS_LUN);
		*nphysicals = HPSA_MAX_PHYS_LUN;
	}
	if (hpsa_scsi_do_report_log_luns(h, logdev, reportlunsize)) {
		dev_err(&h->pdev->dev, "report logical LUNs failed.\n");
		return -1;
	}
1805
	*nlogicals = be32_to_cpu(*((__be32 *) logdev->LUNListLength)) / 8;
1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
	/* Reject Logicals in excess of our max capability. */
	if (*nlogicals > HPSA_MAX_LUN) {
		dev_warn(&h->pdev->dev,
			"maximum logical LUNs (%d) exceeded.  "
			"%d LUNs ignored.\n", HPSA_MAX_LUN,
			*nlogicals - HPSA_MAX_LUN);
			*nlogicals = HPSA_MAX_LUN;
	}
	if (*nlogicals + *nphysicals > HPSA_MAX_PHYS_LUN) {
		dev_warn(&h->pdev->dev,
			"maximum logical + physical LUNs (%d) exceeded. "
			"%d LUNs ignored.\n", HPSA_MAX_PHYS_LUN,
			*nphysicals + *nlogicals - HPSA_MAX_PHYS_LUN);
		*nlogicals = HPSA_MAX_PHYS_LUN - *nphysicals;
	}
	return 0;
}

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
u8 *figure_lunaddrbytes(struct ctlr_info *h, int raid_ctlr_position, int i,
	int nphysicals, int nlogicals, struct ReportLUNdata *physdev_list,
	struct ReportLUNdata *logdev_list)
{
	/* Helper function, figure out where the LUN ID info is coming from
	 * given index i, lists of physical and logical devices, where in
	 * the list the raid controller is supposed to appear (first or last)
	 */

	int logicals_start = nphysicals + (raid_ctlr_position == 0);
	int last_device = nphysicals + nlogicals + (raid_ctlr_position == 0);

	if (i == raid_ctlr_position)
		return RAID_CTLR_LUNID;

	if (i < logicals_start)
		return &physdev_list->LUN[i - (raid_ctlr_position == 0)][0];

	if (i < last_device)
		return &logdev_list->LUN[i - nphysicals -
			(raid_ctlr_position == 0)][0];
	BUG();
	return NULL;
}

1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862
static void hpsa_update_scsi_devices(struct ctlr_info *h, int hostno)
{
	/* the idea here is we could get notified
	 * that some devices have changed, so we do a report
	 * physical luns and report logical luns cmd, and adjust
	 * our list of devices accordingly.
	 *
	 * The scsi3addr's of devices won't change so long as the
	 * adapter is not reset.  That means we can rescan and
	 * tell which devices we already know about, vs. new
	 * devices, vs.  disappearing devices.
	 */
	struct ReportLUNdata *physdev_list = NULL;
	struct ReportLUNdata *logdev_list = NULL;
1863 1864 1865
	u32 nphysicals = 0;
	u32 nlogicals = 0;
	u32 ndev_allocated = 0;
1866 1867 1868
	struct hpsa_scsi_dev_t **currentsd, *this_device, *tmpdevice;
	int ncurrent = 0;
	int reportlunsize = sizeof(*physdev_list) + HPSA_MAX_PHYS_LUN * 8;
1869
	int i, n_ext_target_devs, ndevs_to_allocate;
1870
	int raid_ctlr_position;
1871
	DECLARE_BITMAP(lunzerobits, MAX_EXT_TARGETS);
1872

1873
	currentsd = kzalloc(sizeof(*currentsd) * HPSA_MAX_DEVICES, GFP_KERNEL);
1874 1875 1876 1877
	physdev_list = kzalloc(reportlunsize, GFP_KERNEL);
	logdev_list = kzalloc(reportlunsize, GFP_KERNEL);
	tmpdevice = kzalloc(sizeof(*tmpdevice), GFP_KERNEL);

1878
	if (!currentsd || !physdev_list || !logdev_list || !tmpdevice) {
1879 1880 1881 1882 1883 1884 1885 1886 1887
		dev_err(&h->pdev->dev, "out of memory\n");
		goto out;
	}
	memset(lunzerobits, 0, sizeof(lunzerobits));

	if (hpsa_gather_lun_info(h, reportlunsize, physdev_list, &nphysicals,
			logdev_list, &nlogicals))
		goto out;

1888 1889 1890
	/* We might see up to the maximum number of logical and physical disks
	 * plus external target devices, and a device for the local RAID
	 * controller.
1891
	 */
1892
	ndevs_to_allocate = nphysicals + nlogicals + MAX_EXT_TARGETS + 1;
1893 1894 1895

	/* Allocate the per device structures */
	for (i = 0; i < ndevs_to_allocate; i++) {
1896 1897 1898 1899 1900 1901 1902
		if (i >= HPSA_MAX_DEVICES) {
			dev_warn(&h->pdev->dev, "maximum devices (%d) exceeded."
				"  %d devices ignored.\n", HPSA_MAX_DEVICES,
				ndevs_to_allocate - HPSA_MAX_DEVICES);
			break;
		}

1903 1904 1905 1906 1907 1908 1909 1910 1911
		currentsd[i] = kzalloc(sizeof(*currentsd[i]), GFP_KERNEL);
		if (!currentsd[i]) {
			dev_warn(&h->pdev->dev, "out of memory at %s:%d\n",
				__FILE__, __LINE__);
			goto out;
		}
		ndev_allocated++;
	}

1912 1913 1914 1915 1916
	if (unlikely(is_scsi_rev_5(h)))
		raid_ctlr_position = 0;
	else
		raid_ctlr_position = nphysicals + nlogicals;

1917
	/* adjust our table of devices */
1918
	n_ext_target_devs = 0;
1919
	for (i = 0; i < nphysicals + nlogicals + 1; i++) {
1920
		u8 *lunaddrbytes, is_OBDR = 0;
1921 1922

		/* Figure out where the LUN ID info is coming from */
1923 1924
		lunaddrbytes = figure_lunaddrbytes(h, raid_ctlr_position,
			i, nphysicals, nlogicals, physdev_list, logdev_list);
1925
		/* skip masked physical devices. */
1926 1927
		if (lunaddrbytes[3] & 0xC0 &&
			i < nphysicals + (raid_ctlr_position == 0))
1928 1929 1930
			continue;

		/* Get device type, vendor, model, device id */
1931 1932
		if (hpsa_update_device_info(h, lunaddrbytes, tmpdevice,
							&is_OBDR))
1933
			continue; /* skip it if we can't talk to it. */
1934
		figure_bus_target_lun(h, lunaddrbytes, tmpdevice);
1935 1936 1937
		this_device = currentsd[ncurrent];

		/*
1938
		 * For external target devices, we have to insert a LUN 0 which
1939 1940 1941 1942 1943
		 * doesn't show up in CCISS_REPORT_PHYSICAL data, but there
		 * is nonetheless an enclosure device there.  We have to
		 * present that otherwise linux won't find anything if
		 * there is no lun 0.
		 */
1944
		if (add_ext_target_dev(h, tmpdevice, this_device,
1945
				lunaddrbytes, lunzerobits,
1946
				&n_ext_target_devs)) {
1947 1948 1949 1950 1951 1952 1953
			ncurrent++;
			this_device = currentsd[ncurrent];
		}

		*this_device = *tmpdevice;

		switch (this_device->devtype) {
1954
		case TYPE_ROM:
1955 1956 1957 1958 1959 1960 1961
			/* We don't *really* support actual CD-ROM devices,
			 * just "One Button Disaster Recovery" tape drive
			 * which temporarily pretends to be a CD-ROM drive.
			 * So we check that the device is really an OBDR tape
			 * device by checking for "$DR-10" in bytes 43-48 of
			 * the inquiry data.
			 */
1962 1963
			if (is_OBDR)
				ncurrent++;
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986
			break;
		case TYPE_DISK:
			if (i < nphysicals)
				break;
			ncurrent++;
			break;
		case TYPE_TAPE:
		case TYPE_MEDIUM_CHANGER:
			ncurrent++;
			break;
		case TYPE_RAID:
			/* Only present the Smartarray HBA as a RAID controller.
			 * If it's a RAID controller other than the HBA itself
			 * (an external RAID controller, MSA500 or similar)
			 * don't present it.
			 */
			if (!is_hba_lunid(lunaddrbytes))
				break;
			ncurrent++;
			break;
		default:
			break;
		}
1987
		if (ncurrent >= HPSA_MAX_DEVICES)
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
			break;
	}
	adjust_hpsa_scsi_table(h, hostno, currentsd, ncurrent);
out:
	kfree(tmpdevice);
	for (i = 0; i < ndev_allocated; i++)
		kfree(currentsd[i]);
	kfree(currentsd);
	kfree(physdev_list);
	kfree(logdev_list);
}

/* hpsa_scatter_gather takes a struct scsi_cmnd, (cmd), and does the pci
 * dma mapping  and fills in the scatter gather entries of the
 * hpsa command, cp.
 */
2004
static int hpsa_scatter_gather(struct ctlr_info *h,
2005 2006 2007 2008 2009
		struct CommandList *cp,
		struct scsi_cmnd *cmd)
{
	unsigned int len;
	struct scatterlist *sg;
2010
	u64 addr64;
2011 2012
	int use_sg, i, sg_index, chained;
	struct SGDescriptor *curr_sg;
2013

2014
	BUG_ON(scsi_sg_count(cmd) > h->maxsgentries);
2015 2016 2017 2018 2019 2020 2021 2022

	use_sg = scsi_dma_map(cmd);
	if (use_sg < 0)
		return use_sg;

	if (!use_sg)
		goto sglist_finished;

2023 2024 2025
	curr_sg = cp->SG;
	chained = 0;
	sg_index = 0;
2026
	scsi_for_each_sg(cmd, sg, use_sg, i) {
2027 2028 2029 2030 2031 2032
		if (i == h->max_cmd_sg_entries - 1 &&
			use_sg > h->max_cmd_sg_entries) {
			chained = 1;
			curr_sg = h->cmd_sg_list[cp->cmdindex];
			sg_index = 0;
		}
2033
		addr64 = (u64) sg_dma_address(sg);
2034
		len  = sg_dma_len(sg);
2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049
		curr_sg->Addr.lower = (u32) (addr64 & 0x0FFFFFFFFULL);
		curr_sg->Addr.upper = (u32) ((addr64 >> 32) & 0x0FFFFFFFFULL);
		curr_sg->Len = len;
		curr_sg->Ext = 0;  /* we are not chaining */
		curr_sg++;
	}

	if (use_sg + chained > h->maxSG)
		h->maxSG = use_sg + chained;

	if (chained) {
		cp->Header.SGList = h->max_cmd_sg_entries;
		cp->Header.SGTotal = (u16) (use_sg + 1);
		hpsa_map_sg_chain_block(h, cp);
		return 0;
2050 2051 2052 2053
	}

sglist_finished:

2054 2055
	cp->Header.SGList = (u8) use_sg;   /* no. SGs contig in this cmd */
	cp->Header.SGTotal = (u16) use_sg; /* total sgs in this cmd list */
2056 2057 2058 2059
	return 0;
}


J
Jeff Garzik 已提交
2060
static int hpsa_scsi_queue_command_lck(struct scsi_cmnd *cmd,
2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
	void (*done)(struct scsi_cmnd *))
{
	struct ctlr_info *h;
	struct hpsa_scsi_dev_t *dev;
	unsigned char scsi3addr[8];
	struct CommandList *c;
	unsigned long flags;

	/* Get the ptr to our adapter structure out of cmd->host. */
	h = sdev_to_hba(cmd->device);
	dev = cmd->device->hostdata;
	if (!dev) {
		cmd->result = DID_NO_CONNECT << 16;
		done(cmd);
		return 0;
	}
	memcpy(scsi3addr, dev->scsi3addr, sizeof(scsi3addr));

	spin_lock_irqsave(&h->lock, flags);
2080 2081 2082 2083 2084 2085 2086
	if (unlikely(h->lockup_detected)) {
		spin_unlock_irqrestore(&h->lock, flags);
		cmd->result = DID_ERROR << 16;
		done(cmd);
		return 0;
	}
	/* Need a lock as this is being allocated from the pool */
2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104
	c = cmd_alloc(h);
	spin_unlock_irqrestore(&h->lock, flags);
	if (c == NULL) {			/* trouble... */
		dev_err(&h->pdev->dev, "cmd_alloc returned NULL!\n");
		return SCSI_MLQUEUE_HOST_BUSY;
	}

	/* Fill in the command list header */

	cmd->scsi_done = done;    /* save this for use by completion code */

	/* save c in case we have to abort it  */
	cmd->host_scribble = (unsigned char *) c;

	c->cmd_type = CMD_SCSI;
	c->scsi_cmd = cmd;
	c->Header.ReplyQueue = 0;  /* unused in simple mode */
	memcpy(&c->Header.LUN.LunAddrBytes[0], &scsi3addr[0], 8);
2105 2106
	c->Header.Tag.lower = (c->cmdindex << DIRECT_LOOKUP_SHIFT);
	c->Header.Tag.lower |= DIRECT_LOOKUP_BIT;
2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150

	/* Fill in the request block... */

	c->Request.Timeout = 0;
	memset(c->Request.CDB, 0, sizeof(c->Request.CDB));
	BUG_ON(cmd->cmd_len > sizeof(c->Request.CDB));
	c->Request.CDBLen = cmd->cmd_len;
	memcpy(c->Request.CDB, cmd->cmnd, cmd->cmd_len);
	c->Request.Type.Type = TYPE_CMD;
	c->Request.Type.Attribute = ATTR_SIMPLE;
	switch (cmd->sc_data_direction) {
	case DMA_TO_DEVICE:
		c->Request.Type.Direction = XFER_WRITE;
		break;
	case DMA_FROM_DEVICE:
		c->Request.Type.Direction = XFER_READ;
		break;
	case DMA_NONE:
		c->Request.Type.Direction = XFER_NONE;
		break;
	case DMA_BIDIRECTIONAL:
		/* This can happen if a buggy application does a scsi passthru
		 * and sets both inlen and outlen to non-zero. ( see
		 * ../scsi/scsi_ioctl.c:scsi_ioctl_send_command() )
		 */

		c->Request.Type.Direction = XFER_RSVD;
		/* This is technically wrong, and hpsa controllers should
		 * reject it with CMD_INVALID, which is the most correct
		 * response, but non-fibre backends appear to let it
		 * slide by, and give the same results as if this field
		 * were set correctly.  Either way is acceptable for
		 * our purposes here.
		 */

		break;

	default:
		dev_err(&h->pdev->dev, "unknown data direction: %d\n",
			cmd->sc_data_direction);
		BUG();
		break;
	}

2151
	if (hpsa_scatter_gather(h, c, cmd) < 0) { /* Fill SG list */
2152 2153 2154 2155 2156 2157 2158 2159
		cmd_free(h, c);
		return SCSI_MLQUEUE_HOST_BUSY;
	}
	enqueue_cmd_and_start_io(h, c);
	/* the cmd'll come back via intr handler in complete_scsi_command()  */
	return 0;
}

J
Jeff Garzik 已提交
2160 2161
static DEF_SCSI_QCMD(hpsa_scsi_queue_command)

2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203
static void hpsa_scan_start(struct Scsi_Host *sh)
{
	struct ctlr_info *h = shost_to_hba(sh);
	unsigned long flags;

	/* wait until any scan already in progress is finished. */
	while (1) {
		spin_lock_irqsave(&h->scan_lock, flags);
		if (h->scan_finished)
			break;
		spin_unlock_irqrestore(&h->scan_lock, flags);
		wait_event(h->scan_wait_queue, h->scan_finished);
		/* Note: We don't need to worry about a race between this
		 * thread and driver unload because the midlayer will
		 * have incremented the reference count, so unload won't
		 * happen if we're in here.
		 */
	}
	h->scan_finished = 0; /* mark scan as in progress */
	spin_unlock_irqrestore(&h->scan_lock, flags);

	hpsa_update_scsi_devices(h, h->scsi_host->host_no);

	spin_lock_irqsave(&h->scan_lock, flags);
	h->scan_finished = 1; /* mark scan as finished. */
	wake_up_all(&h->scan_wait_queue);
	spin_unlock_irqrestore(&h->scan_lock, flags);
}

static int hpsa_scan_finished(struct Scsi_Host *sh,
	unsigned long elapsed_time)
{
	struct ctlr_info *h = shost_to_hba(sh);
	unsigned long flags;
	int finished;

	spin_lock_irqsave(&h->scan_lock, flags);
	finished = h->scan_finished;
	spin_unlock_irqrestore(&h->scan_lock, flags);
	return finished;
}

2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220
static int hpsa_change_queue_depth(struct scsi_device *sdev,
	int qdepth, int reason)
{
	struct ctlr_info *h = sdev_to_hba(sdev);

	if (reason != SCSI_QDEPTH_DEFAULT)
		return -ENOTSUPP;

	if (qdepth < 1)
		qdepth = 1;
	else
		if (qdepth > h->nr_cmds)
			qdepth = h->nr_cmds;
	scsi_adjust_queue_depth(sdev, scsi_get_tag_type(sdev), qdepth);
	return sdev->queue_depth;
}

2221 2222 2223 2224 2225 2226 2227 2228 2229 2230
static void hpsa_unregister_scsi(struct ctlr_info *h)
{
	/* we are being forcibly unloaded, and may not refuse. */
	scsi_remove_host(h->scsi_host);
	scsi_host_put(h->scsi_host);
	h->scsi_host = NULL;
}

static int hpsa_register_scsi(struct ctlr_info *h)
{
2231 2232
	struct Scsi_Host *sh;
	int error;
2233

2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266
	sh = scsi_host_alloc(&hpsa_driver_template, sizeof(h));
	if (sh == NULL)
		goto fail;

	sh->io_port = 0;
	sh->n_io_port = 0;
	sh->this_id = -1;
	sh->max_channel = 3;
	sh->max_cmd_len = MAX_COMMAND_SIZE;
	sh->max_lun = HPSA_MAX_LUN;
	sh->max_id = HPSA_MAX_LUN;
	sh->can_queue = h->nr_cmds;
	sh->cmd_per_lun = h->nr_cmds;
	sh->sg_tablesize = h->maxsgentries;
	h->scsi_host = sh;
	sh->hostdata[0] = (unsigned long) h;
	sh->irq = h->intr[h->intr_mode];
	sh->unique_id = sh->irq;
	error = scsi_add_host(sh, &h->pdev->dev);
	if (error)
		goto fail_host_put;
	scsi_scan_host(sh);
	return 0;

 fail_host_put:
	dev_err(&h->pdev->dev, "%s: scsi_add_host"
		" failed for controller %d\n", __func__, h->ctlr);
	scsi_host_put(sh);
	return error;
 fail:
	dev_err(&h->pdev->dev, "%s: scsi_host_alloc"
		" failed for controller %d\n", __func__, h->ctlr);
	return -ENOMEM;
2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343
}

static int wait_for_device_to_become_ready(struct ctlr_info *h,
	unsigned char lunaddr[])
{
	int rc = 0;
	int count = 0;
	int waittime = 1; /* seconds */
	struct CommandList *c;

	c = cmd_special_alloc(h);
	if (!c) {
		dev_warn(&h->pdev->dev, "out of memory in "
			"wait_for_device_to_become_ready.\n");
		return IO_ERROR;
	}

	/* Send test unit ready until device ready, or give up. */
	while (count < HPSA_TUR_RETRY_LIMIT) {

		/* Wait for a bit.  do this first, because if we send
		 * the TUR right away, the reset will just abort it.
		 */
		msleep(1000 * waittime);
		count++;

		/* Increase wait time with each try, up to a point. */
		if (waittime < HPSA_MAX_WAIT_INTERVAL_SECS)
			waittime = waittime * 2;

		/* Send the Test Unit Ready */
		fill_cmd(c, TEST_UNIT_READY, h, NULL, 0, 0, lunaddr, TYPE_CMD);
		hpsa_scsi_do_simple_cmd_core(h, c);
		/* no unmap needed here because no data xfer. */

		if (c->err_info->CommandStatus == CMD_SUCCESS)
			break;

		if (c->err_info->CommandStatus == CMD_TARGET_STATUS &&
			c->err_info->ScsiStatus == SAM_STAT_CHECK_CONDITION &&
			(c->err_info->SenseInfo[2] == NO_SENSE ||
			c->err_info->SenseInfo[2] == UNIT_ATTENTION))
			break;

		dev_warn(&h->pdev->dev, "waiting %d secs "
			"for device to become ready.\n", waittime);
		rc = 1; /* device not ready. */
	}

	if (rc)
		dev_warn(&h->pdev->dev, "giving up on device.\n");
	else
		dev_warn(&h->pdev->dev, "device is ready.\n");

	cmd_special_free(h, c);
	return rc;
}

/* Need at least one of these error handlers to keep ../scsi/hosts.c from
 * complaining.  Doing a host- or bus-reset can't do anything good here.
 */
static int hpsa_eh_device_reset_handler(struct scsi_cmnd *scsicmd)
{
	int rc;
	struct ctlr_info *h;
	struct hpsa_scsi_dev_t *dev;

	/* find the controller to which the command to be aborted was sent */
	h = sdev_to_hba(scsicmd->device);
	if (h == NULL) /* paranoia */
		return FAILED;
	dev = scsicmd->device->hostdata;
	if (!dev) {
		dev_err(&h->pdev->dev, "hpsa_eh_device_reset_handler: "
			"device lookup failed.\n");
		return FAILED;
	}
2344 2345
	dev_warn(&h->pdev->dev, "resetting device %d:%d:%d:%d\n",
		h->scsi_host->host_no, dev->bus, dev->target, dev->lun);
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386
	/* send a reset to the SCSI LUN which the command was sent to */
	rc = hpsa_send_reset(h, dev->scsi3addr);
	if (rc == 0 && wait_for_device_to_become_ready(h, dev->scsi3addr) == 0)
		return SUCCESS;

	dev_warn(&h->pdev->dev, "resetting device failed.\n");
	return FAILED;
}

/*
 * For operations that cannot sleep, a command block is allocated at init,
 * and managed by cmd_alloc() and cmd_free() using a simple bitmap to track
 * which ones are free or in use.  Lock must be held when calling this.
 * cmd_free() is the complement.
 */
static struct CommandList *cmd_alloc(struct ctlr_info *h)
{
	struct CommandList *c;
	int i;
	union u64bit temp64;
	dma_addr_t cmd_dma_handle, err_dma_handle;

	do {
		i = find_first_zero_bit(h->cmd_pool_bits, h->nr_cmds);
		if (i == h->nr_cmds)
			return NULL;
	} while (test_and_set_bit
		 (i & (BITS_PER_LONG - 1),
		  h->cmd_pool_bits + (i / BITS_PER_LONG)) != 0);
	c = h->cmd_pool + i;
	memset(c, 0, sizeof(*c));
	cmd_dma_handle = h->cmd_pool_dhandle
	    + i * sizeof(*c);
	c->err_info = h->errinfo_pool + i;
	memset(c->err_info, 0, sizeof(*c->err_info));
	err_dma_handle = h->errinfo_pool_dhandle
	    + i * sizeof(*c->err_info);
	h->nr_allocs++;

	c->cmdindex = i;

2387
	INIT_LIST_HEAD(&c->list);
2388 2389
	c->busaddr = (u32) cmd_dma_handle;
	temp64.val = (u64) err_dma_handle;
2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424
	c->ErrDesc.Addr.lower = temp64.val32.lower;
	c->ErrDesc.Addr.upper = temp64.val32.upper;
	c->ErrDesc.Len = sizeof(*c->err_info);

	c->h = h;
	return c;
}

/* For operations that can wait for kmalloc to possibly sleep,
 * this routine can be called. Lock need not be held to call
 * cmd_special_alloc. cmd_special_free() is the complement.
 */
static struct CommandList *cmd_special_alloc(struct ctlr_info *h)
{
	struct CommandList *c;
	union u64bit temp64;
	dma_addr_t cmd_dma_handle, err_dma_handle;

	c = pci_alloc_consistent(h->pdev, sizeof(*c), &cmd_dma_handle);
	if (c == NULL)
		return NULL;
	memset(c, 0, sizeof(*c));

	c->cmdindex = -1;

	c->err_info = pci_alloc_consistent(h->pdev, sizeof(*c->err_info),
		    &err_dma_handle);

	if (c->err_info == NULL) {
		pci_free_consistent(h->pdev,
			sizeof(*c), c, cmd_dma_handle);
		return NULL;
	}
	memset(c->err_info, 0, sizeof(*c->err_info));

2425
	INIT_LIST_HEAD(&c->list);
2426 2427
	c->busaddr = (u32) cmd_dma_handle;
	temp64.val = (u64) err_dma_handle;
2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454
	c->ErrDesc.Addr.lower = temp64.val32.lower;
	c->ErrDesc.Addr.upper = temp64.val32.upper;
	c->ErrDesc.Len = sizeof(*c->err_info);

	c->h = h;
	return c;
}

static void cmd_free(struct ctlr_info *h, struct CommandList *c)
{
	int i;

	i = c - h->cmd_pool;
	clear_bit(i & (BITS_PER_LONG - 1),
		  h->cmd_pool_bits + (i / BITS_PER_LONG));
	h->nr_frees++;
}

static void cmd_special_free(struct ctlr_info *h, struct CommandList *c)
{
	union u64bit temp64;

	temp64.val32.lower = c->ErrDesc.Addr.lower;
	temp64.val32.upper = c->ErrDesc.Addr.upper;
	pci_free_consistent(h->pdev, sizeof(*c->err_info),
			    c->err_info, (dma_addr_t) temp64.val);
	pci_free_consistent(h->pdev, sizeof(*c),
2455
			    c, (dma_addr_t) (c->busaddr & DIRECT_LOOKUP_MASK));
2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468
}

#ifdef CONFIG_COMPAT

static int hpsa_ioctl32_passthru(struct scsi_device *dev, int cmd, void *arg)
{
	IOCTL32_Command_struct __user *arg32 =
	    (IOCTL32_Command_struct __user *) arg;
	IOCTL_Command_struct arg64;
	IOCTL_Command_struct __user *p = compat_alloc_user_space(sizeof(arg64));
	int err;
	u32 cp;

2469
	memset(&arg64, 0, sizeof(arg64));
2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484
	err = 0;
	err |= copy_from_user(&arg64.LUN_info, &arg32->LUN_info,
			   sizeof(arg64.LUN_info));
	err |= copy_from_user(&arg64.Request, &arg32->Request,
			   sizeof(arg64.Request));
	err |= copy_from_user(&arg64.error_info, &arg32->error_info,
			   sizeof(arg64.error_info));
	err |= get_user(arg64.buf_size, &arg32->buf_size);
	err |= get_user(cp, &arg32->buf);
	arg64.buf = compat_ptr(cp);
	err |= copy_to_user(p, &arg64, sizeof(arg64));

	if (err)
		return -EFAULT;

2485
	err = hpsa_ioctl(dev, CCISS_PASSTHRU, (void *)p);
2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505
	if (err)
		return err;
	err |= copy_in_user(&arg32->error_info, &p->error_info,
			 sizeof(arg32->error_info));
	if (err)
		return -EFAULT;
	return err;
}

static int hpsa_ioctl32_big_passthru(struct scsi_device *dev,
	int cmd, void *arg)
{
	BIG_IOCTL32_Command_struct __user *arg32 =
	    (BIG_IOCTL32_Command_struct __user *) arg;
	BIG_IOCTL_Command_struct arg64;
	BIG_IOCTL_Command_struct __user *p =
	    compat_alloc_user_space(sizeof(arg64));
	int err;
	u32 cp;

2506
	memset(&arg64, 0, sizeof(arg64));
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522
	err = 0;
	err |= copy_from_user(&arg64.LUN_info, &arg32->LUN_info,
			   sizeof(arg64.LUN_info));
	err |= copy_from_user(&arg64.Request, &arg32->Request,
			   sizeof(arg64.Request));
	err |= copy_from_user(&arg64.error_info, &arg32->error_info,
			   sizeof(arg64.error_info));
	err |= get_user(arg64.buf_size, &arg32->buf_size);
	err |= get_user(arg64.malloc_size, &arg32->malloc_size);
	err |= get_user(cp, &arg32->buf);
	arg64.buf = compat_ptr(cp);
	err |= copy_to_user(p, &arg64, sizeof(arg64));

	if (err)
		return -EFAULT;

2523
	err = hpsa_ioctl(dev, CCISS_BIG_PASSTHRU, (void *)p);
2524 2525 2526 2527 2528 2529 2530 2531
	if (err)
		return err;
	err |= copy_in_user(&arg32->error_info, &p->error_info,
			 sizeof(arg32->error_info));
	if (err)
		return -EFAULT;
	return err;
}
2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561

static int hpsa_compat_ioctl(struct scsi_device *dev, int cmd, void *arg)
{
	switch (cmd) {
	case CCISS_GETPCIINFO:
	case CCISS_GETINTINFO:
	case CCISS_SETINTINFO:
	case CCISS_GETNODENAME:
	case CCISS_SETNODENAME:
	case CCISS_GETHEARTBEAT:
	case CCISS_GETBUSTYPES:
	case CCISS_GETFIRMVER:
	case CCISS_GETDRIVVER:
	case CCISS_REVALIDVOLS:
	case CCISS_DEREGDISK:
	case CCISS_REGNEWDISK:
	case CCISS_REGNEWD:
	case CCISS_RESCANDISK:
	case CCISS_GETLUNINFO:
		return hpsa_ioctl(dev, cmd, arg);

	case CCISS_PASSTHRU32:
		return hpsa_ioctl32_passthru(dev, cmd, arg);
	case CCISS_BIG_PASSTHRU32:
		return hpsa_ioctl32_big_passthru(dev, cmd, arg);

	default:
		return -ENOIOCTLCMD;
	}
}
2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622
#endif

static int hpsa_getpciinfo_ioctl(struct ctlr_info *h, void __user *argp)
{
	struct hpsa_pci_info pciinfo;

	if (!argp)
		return -EINVAL;
	pciinfo.domain = pci_domain_nr(h->pdev->bus);
	pciinfo.bus = h->pdev->bus->number;
	pciinfo.dev_fn = h->pdev->devfn;
	pciinfo.board_id = h->board_id;
	if (copy_to_user(argp, &pciinfo, sizeof(pciinfo)))
		return -EFAULT;
	return 0;
}

static int hpsa_getdrivver_ioctl(struct ctlr_info *h, void __user *argp)
{
	DriverVer_type DriverVer;
	unsigned char vmaj, vmin, vsubmin;
	int rc;

	rc = sscanf(HPSA_DRIVER_VERSION, "%hhu.%hhu.%hhu",
		&vmaj, &vmin, &vsubmin);
	if (rc != 3) {
		dev_info(&h->pdev->dev, "driver version string '%s' "
			"unrecognized.", HPSA_DRIVER_VERSION);
		vmaj = 0;
		vmin = 0;
		vsubmin = 0;
	}
	DriverVer = (vmaj << 16) | (vmin << 8) | vsubmin;
	if (!argp)
		return -EINVAL;
	if (copy_to_user(argp, &DriverVer, sizeof(DriverVer_type)))
		return -EFAULT;
	return 0;
}

static int hpsa_passthru_ioctl(struct ctlr_info *h, void __user *argp)
{
	IOCTL_Command_struct iocommand;
	struct CommandList *c;
	char *buff = NULL;
	union u64bit temp64;

	if (!argp)
		return -EINVAL;
	if (!capable(CAP_SYS_RAWIO))
		return -EPERM;
	if (copy_from_user(&iocommand, argp, sizeof(iocommand)))
		return -EFAULT;
	if ((iocommand.buf_size < 1) &&
	    (iocommand.Request.Type.Direction != XFER_NONE)) {
		return -EINVAL;
	}
	if (iocommand.buf_size > 0) {
		buff = kmalloc(iocommand.buf_size, GFP_KERNEL);
		if (buff == NULL)
			return -EFAULT;
2623 2624 2625 2626 2627 2628 2629 2630 2631
		if (iocommand.Request.Type.Direction == XFER_WRITE) {
			/* Copy the data into the buffer we created */
			if (copy_from_user(buff, iocommand.buf,
				iocommand.buf_size)) {
				kfree(buff);
				return -EFAULT;
			}
		} else {
			memset(buff, 0, iocommand.buf_size);
2632
		}
2633
	}
2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666
	c = cmd_special_alloc(h);
	if (c == NULL) {
		kfree(buff);
		return -ENOMEM;
	}
	/* Fill in the command type */
	c->cmd_type = CMD_IOCTL_PEND;
	/* Fill in Command Header */
	c->Header.ReplyQueue = 0; /* unused in simple mode */
	if (iocommand.buf_size > 0) {	/* buffer to fill */
		c->Header.SGList = 1;
		c->Header.SGTotal = 1;
	} else	{ /* no buffers to fill */
		c->Header.SGList = 0;
		c->Header.SGTotal = 0;
	}
	memcpy(&c->Header.LUN, &iocommand.LUN_info, sizeof(c->Header.LUN));
	/* use the kernel address the cmd block for tag */
	c->Header.Tag.lower = c->busaddr;

	/* Fill in Request block */
	memcpy(&c->Request, &iocommand.Request,
		sizeof(c->Request));

	/* Fill in the scatter gather information */
	if (iocommand.buf_size > 0) {
		temp64.val = pci_map_single(h->pdev, buff,
			iocommand.buf_size, PCI_DMA_BIDIRECTIONAL);
		c->SG[0].Addr.lower = temp64.val32.lower;
		c->SG[0].Addr.upper = temp64.val32.upper;
		c->SG[0].Len = iocommand.buf_size;
		c->SG[0].Ext = 0; /* we are not chaining*/
	}
2667
	hpsa_scsi_do_simple_cmd_core_if_no_lockup(h, c);
2668 2669
	if (iocommand.buf_size > 0)
		hpsa_pci_unmap(h->pdev, c, 1, PCI_DMA_BIDIRECTIONAL);
2670 2671 2672 2673 2674 2675 2676 2677 2678 2679
	check_ioctl_unit_attention(h, c);

	/* Copy the error information out */
	memcpy(&iocommand.error_info, c->err_info,
		sizeof(iocommand.error_info));
	if (copy_to_user(argp, &iocommand, sizeof(iocommand))) {
		kfree(buff);
		cmd_special_free(h, c);
		return -EFAULT;
	}
2680 2681
	if (iocommand.Request.Type.Direction == XFER_READ &&
		iocommand.buf_size > 0) {
2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703
		/* Copy the data out of the buffer we created */
		if (copy_to_user(iocommand.buf, buff, iocommand.buf_size)) {
			kfree(buff);
			cmd_special_free(h, c);
			return -EFAULT;
		}
	}
	kfree(buff);
	cmd_special_free(h, c);
	return 0;
}

static int hpsa_big_passthru_ioctl(struct ctlr_info *h, void __user *argp)
{
	BIG_IOCTL_Command_struct *ioc;
	struct CommandList *c;
	unsigned char **buff = NULL;
	int *buff_size = NULL;
	union u64bit temp64;
	BYTE sg_used = 0;
	int status = 0;
	int i;
2704 2705
	u32 left;
	u32 sz;
2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731
	BYTE __user *data_ptr;

	if (!argp)
		return -EINVAL;
	if (!capable(CAP_SYS_RAWIO))
		return -EPERM;
	ioc = (BIG_IOCTL_Command_struct *)
	    kmalloc(sizeof(*ioc), GFP_KERNEL);
	if (!ioc) {
		status = -ENOMEM;
		goto cleanup1;
	}
	if (copy_from_user(ioc, argp, sizeof(*ioc))) {
		status = -EFAULT;
		goto cleanup1;
	}
	if ((ioc->buf_size < 1) &&
	    (ioc->Request.Type.Direction != XFER_NONE)) {
		status = -EINVAL;
		goto cleanup1;
	}
	/* Check kmalloc limits  using all SGs */
	if (ioc->malloc_size > MAX_KMALLOC_SIZE) {
		status = -EINVAL;
		goto cleanup1;
	}
2732
	if (ioc->buf_size > ioc->malloc_size * SG_ENTRIES_IN_CMD) {
2733 2734 2735
		status = -EINVAL;
		goto cleanup1;
	}
2736
	buff = kzalloc(SG_ENTRIES_IN_CMD * sizeof(char *), GFP_KERNEL);
2737 2738 2739 2740
	if (!buff) {
		status = -ENOMEM;
		goto cleanup1;
	}
2741
	buff_size = kmalloc(SG_ENTRIES_IN_CMD * sizeof(int), GFP_KERNEL);
2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773
	if (!buff_size) {
		status = -ENOMEM;
		goto cleanup1;
	}
	left = ioc->buf_size;
	data_ptr = ioc->buf;
	while (left) {
		sz = (left > ioc->malloc_size) ? ioc->malloc_size : left;
		buff_size[sg_used] = sz;
		buff[sg_used] = kmalloc(sz, GFP_KERNEL);
		if (buff[sg_used] == NULL) {
			status = -ENOMEM;
			goto cleanup1;
		}
		if (ioc->Request.Type.Direction == XFER_WRITE) {
			if (copy_from_user(buff[sg_used], data_ptr, sz)) {
				status = -ENOMEM;
				goto cleanup1;
			}
		} else
			memset(buff[sg_used], 0, sz);
		left -= sz;
		data_ptr += sz;
		sg_used++;
	}
	c = cmd_special_alloc(h);
	if (c == NULL) {
		status = -ENOMEM;
		goto cleanup1;
	}
	c->cmd_type = CMD_IOCTL_PEND;
	c->Header.ReplyQueue = 0;
2774
	c->Header.SGList = c->Header.SGTotal = sg_used;
2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789
	memcpy(&c->Header.LUN, &ioc->LUN_info, sizeof(c->Header.LUN));
	c->Header.Tag.lower = c->busaddr;
	memcpy(&c->Request, &ioc->Request, sizeof(c->Request));
	if (ioc->buf_size > 0) {
		int i;
		for (i = 0; i < sg_used; i++) {
			temp64.val = pci_map_single(h->pdev, buff[i],
				    buff_size[i], PCI_DMA_BIDIRECTIONAL);
			c->SG[i].Addr.lower = temp64.val32.lower;
			c->SG[i].Addr.upper = temp64.val32.upper;
			c->SG[i].Len = buff_size[i];
			/* we are not chaining */
			c->SG[i].Ext = 0;
		}
	}
2790
	hpsa_scsi_do_simple_cmd_core_if_no_lockup(h, c);
2791 2792
	if (sg_used)
		hpsa_pci_unmap(h->pdev, c, sg_used, PCI_DMA_BIDIRECTIONAL);
2793 2794 2795 2796 2797 2798 2799 2800
	check_ioctl_unit_attention(h, c);
	/* Copy the error information out */
	memcpy(&ioc->error_info, c->err_info, sizeof(ioc->error_info));
	if (copy_to_user(argp, ioc, sizeof(*ioc))) {
		cmd_special_free(h, c);
		status = -EFAULT;
		goto cleanup1;
	}
2801
	if (ioc->Request.Type.Direction == XFER_READ && ioc->buf_size > 0) {
2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846
		/* Copy the data out of the buffer we created */
		BYTE __user *ptr = ioc->buf;
		for (i = 0; i < sg_used; i++) {
			if (copy_to_user(ptr, buff[i], buff_size[i])) {
				cmd_special_free(h, c);
				status = -EFAULT;
				goto cleanup1;
			}
			ptr += buff_size[i];
		}
	}
	cmd_special_free(h, c);
	status = 0;
cleanup1:
	if (buff) {
		for (i = 0; i < sg_used; i++)
			kfree(buff[i]);
		kfree(buff);
	}
	kfree(buff_size);
	kfree(ioc);
	return status;
}

static void check_ioctl_unit_attention(struct ctlr_info *h,
	struct CommandList *c)
{
	if (c->err_info->CommandStatus == CMD_TARGET_STATUS &&
			c->err_info->ScsiStatus != SAM_STAT_CHECK_CONDITION)
		(void) check_for_unit_attention(h, c);
}
/*
 * ioctl
 */
static int hpsa_ioctl(struct scsi_device *dev, int cmd, void *arg)
{
	struct ctlr_info *h;
	void __user *argp = (void __user *)arg;

	h = sdev_to_hba(dev);

	switch (cmd) {
	case CCISS_DEREGDISK:
	case CCISS_REGNEWDISK:
	case CCISS_REGNEWD:
2847
		hpsa_scan_start(h->scsi_host);
2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861
		return 0;
	case CCISS_GETPCIINFO:
		return hpsa_getpciinfo_ioctl(h, argp);
	case CCISS_GETDRIVVER:
		return hpsa_getdrivver_ioctl(h, argp);
	case CCISS_PASSTHRU:
		return hpsa_passthru_ioctl(h, argp);
	case CCISS_BIG_PASSTHRU:
		return hpsa_big_passthru_ioctl(h, argp);
	default:
		return -ENOTTY;
	}
}

2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881
static int __devinit hpsa_send_host_reset(struct ctlr_info *h,
	unsigned char *scsi3addr, u8 reset_type)
{
	struct CommandList *c;

	c = cmd_alloc(h);
	if (!c)
		return -ENOMEM;
	fill_cmd(c, HPSA_DEVICE_RESET_MSG, h, NULL, 0, 0,
		RAID_CTLR_LUNID, TYPE_MSG);
	c->Request.CDB[1] = reset_type; /* fill_cmd defaults to target reset */
	c->waiting = NULL;
	enqueue_cmd_and_start_io(h, c);
	/* Don't wait for completion, the reset won't complete.  Don't free
	 * the command either.  This is the last command we will send before
	 * re-initializing everything, so it doesn't matter and won't leak.
	 */
	return 0;
}

2882 2883
static void fill_cmd(struct CommandList *c, u8 cmd, struct ctlr_info *h,
	void *buff, size_t size, u8 page_code, unsigned char *scsi3addr,
2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937
	int cmd_type)
{
	int pci_dir = XFER_NONE;

	c->cmd_type = CMD_IOCTL_PEND;
	c->Header.ReplyQueue = 0;
	if (buff != NULL && size > 0) {
		c->Header.SGList = 1;
		c->Header.SGTotal = 1;
	} else {
		c->Header.SGList = 0;
		c->Header.SGTotal = 0;
	}
	c->Header.Tag.lower = c->busaddr;
	memcpy(c->Header.LUN.LunAddrBytes, scsi3addr, 8);

	c->Request.Type.Type = cmd_type;
	if (cmd_type == TYPE_CMD) {
		switch (cmd) {
		case HPSA_INQUIRY:
			/* are we trying to read a vital product page */
			if (page_code != 0) {
				c->Request.CDB[1] = 0x01;
				c->Request.CDB[2] = page_code;
			}
			c->Request.CDBLen = 6;
			c->Request.Type.Attribute = ATTR_SIMPLE;
			c->Request.Type.Direction = XFER_READ;
			c->Request.Timeout = 0;
			c->Request.CDB[0] = HPSA_INQUIRY;
			c->Request.CDB[4] = size & 0xFF;
			break;
		case HPSA_REPORT_LOG:
		case HPSA_REPORT_PHYS:
			/* Talking to controller so It's a physical command
			   mode = 00 target = 0.  Nothing to write.
			 */
			c->Request.CDBLen = 12;
			c->Request.Type.Attribute = ATTR_SIMPLE;
			c->Request.Type.Direction = XFER_READ;
			c->Request.Timeout = 0;
			c->Request.CDB[0] = cmd;
			c->Request.CDB[6] = (size >> 24) & 0xFF; /* MSB */
			c->Request.CDB[7] = (size >> 16) & 0xFF;
			c->Request.CDB[8] = (size >> 8) & 0xFF;
			c->Request.CDB[9] = size & 0xFF;
			break;
		case HPSA_CACHE_FLUSH:
			c->Request.CDBLen = 12;
			c->Request.Type.Attribute = ATTR_SIMPLE;
			c->Request.Type.Direction = XFER_WRITE;
			c->Request.Timeout = 0;
			c->Request.CDB[0] = BMIC_WRITE;
			c->Request.CDB[6] = BMIC_CACHE_FLUSH;
2938 2939
			c->Request.CDB[7] = (size >> 8) & 0xFF;
			c->Request.CDB[8] = size & 0xFF;
2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960
			break;
		case TEST_UNIT_READY:
			c->Request.CDBLen = 6;
			c->Request.Type.Attribute = ATTR_SIMPLE;
			c->Request.Type.Direction = XFER_NONE;
			c->Request.Timeout = 0;
			break;
		default:
			dev_warn(&h->pdev->dev, "unknown command 0x%c\n", cmd);
			BUG();
			return;
		}
	} else if (cmd_type == TYPE_MSG) {
		switch (cmd) {

		case  HPSA_DEVICE_RESET_MSG:
			c->Request.CDBLen = 16;
			c->Request.Type.Type =  1; /* It is a MSG not a CMD */
			c->Request.Type.Attribute = ATTR_SIMPLE;
			c->Request.Type.Direction = XFER_NONE;
			c->Request.Timeout = 0; /* Don't time out */
2961 2962
			memset(&c->Request.CDB[0], 0, sizeof(c->Request.CDB));
			c->Request.CDB[0] =  cmd;
2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019
			c->Request.CDB[1] = 0x03;  /* Reset target above */
			/* If bytes 4-7 are zero, it means reset the */
			/* LunID device */
			c->Request.CDB[4] = 0x00;
			c->Request.CDB[5] = 0x00;
			c->Request.CDB[6] = 0x00;
			c->Request.CDB[7] = 0x00;
		break;

		default:
			dev_warn(&h->pdev->dev, "unknown message type %d\n",
				cmd);
			BUG();
		}
	} else {
		dev_warn(&h->pdev->dev, "unknown command type %d\n", cmd_type);
		BUG();
	}

	switch (c->Request.Type.Direction) {
	case XFER_READ:
		pci_dir = PCI_DMA_FROMDEVICE;
		break;
	case XFER_WRITE:
		pci_dir = PCI_DMA_TODEVICE;
		break;
	case XFER_NONE:
		pci_dir = PCI_DMA_NONE;
		break;
	default:
		pci_dir = PCI_DMA_BIDIRECTIONAL;
	}

	hpsa_map_one(h->pdev, c, buff, size, pci_dir);

	return;
}

/*
 * Map (physical) PCI mem into (virtual) kernel space
 */
static void __iomem *remap_pci_mem(ulong base, ulong size)
{
	ulong page_base = ((ulong) base) & PAGE_MASK;
	ulong page_offs = ((ulong) base) - page_base;
	void __iomem *page_remapped = ioremap(page_base, page_offs + size);

	return page_remapped ? (page_remapped + page_offs) : NULL;
}

/* Takes cmds off the submission queue and sends them to the hardware,
 * then puts them on the queue of cmds waiting for completion.
 */
static void start_io(struct ctlr_info *h)
{
	struct CommandList *c;

3020 3021
	while (!list_empty(&h->reqQ)) {
		c = list_entry(h->reqQ.next, struct CommandList, list);
3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044
		/* can't do anything if fifo is full */
		if ((h->access.fifo_full(h))) {
			dev_warn(&h->pdev->dev, "fifo full\n");
			break;
		}

		/* Get the first entry from the Request Q */
		removeQ(c);
		h->Qdepth--;

		/* Tell the controller execute command */
		h->access.submit_command(h, c);

		/* Put job onto the completed Q */
		addQ(&h->cmpQ, c);
	}
}

static inline unsigned long get_next_completion(struct ctlr_info *h)
{
	return h->access.command_completed(h);
}

3045
static inline bool interrupt_pending(struct ctlr_info *h)
3046 3047 3048 3049 3050 3051
{
	return h->access.intr_pending(h);
}

static inline long interrupt_not_for_us(struct ctlr_info *h)
{
3052 3053
	return (h->access.intr_pending(h) == 0) ||
		(h->interrupts_enabled == 0);
3054 3055
}

3056 3057
static inline int bad_tag(struct ctlr_info *h, u32 tag_index,
	u32 raw_tag)
3058 3059 3060 3061 3062 3063 3064 3065
{
	if (unlikely(tag_index >= h->nr_cmds)) {
		dev_warn(&h->pdev->dev, "bad tag 0x%08x ignored.\n", raw_tag);
		return 1;
	}
	return 0;
}

3066
static inline void finish_cmd(struct CommandList *c, u32 raw_tag)
3067 3068 3069
{
	removeQ(c);
	if (likely(c->cmd_type == CMD_SCSI))
3070
		complete_scsi_command(c);
3071 3072 3073 3074
	else if (c->cmd_type == CMD_IOCTL_PEND)
		complete(c->waiting);
}

3075 3076 3077 3078 3079 3080 3081 3082 3083 3084
static inline u32 hpsa_tag_contains_index(u32 tag)
{
	return tag & DIRECT_LOOKUP_BIT;
}

static inline u32 hpsa_tag_to_index(u32 tag)
{
	return tag >> DIRECT_LOOKUP_SHIFT;
}

3085 3086

static inline u32 hpsa_tag_discard_error_bits(struct ctlr_info *h, u32 tag)
3087
{
3088 3089
#define HPSA_PERF_ERROR_BITS ((1 << DIRECT_LOOKUP_SHIFT) - 1)
#define HPSA_SIMPLE_ERROR_BITS 0x03
3090
	if (unlikely(!(h->transMethod & CFGTBL_Trans_Performant)))
3091 3092
		return tag & ~HPSA_SIMPLE_ERROR_BITS;
	return tag & ~HPSA_PERF_ERROR_BITS;
3093 3094
}

3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116
/* process completion of an indexed ("direct lookup") command */
static inline u32 process_indexed_cmd(struct ctlr_info *h,
	u32 raw_tag)
{
	u32 tag_index;
	struct CommandList *c;

	tag_index = hpsa_tag_to_index(raw_tag);
	if (bad_tag(h, tag_index, raw_tag))
		return next_command(h);
	c = h->cmd_pool + tag_index;
	finish_cmd(c, raw_tag);
	return next_command(h);
}

/* process completion of a non-indexed command */
static inline u32 process_nonindexed_cmd(struct ctlr_info *h,
	u32 raw_tag)
{
	u32 tag;
	struct CommandList *c = NULL;

3117
	tag = hpsa_tag_discard_error_bits(h, raw_tag);
3118
	list_for_each_entry(c, &h->cmpQ, list) {
3119 3120 3121 3122 3123 3124 3125 3126 3127
		if ((c->busaddr & 0xFFFFFFE0) == (tag & 0xFFFFFFE0)) {
			finish_cmd(c, raw_tag);
			return next_command(h);
		}
	}
	bad_tag(h, h->nr_cmds + 1, raw_tag);
	return next_command(h);
}

3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158
/* Some controllers, like p400, will give us one interrupt
 * after a soft reset, even if we turned interrupts off.
 * Only need to check for this in the hpsa_xxx_discard_completions
 * functions.
 */
static int ignore_bogus_interrupt(struct ctlr_info *h)
{
	if (likely(!reset_devices))
		return 0;

	if (likely(h->interrupts_enabled))
		return 0;

	dev_info(&h->pdev->dev, "Received interrupt while interrupts disabled "
		"(known firmware bug.)  Ignoring.\n");

	return 1;
}

static irqreturn_t hpsa_intx_discard_completions(int irq, void *dev_id)
{
	struct ctlr_info *h = dev_id;
	unsigned long flags;
	u32 raw_tag;

	if (ignore_bogus_interrupt(h))
		return IRQ_NONE;

	if (interrupt_not_for_us(h))
		return IRQ_NONE;
	spin_lock_irqsave(&h->lock, flags);
3159
	h->last_intr_timestamp = get_jiffies_64();
3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178
	while (interrupt_pending(h)) {
		raw_tag = get_next_completion(h);
		while (raw_tag != FIFO_EMPTY)
			raw_tag = next_command(h);
	}
	spin_unlock_irqrestore(&h->lock, flags);
	return IRQ_HANDLED;
}

static irqreturn_t hpsa_msix_discard_completions(int irq, void *dev_id)
{
	struct ctlr_info *h = dev_id;
	unsigned long flags;
	u32 raw_tag;

	if (ignore_bogus_interrupt(h))
		return IRQ_NONE;

	spin_lock_irqsave(&h->lock, flags);
3179
	h->last_intr_timestamp = get_jiffies_64();
3180 3181 3182 3183 3184 3185 3186
	raw_tag = get_next_completion(h);
	while (raw_tag != FIFO_EMPTY)
		raw_tag = next_command(h);
	spin_unlock_irqrestore(&h->lock, flags);
	return IRQ_HANDLED;
}

3187
static irqreturn_t do_hpsa_intr_intx(int irq, void *dev_id)
3188 3189 3190
{
	struct ctlr_info *h = dev_id;
	unsigned long flags;
3191
	u32 raw_tag;
3192 3193 3194

	if (interrupt_not_for_us(h))
		return IRQ_NONE;
3195
	spin_lock_irqsave(&h->lock, flags);
3196
	h->last_intr_timestamp = get_jiffies_64();
3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215
	while (interrupt_pending(h)) {
		raw_tag = get_next_completion(h);
		while (raw_tag != FIFO_EMPTY) {
			if (hpsa_tag_contains_index(raw_tag))
				raw_tag = process_indexed_cmd(h, raw_tag);
			else
				raw_tag = process_nonindexed_cmd(h, raw_tag);
		}
	}
	spin_unlock_irqrestore(&h->lock, flags);
	return IRQ_HANDLED;
}

static irqreturn_t do_hpsa_intr_msi(int irq, void *dev_id)
{
	struct ctlr_info *h = dev_id;
	unsigned long flags;
	u32 raw_tag;

3216
	spin_lock_irqsave(&h->lock, flags);
3217
	h->last_intr_timestamp = get_jiffies_64();
3218 3219 3220 3221 3222 3223
	raw_tag = get_next_completion(h);
	while (raw_tag != FIFO_EMPTY) {
		if (hpsa_tag_contains_index(raw_tag))
			raw_tag = process_indexed_cmd(h, raw_tag);
		else
			raw_tag = process_nonindexed_cmd(h, raw_tag);
3224 3225 3226 3227 3228
	}
	spin_unlock_irqrestore(&h->lock, flags);
	return IRQ_HANDLED;
}

3229 3230 3231 3232
/* Send a message CDB to the firmware. Careful, this only works
 * in simple mode, not performant mode due to the tag lookup.
 * We only ever use this immediately after a controller reset.
 */
3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297
static __devinit int hpsa_message(struct pci_dev *pdev, unsigned char opcode,
						unsigned char type)
{
	struct Command {
		struct CommandListHeader CommandHeader;
		struct RequestBlock Request;
		struct ErrDescriptor ErrorDescriptor;
	};
	struct Command *cmd;
	static const size_t cmd_sz = sizeof(*cmd) +
					sizeof(cmd->ErrorDescriptor);
	dma_addr_t paddr64;
	uint32_t paddr32, tag;
	void __iomem *vaddr;
	int i, err;

	vaddr = pci_ioremap_bar(pdev, 0);
	if (vaddr == NULL)
		return -ENOMEM;

	/* The Inbound Post Queue only accepts 32-bit physical addresses for the
	 * CCISS commands, so they must be allocated from the lower 4GiB of
	 * memory.
	 */
	err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
	if (err) {
		iounmap(vaddr);
		return -ENOMEM;
	}

	cmd = pci_alloc_consistent(pdev, cmd_sz, &paddr64);
	if (cmd == NULL) {
		iounmap(vaddr);
		return -ENOMEM;
	}

	/* This must fit, because of the 32-bit consistent DMA mask.  Also,
	 * although there's no guarantee, we assume that the address is at
	 * least 4-byte aligned (most likely, it's page-aligned).
	 */
	paddr32 = paddr64;

	cmd->CommandHeader.ReplyQueue = 0;
	cmd->CommandHeader.SGList = 0;
	cmd->CommandHeader.SGTotal = 0;
	cmd->CommandHeader.Tag.lower = paddr32;
	cmd->CommandHeader.Tag.upper = 0;
	memset(&cmd->CommandHeader.LUN.LunAddrBytes, 0, 8);

	cmd->Request.CDBLen = 16;
	cmd->Request.Type.Type = TYPE_MSG;
	cmd->Request.Type.Attribute = ATTR_HEADOFQUEUE;
	cmd->Request.Type.Direction = XFER_NONE;
	cmd->Request.Timeout = 0; /* Don't time out */
	cmd->Request.CDB[0] = opcode;
	cmd->Request.CDB[1] = type;
	memset(&cmd->Request.CDB[2], 0, 14); /* rest of the CDB is reserved */
	cmd->ErrorDescriptor.Addr.lower = paddr32 + sizeof(*cmd);
	cmd->ErrorDescriptor.Addr.upper = 0;
	cmd->ErrorDescriptor.Len = sizeof(struct ErrorInfo);

	writel(paddr32, vaddr + SA5_REQUEST_PORT_OFFSET);

	for (i = 0; i < HPSA_MSG_SEND_RETRY_LIMIT; i++) {
		tag = readl(vaddr + SA5_REPLY_PORT_OFFSET);
3298
		if ((tag & ~HPSA_SIMPLE_ERROR_BITS) == paddr32)
3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328
			break;
		msleep(HPSA_MSG_SEND_RETRY_INTERVAL_MSECS);
	}

	iounmap(vaddr);

	/* we leak the DMA buffer here ... no choice since the controller could
	 *  still complete the command.
	 */
	if (i == HPSA_MSG_SEND_RETRY_LIMIT) {
		dev_err(&pdev->dev, "controller message %02x:%02x timed out\n",
			opcode, type);
		return -ETIMEDOUT;
	}

	pci_free_consistent(pdev, cmd_sz, cmd, paddr64);

	if (tag & HPSA_ERROR_BIT) {
		dev_err(&pdev->dev, "controller message %02x:%02x failed\n",
			opcode, type);
		return -EIO;
	}

	dev_info(&pdev->dev, "controller message %02x:%02x succeeded\n",
		opcode, type);
	return 0;
}

#define hpsa_noop(p) hpsa_message(p, 3, 0)

3329
static int hpsa_controller_hard_reset(struct pci_dev *pdev,
3330
	void * __iomem vaddr, u32 use_doorbell)
3331 3332 3333 3334 3335 3336 3337 3338 3339 3340
{
	u16 pmcsr;
	int pos;

	if (use_doorbell) {
		/* For everything after the P600, the PCI power state method
		 * of resetting the controller doesn't work, so we have this
		 * other way using the doorbell register.
		 */
		dev_info(&pdev->dev, "using doorbell to reset controller\n");
3341
		writel(use_doorbell, vaddr + SA5_DOORBELL);
3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371
	} else { /* Try to do it the PCI power state way */

		/* Quoting from the Open CISS Specification: "The Power
		 * Management Control/Status Register (CSR) controls the power
		 * state of the device.  The normal operating state is D0,
		 * CSR=00h.  The software off state is D3, CSR=03h.  To reset
		 * the controller, place the interface device in D3 then to D0,
		 * this causes a secondary PCI reset which will reset the
		 * controller." */

		pos = pci_find_capability(pdev, PCI_CAP_ID_PM);
		if (pos == 0) {
			dev_err(&pdev->dev,
				"hpsa_reset_controller: "
				"PCI PM not supported\n");
			return -ENODEV;
		}
		dev_info(&pdev->dev, "using PCI PM to reset controller\n");
		/* enter the D3hot power management state */
		pci_read_config_word(pdev, pos + PCI_PM_CTRL, &pmcsr);
		pmcsr &= ~PCI_PM_CTRL_STATE_MASK;
		pmcsr |= PCI_D3hot;
		pci_write_config_word(pdev, pos + PCI_PM_CTRL, pmcsr);

		msleep(500);

		/* enter the D0 power management state */
		pmcsr &= ~PCI_PM_CTRL_STATE_MASK;
		pmcsr |= PCI_D0;
		pci_write_config_word(pdev, pos + PCI_PM_CTRL, pmcsr);
3372 3373 3374 3375 3376 3377 3378

		/*
		 * The P600 requires a small delay when changing states.
		 * Otherwise we may think the board did not reset and we bail.
		 * This for kdump only and is particular to the P600.
		 */
		msleep(500);
3379 3380 3381 3382
	}
	return 0;
}

3383 3384 3385
static __devinit void init_driver_version(char *driver_version, int len)
{
	memset(driver_version, 0, len);
3386
	strncpy(driver_version, HPSA " " HPSA_DRIVER_VERSION, len - 1);
3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435
}

static __devinit int write_driver_ver_to_cfgtable(
	struct CfgTable __iomem *cfgtable)
{
	char *driver_version;
	int i, size = sizeof(cfgtable->driver_version);

	driver_version = kmalloc(size, GFP_KERNEL);
	if (!driver_version)
		return -ENOMEM;

	init_driver_version(driver_version, size);
	for (i = 0; i < size; i++)
		writeb(driver_version[i], &cfgtable->driver_version[i]);
	kfree(driver_version);
	return 0;
}

static __devinit void read_driver_ver_from_cfgtable(
	struct CfgTable __iomem *cfgtable, unsigned char *driver_ver)
{
	int i;

	for (i = 0; i < sizeof(cfgtable->driver_version); i++)
		driver_ver[i] = readb(&cfgtable->driver_version[i]);
}

static __devinit int controller_reset_failed(
	struct CfgTable __iomem *cfgtable)
{

	char *driver_ver, *old_driver_ver;
	int rc, size = sizeof(cfgtable->driver_version);

	old_driver_ver = kmalloc(2 * size, GFP_KERNEL);
	if (!old_driver_ver)
		return -ENOMEM;
	driver_ver = old_driver_ver + size;

	/* After a reset, the 32 bytes of "driver version" in the cfgtable
	 * should have been changed, otherwise we know the reset failed.
	 */
	init_driver_version(old_driver_ver, size);
	read_driver_ver_from_cfgtable(cfgtable, driver_ver);
	rc = !memcmp(driver_ver, old_driver_ver, size);
	kfree(old_driver_ver);
	return rc;
}
3436
/* This does a hard reset of the controller using PCI power management
3437
 * states or the using the doorbell register.
3438
 */
3439
static __devinit int hpsa_kdump_hard_reset_controller(struct pci_dev *pdev)
3440
{
3441 3442 3443 3444 3445
	u64 cfg_offset;
	u32 cfg_base_addr;
	u64 cfg_base_addr_index;
	void __iomem *vaddr;
	unsigned long paddr;
3446
	u32 misc_fw_support;
3447
	int rc;
3448
	struct CfgTable __iomem *cfgtable;
3449
	u32 use_doorbell;
3450
	u32 board_id;
3451
	u16 command_register;
3452

3453 3454
	/* For controllers as old as the P600, this is very nearly
	 * the same thing as
3455 3456 3457 3458 3459 3460
	 *
	 * pci_save_state(pci_dev);
	 * pci_set_power_state(pci_dev, PCI_D3hot);
	 * pci_set_power_state(pci_dev, PCI_D0);
	 * pci_restore_state(pci_dev);
	 *
3461 3462 3463
	 * For controllers newer than the P600, the pci power state
	 * method of resetting doesn't work so we have another way
	 * using the doorbell register.
3464
	 */
3465

3466
	rc = hpsa_lookup_board_id(pdev, &board_id);
3467
	if (rc < 0 || !ctlr_is_resettable(board_id)) {
3468 3469 3470
		dev_warn(&pdev->dev, "Not resetting device.\n");
		return -ENODEV;
	}
3471 3472 3473 3474

	/* if controller is soft- but not hard resettable... */
	if (!ctlr_is_hard_resettable(board_id))
		return -ENOTSUPP; /* try soft reset later. */
3475

3476 3477 3478 3479 3480 3481 3482
	/* Save the PCI command register */
	pci_read_config_word(pdev, 4, &command_register);
	/* Turn the board off.  This is so that later pci_restore_state()
	 * won't turn the board on before the rest of config space is ready.
	 */
	pci_disable_device(pdev);
	pci_save_state(pdev);
3483

3484 3485 3486 3487 3488 3489 3490
	/* find the first memory BAR, so we can find the cfg table */
	rc = hpsa_pci_find_memory_BAR(pdev, &paddr);
	if (rc)
		return rc;
	vaddr = remap_pci_mem(paddr, 0x250);
	if (!vaddr)
		return -ENOMEM;
3491

3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502
	/* find cfgtable in order to check if reset via doorbell is supported */
	rc = hpsa_find_cfg_addrs(pdev, vaddr, &cfg_base_addr,
					&cfg_base_addr_index, &cfg_offset);
	if (rc)
		goto unmap_vaddr;
	cfgtable = remap_pci_mem(pci_resource_start(pdev,
		       cfg_base_addr_index) + cfg_offset, sizeof(*cfgtable));
	if (!cfgtable) {
		rc = -ENOMEM;
		goto unmap_vaddr;
	}
3503 3504 3505
	rc = write_driver_ver_to_cfgtable(cfgtable);
	if (rc)
		goto unmap_vaddr;
3506

3507 3508 3509
	/* If reset via doorbell register is supported, use that.
	 * There are two such methods.  Favor the newest method.
	 */
3510
	misc_fw_support = readl(&cfgtable->misc_fw_support);
3511 3512 3513 3514 3515 3516
	use_doorbell = misc_fw_support & MISC_FW_DOORBELL_RESET2;
	if (use_doorbell) {
		use_doorbell = DOORBELL_CTLR_RESET2;
	} else {
		use_doorbell = misc_fw_support & MISC_FW_DOORBELL_RESET;
		if (use_doorbell) {
3517 3518
			dev_warn(&pdev->dev, "Soft reset not supported. "
				"Firmware update is required.\n");
3519
			rc = -ENOTSUPP; /* try soft reset */
3520 3521 3522
			goto unmap_cfgtable;
		}
	}
3523

3524 3525 3526
	rc = hpsa_controller_hard_reset(pdev, vaddr, use_doorbell);
	if (rc)
		goto unmap_cfgtable;
3527

3528 3529 3530 3531 3532
	pci_restore_state(pdev);
	rc = pci_enable_device(pdev);
	if (rc) {
		dev_warn(&pdev->dev, "failed to enable device.\n");
		goto unmap_cfgtable;
3533
	}
3534
	pci_write_config_word(pdev, 4, command_register);
3535

3536 3537 3538 3539
	/* Some devices (notably the HP Smart Array 5i Controller)
	   need a little pause here */
	msleep(HPSA_POST_RESET_PAUSE_MSECS);

3540
	/* Wait for board to become not ready, then ready. */
3541
	dev_info(&pdev->dev, "Waiting for board to reset.\n");
3542
	rc = hpsa_wait_for_board_state(pdev, vaddr, BOARD_NOT_READY);
3543
	if (rc) {
3544
		dev_warn(&pdev->dev,
3545 3546 3547 3548 3549
			"failed waiting for board to reset."
			" Will try soft reset.\n");
		rc = -ENOTSUPP; /* Not expected, but try soft reset later */
		goto unmap_cfgtable;
	}
3550 3551 3552
	rc = hpsa_wait_for_board_state(pdev, vaddr, BOARD_READY);
	if (rc) {
		dev_warn(&pdev->dev,
3553 3554
			"failed waiting for board to become ready "
			"after hard reset\n");
3555 3556 3557
		goto unmap_cfgtable;
	}

3558 3559 3560 3561
	rc = controller_reset_failed(vaddr);
	if (rc < 0)
		goto unmap_cfgtable;
	if (rc) {
3562 3563 3564
		dev_warn(&pdev->dev, "Unable to successfully reset "
			"controller. Will try soft reset.\n");
		rc = -ENOTSUPP;
3565
	} else {
3566
		dev_info(&pdev->dev, "board ready after hard reset.\n");
3567 3568 3569 3570 3571 3572 3573 3574
	}

unmap_cfgtable:
	iounmap(cfgtable);

unmap_vaddr:
	iounmap(vaddr);
	return rc;
3575 3576 3577 3578 3579 3580 3581 3582 3583
}

/*
 *  We cannot read the structure directly, for portability we must use
 *   the io functions.
 *   This is for debug only.
 */
static void print_cfg_table(struct device *dev, struct CfgTable *tb)
{
3584
#ifdef HPSA_DEBUG
3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614
	int i;
	char temp_name[17];

	dev_info(dev, "Controller Configuration information\n");
	dev_info(dev, "------------------------------------\n");
	for (i = 0; i < 4; i++)
		temp_name[i] = readb(&(tb->Signature[i]));
	temp_name[4] = '\0';
	dev_info(dev, "   Signature = %s\n", temp_name);
	dev_info(dev, "   Spec Number = %d\n", readl(&(tb->SpecValence)));
	dev_info(dev, "   Transport methods supported = 0x%x\n",
	       readl(&(tb->TransportSupport)));
	dev_info(dev, "   Transport methods active = 0x%x\n",
	       readl(&(tb->TransportActive)));
	dev_info(dev, "   Requested transport Method = 0x%x\n",
	       readl(&(tb->HostWrite.TransportRequest)));
	dev_info(dev, "   Coalesce Interrupt Delay = 0x%x\n",
	       readl(&(tb->HostWrite.CoalIntDelay)));
	dev_info(dev, "   Coalesce Interrupt Count = 0x%x\n",
	       readl(&(tb->HostWrite.CoalIntCount)));
	dev_info(dev, "   Max outstanding commands = 0x%d\n",
	       readl(&(tb->CmdsOutMax)));
	dev_info(dev, "   Bus Types = 0x%x\n", readl(&(tb->BusTypes)));
	for (i = 0; i < 16; i++)
		temp_name[i] = readb(&(tb->ServerName[i]));
	temp_name[16] = '\0';
	dev_info(dev, "   Server Name = %s\n", temp_name);
	dev_info(dev, "   Heartbeat Counter = 0x%x\n\n\n",
		readl(&(tb->HeartBeat)));
#endif				/* HPSA_DEBUG */
3615
}
3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655

static int find_PCI_BAR_index(struct pci_dev *pdev, unsigned long pci_bar_addr)
{
	int i, offset, mem_type, bar_type;

	if (pci_bar_addr == PCI_BASE_ADDRESS_0)	/* looking for BAR zero? */
		return 0;
	offset = 0;
	for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) {
		bar_type = pci_resource_flags(pdev, i) & PCI_BASE_ADDRESS_SPACE;
		if (bar_type == PCI_BASE_ADDRESS_SPACE_IO)
			offset += 4;
		else {
			mem_type = pci_resource_flags(pdev, i) &
			    PCI_BASE_ADDRESS_MEM_TYPE_MASK;
			switch (mem_type) {
			case PCI_BASE_ADDRESS_MEM_TYPE_32:
			case PCI_BASE_ADDRESS_MEM_TYPE_1M:
				offset += 4;	/* 32 bit */
				break;
			case PCI_BASE_ADDRESS_MEM_TYPE_64:
				offset += 8;
				break;
			default:	/* reserved in PCI 2.2 */
				dev_warn(&pdev->dev,
				       "base address is invalid\n");
				return -1;
				break;
			}
		}
		if (offset == pci_bar_addr - PCI_BASE_ADDRESS_0)
			return i + 1;
	}
	return -1;
}

/* If MSI/MSI-X is supported by the kernel we will try to enable it on
 * controllers that are capable. If not, we use IO-APIC mode.
 */

3656
static void __devinit hpsa_interrupt_mode(struct ctlr_info *h)
3657 3658 3659 3660 3661 3662 3663 3664
{
#ifdef CONFIG_PCI_MSI
	int err;
	struct msix_entry hpsa_msix_entries[4] = { {0, 0}, {0, 1},
	{0, 2}, {0, 3}
	};

	/* Some boards advertise MSI but don't really support it */
3665 3666
	if ((h->board_id == 0x40700E11) || (h->board_id == 0x40800E11) ||
	    (h->board_id == 0x40820E11) || (h->board_id == 0x40830E11))
3667
		goto default_int_mode;
3668 3669 3670
	if (pci_find_capability(h->pdev, PCI_CAP_ID_MSIX)) {
		dev_info(&h->pdev->dev, "MSIX\n");
		err = pci_enable_msix(h->pdev, hpsa_msix_entries, 4);
3671 3672 3673 3674 3675 3676 3677 3678 3679
		if (!err) {
			h->intr[0] = hpsa_msix_entries[0].vector;
			h->intr[1] = hpsa_msix_entries[1].vector;
			h->intr[2] = hpsa_msix_entries[2].vector;
			h->intr[3] = hpsa_msix_entries[3].vector;
			h->msix_vector = 1;
			return;
		}
		if (err > 0) {
3680
			dev_warn(&h->pdev->dev, "only %d MSI-X vectors "
3681 3682 3683
			       "available\n", err);
			goto default_int_mode;
		} else {
3684
			dev_warn(&h->pdev->dev, "MSI-X init failed %d\n",
3685 3686 3687 3688
			       err);
			goto default_int_mode;
		}
	}
3689 3690 3691
	if (pci_find_capability(h->pdev, PCI_CAP_ID_MSI)) {
		dev_info(&h->pdev->dev, "MSI\n");
		if (!pci_enable_msi(h->pdev))
3692 3693
			h->msi_vector = 1;
		else
3694
			dev_warn(&h->pdev->dev, "MSI init failed\n");
3695 3696 3697 3698
	}
default_int_mode:
#endif				/* CONFIG_PCI_MSI */
	/* if we get here we're going to use the default interrupt mode */
3699
	h->intr[h->intr_mode] = h->pdev->irq;
3700 3701
}

3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715
static int __devinit hpsa_lookup_board_id(struct pci_dev *pdev, u32 *board_id)
{
	int i;
	u32 subsystem_vendor_id, subsystem_device_id;

	subsystem_vendor_id = pdev->subsystem_vendor;
	subsystem_device_id = pdev->subsystem_device;
	*board_id = ((subsystem_device_id << 16) & 0xffff0000) |
		    subsystem_vendor_id;

	for (i = 0; i < ARRAY_SIZE(products); i++)
		if (*board_id == products[i].board_id)
			return i;

3716 3717 3718
	if ((subsystem_vendor_id != PCI_VENDOR_ID_HP &&
		subsystem_vendor_id != PCI_VENDOR_ID_COMPAQ) ||
		!hpsa_allow_any) {
3719 3720 3721 3722 3723 3724 3725
		dev_warn(&pdev->dev, "unrecognized board ID: "
			"0x%08x, ignoring.\n", *board_id);
			return -ENODEV;
	}
	return ARRAY_SIZE(products) - 1; /* generic unknown smart array */
}

3726
static int __devinit hpsa_pci_find_memory_BAR(struct pci_dev *pdev,
3727 3728 3729 3730 3731
	unsigned long *memory_bar)
{
	int i;

	for (i = 0; i < DEVICE_COUNT_RESOURCE; i++)
3732
		if (pci_resource_flags(pdev, i) & IORESOURCE_MEM) {
3733
			/* addressing mode bits already removed */
3734 3735
			*memory_bar = pci_resource_start(pdev, i);
			dev_dbg(&pdev->dev, "memory BAR = %lx\n",
3736 3737 3738
				*memory_bar);
			return 0;
		}
3739
	dev_warn(&pdev->dev, "no memory BAR found\n");
3740 3741 3742
	return -ENODEV;
}

3743 3744
static int __devinit hpsa_wait_for_board_state(struct pci_dev *pdev,
	void __iomem *vaddr, int wait_for_ready)
3745
{
3746
	int i, iterations;
3747
	u32 scratchpad;
3748 3749 3750 3751
	if (wait_for_ready)
		iterations = HPSA_BOARD_READY_ITERATIONS;
	else
		iterations = HPSA_BOARD_NOT_READY_ITERATIONS;
3752

3753 3754 3755 3756 3757 3758 3759 3760 3761
	for (i = 0; i < iterations; i++) {
		scratchpad = readl(vaddr + SA5_SCRATCHPAD_OFFSET);
		if (wait_for_ready) {
			if (scratchpad == HPSA_FIRMWARE_READY)
				return 0;
		} else {
			if (scratchpad != HPSA_FIRMWARE_READY)
				return 0;
		}
3762 3763
		msleep(HPSA_BOARD_READY_POLL_INTERVAL_MSECS);
	}
3764
	dev_warn(&pdev->dev, "board not ready, timed out.\n");
3765 3766 3767
	return -ENODEV;
}

3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782
static int __devinit hpsa_find_cfg_addrs(struct pci_dev *pdev,
	void __iomem *vaddr, u32 *cfg_base_addr, u64 *cfg_base_addr_index,
	u64 *cfg_offset)
{
	*cfg_base_addr = readl(vaddr + SA5_CTCFG_OFFSET);
	*cfg_offset = readl(vaddr + SA5_CTMEM_OFFSET);
	*cfg_base_addr &= (u32) 0x0000ffff;
	*cfg_base_addr_index = find_PCI_BAR_index(pdev, *cfg_base_addr);
	if (*cfg_base_addr_index == -1) {
		dev_warn(&pdev->dev, "cannot find cfg_base_addr_index\n");
		return -ENODEV;
	}
	return 0;
}

3783
static int __devinit hpsa_find_cfgtables(struct ctlr_info *h)
3784
{
3785 3786 3787
	u64 cfg_offset;
	u32 cfg_base_addr;
	u64 cfg_base_addr_index;
3788
	u32 trans_offset;
3789
	int rc;
3790

3791 3792 3793 3794
	rc = hpsa_find_cfg_addrs(h->pdev, h->vaddr, &cfg_base_addr,
		&cfg_base_addr_index, &cfg_offset);
	if (rc)
		return rc;
3795
	h->cfgtable = remap_pci_mem(pci_resource_start(h->pdev,
3796
		       cfg_base_addr_index) + cfg_offset, sizeof(*h->cfgtable));
3797 3798
	if (!h->cfgtable)
		return -ENOMEM;
3799 3800 3801
	rc = write_driver_ver_to_cfgtable(h->cfgtable);
	if (rc)
		return rc;
3802
	/* Find performant mode table. */
3803
	trans_offset = readl(&h->cfgtable->TransMethodOffset);
3804 3805 3806 3807 3808 3809 3810 3811
	h->transtable = remap_pci_mem(pci_resource_start(h->pdev,
				cfg_base_addr_index)+cfg_offset+trans_offset,
				sizeof(*h->transtable));
	if (!h->transtable)
		return -ENOMEM;
	return 0;
}

3812 3813 3814
static void __devinit hpsa_get_max_perf_mode_cmds(struct ctlr_info *h)
{
	h->max_commands = readl(&(h->cfgtable->MaxPerformantModeCommands));
3815 3816 3817 3818 3819

	/* Limit commands in memory limited kdump scenario. */
	if (reset_devices && h->max_commands > 32)
		h->max_commands = 32;

3820 3821 3822 3823 3824 3825 3826 3827 3828
	if (h->max_commands < 16) {
		dev_warn(&h->pdev->dev, "Controller reports "
			"max supported commands of %d, an obvious lie. "
			"Using 16.  Ensure that firmware is up to date.\n",
			h->max_commands);
		h->max_commands = 16;
	}
}

3829 3830 3831 3832 3833 3834
/* Interrogate the hardware for some limits:
 * max commands, max SG elements without chaining, and with chaining,
 * SG chain block size, etc.
 */
static void __devinit hpsa_find_board_params(struct ctlr_info *h)
{
3835
	hpsa_get_max_perf_mode_cmds(h);
3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852
	h->nr_cmds = h->max_commands - 4; /* Allow room for some ioctls */
	h->maxsgentries = readl(&(h->cfgtable->MaxScatterGatherElements));
	/*
	 * Limit in-command s/g elements to 32 save dma'able memory.
	 * Howvever spec says if 0, use 31
	 */
	h->max_cmd_sg_entries = 31;
	if (h->maxsgentries > 512) {
		h->max_cmd_sg_entries = 32;
		h->chainsize = h->maxsgentries - h->max_cmd_sg_entries + 1;
		h->maxsgentries--; /* save one for chain pointer */
	} else {
		h->maxsgentries = 31; /* default to traditional values */
		h->chainsize = 0;
	}
}

3853 3854
static inline bool hpsa_CISS_signature_present(struct ctlr_info *h)
{
A
Akinobu Mita 已提交
3855
	if (!check_signature(h->cfgtable->Signature, "CISS", 4)) {
3856 3857 3858 3859 3860 3861
		dev_warn(&h->pdev->dev, "not a valid CISS config table\n");
		return false;
	}
	return true;
}

3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873
/* Need to enable prefetch in the SCSI core for 6400 in x86 */
static inline void hpsa_enable_scsi_prefetch(struct ctlr_info *h)
{
#ifdef CONFIG_X86
	u32 prefetch;

	prefetch = readl(&(h->cfgtable->SCSI_Prefetch));
	prefetch |= 0x100;
	writel(prefetch, &(h->cfgtable->SCSI_Prefetch));
#endif
}

3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887
/* Disable DMA prefetch for the P600.  Otherwise an ASIC bug may result
 * in a prefetch beyond physical memory.
 */
static inline void hpsa_p600_dma_prefetch_quirk(struct ctlr_info *h)
{
	u32 dma_prefetch;

	if (h->board_id != 0x3225103C)
		return;
	dma_prefetch = readl(h->vaddr + I2O_DMA1_CFG);
	dma_prefetch |= 0x8000;
	writel(dma_prefetch, h->vaddr + I2O_DMA1_CFG);
}

3888
static void __devinit hpsa_wait_for_mode_change_ack(struct ctlr_info *h)
3889 3890
{
	int i;
3891 3892
	u32 doorbell_value;
	unsigned long flags;
3893 3894 3895 3896 3897 3898

	/* under certain very rare conditions, this can take awhile.
	 * (e.g.: hot replace a failed 144GB drive in a RAID 5 set right
	 * as we enter this code.)
	 */
	for (i = 0; i < MAX_CONFIG_WAIT; i++) {
3899 3900 3901
		spin_lock_irqsave(&h->lock, flags);
		doorbell_value = readl(h->vaddr + SA5_DOORBELL);
		spin_unlock_irqrestore(&h->lock, flags);
D
Dan Carpenter 已提交
3902
		if (!(doorbell_value & CFGTBL_ChangeReq))
3903 3904
			break;
		/* delay and try again */
3905
		usleep_range(10000, 20000);
3906
	}
3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921
}

static int __devinit hpsa_enter_simple_mode(struct ctlr_info *h)
{
	u32 trans_support;

	trans_support = readl(&(h->cfgtable->TransportSupport));
	if (!(trans_support & SIMPLE_MODE))
		return -ENOTSUPP;

	h->max_commands = readl(&(h->cfgtable->CmdsOutMax));
	/* Update the field, and then ring the doorbell */
	writel(CFGTBL_Trans_Simple, &(h->cfgtable->HostWrite.TransportRequest));
	writel(CFGTBL_ChangeReq, h->vaddr + SA5_DOORBELL);
	hpsa_wait_for_mode_change_ack(h);
3922 3923 3924 3925 3926 3927
	print_cfg_table(&h->pdev->dev, h->cfgtable);
	if (!(readl(&(h->cfgtable->TransportActive)) & CFGTBL_Trans_Simple)) {
		dev_warn(&h->pdev->dev,
			"unable to get board into simple mode\n");
		return -ENODEV;
	}
3928
	h->transMethod = CFGTBL_Trans_Simple;
3929 3930 3931
	return 0;
}

3932 3933
static int __devinit hpsa_pci_init(struct ctlr_info *h)
{
3934
	int prod_index, err;
3935

3936 3937 3938 3939 3940
	prod_index = hpsa_lookup_board_id(h->pdev, &h->board_id);
	if (prod_index < 0)
		return -ENODEV;
	h->product_name = products[prod_index].product_name;
	h->access = *(products[prod_index].access);
3941

M
Matthew Garrett 已提交
3942 3943 3944
	pci_disable_link_state(h->pdev, PCIE_LINK_STATE_L0S |
			       PCIE_LINK_STATE_L1 | PCIE_LINK_STATE_CLKPM);

3945
	err = pci_enable_device(h->pdev);
3946
	if (err) {
3947
		dev_warn(&h->pdev->dev, "unable to enable PCI device\n");
3948 3949 3950
		return err;
	}

3951 3952 3953
	/* Enable bus mastering (pci_disable_device may disable this) */
	pci_set_master(h->pdev);

3954
	err = pci_request_regions(h->pdev, HPSA);
3955
	if (err) {
3956 3957
		dev_err(&h->pdev->dev,
			"cannot obtain PCI resources, aborting\n");
3958 3959
		return err;
	}
3960
	hpsa_interrupt_mode(h);
3961
	err = hpsa_pci_find_memory_BAR(h->pdev, &h->paddr);
3962
	if (err)
3963 3964
		goto err_out_free_res;
	h->vaddr = remap_pci_mem(h->paddr, 0x250);
3965 3966 3967 3968
	if (!h->vaddr) {
		err = -ENOMEM;
		goto err_out_free_res;
	}
3969
	err = hpsa_wait_for_board_state(h->pdev, h->vaddr, BOARD_READY);
3970
	if (err)
3971
		goto err_out_free_res;
3972 3973
	err = hpsa_find_cfgtables(h);
	if (err)
3974
		goto err_out_free_res;
3975
	hpsa_find_board_params(h);
3976

3977
	if (!hpsa_CISS_signature_present(h)) {
3978 3979 3980
		err = -ENODEV;
		goto err_out_free_res;
	}
3981
	hpsa_enable_scsi_prefetch(h);
3982
	hpsa_p600_dma_prefetch_quirk(h);
3983 3984
	err = hpsa_enter_simple_mode(h);
	if (err)
3985 3986 3987 3988
		goto err_out_free_res;
	return 0;

err_out_free_res:
3989 3990 3991 3992 3993 3994
	if (h->transtable)
		iounmap(h->transtable);
	if (h->cfgtable)
		iounmap(h->cfgtable);
	if (h->vaddr)
		iounmap(h->vaddr);
3995
	pci_disable_device(h->pdev);
3996
	pci_release_regions(h->pdev);
3997 3998 3999
	return err;
}

4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015
static void __devinit hpsa_hba_inquiry(struct ctlr_info *h)
{
	int rc;

#define HBA_INQUIRY_BYTE_COUNT 64
	h->hba_inquiry_data = kmalloc(HBA_INQUIRY_BYTE_COUNT, GFP_KERNEL);
	if (!h->hba_inquiry_data)
		return;
	rc = hpsa_scsi_do_inquiry(h, RAID_CTLR_LUNID, 0,
		h->hba_inquiry_data, HBA_INQUIRY_BYTE_COUNT);
	if (rc != 0) {
		kfree(h->hba_inquiry_data);
		h->hba_inquiry_data = NULL;
	}
}

4016 4017
static __devinit int hpsa_init_reset_devices(struct pci_dev *pdev)
{
4018
	int rc, i;
4019 4020 4021 4022

	if (!reset_devices)
		return 0;

4023 4024
	/* Reset the controller with a PCI power-cycle or via doorbell */
	rc = hpsa_kdump_hard_reset_controller(pdev);
4025

4026 4027
	/* -ENOTSUPP here means we cannot reset the controller
	 * but it's already (and still) up and running in
4028 4029
	 * "performant mode".  Or, it might be 640x, which can't reset
	 * due to concerns about shared bbwc between 6402/6404 pair.
4030 4031
	 */
	if (rc == -ENOTSUPP)
4032
		return rc; /* just try to do the kdump anyhow. */
4033 4034
	if (rc)
		return -ENODEV;
4035 4036

	/* Now try to get the controller to respond to a no-op */
4037
	dev_warn(&pdev->dev, "Waiting for controller to respond to no-op\n");
4038 4039 4040 4041 4042 4043 4044 4045 4046 4047
	for (i = 0; i < HPSA_POST_RESET_NOOP_RETRIES; i++) {
		if (hpsa_noop(pdev) == 0)
			break;
		else
			dev_warn(&pdev->dev, "no-op failed%s\n",
					(i < 11 ? "; re-trying" : ""));
	}
	return 0;
}

4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081
static __devinit int hpsa_allocate_cmd_pool(struct ctlr_info *h)
{
	h->cmd_pool_bits = kzalloc(
		DIV_ROUND_UP(h->nr_cmds, BITS_PER_LONG) *
		sizeof(unsigned long), GFP_KERNEL);
	h->cmd_pool = pci_alloc_consistent(h->pdev,
		    h->nr_cmds * sizeof(*h->cmd_pool),
		    &(h->cmd_pool_dhandle));
	h->errinfo_pool = pci_alloc_consistent(h->pdev,
		    h->nr_cmds * sizeof(*h->errinfo_pool),
		    &(h->errinfo_pool_dhandle));
	if ((h->cmd_pool_bits == NULL)
	    || (h->cmd_pool == NULL)
	    || (h->errinfo_pool == NULL)) {
		dev_err(&h->pdev->dev, "out of memory in %s", __func__);
		return -ENOMEM;
	}
	return 0;
}

static void hpsa_free_cmd_pool(struct ctlr_info *h)
{
	kfree(h->cmd_pool_bits);
	if (h->cmd_pool)
		pci_free_consistent(h->pdev,
			    h->nr_cmds * sizeof(struct CommandList),
			    h->cmd_pool, h->cmd_pool_dhandle);
	if (h->errinfo_pool)
		pci_free_consistent(h->pdev,
			    h->nr_cmds * sizeof(struct ErrorInfo),
			    h->errinfo_pool,
			    h->errinfo_pool_dhandle);
}

4082 4083 4084 4085 4086 4087 4088 4089
static int hpsa_request_irq(struct ctlr_info *h,
	irqreturn_t (*msixhandler)(int, void *),
	irqreturn_t (*intxhandler)(int, void *))
{
	int rc;

	if (h->msix_vector || h->msi_vector)
		rc = request_irq(h->intr[h->intr_mode], msixhandler,
4090
				0, h->devname, h);
4091 4092
	else
		rc = request_irq(h->intr[h->intr_mode], intxhandler,
4093
				IRQF_SHARED, h->devname, h);
4094 4095 4096 4097 4098 4099 4100 4101
	if (rc) {
		dev_err(&h->pdev->dev, "unable to get irq %d for %s\n",
		       h->intr[h->intr_mode], h->devname);
		return -ENODEV;
	}
	return 0;
}

4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149
static int __devinit hpsa_kdump_soft_reset(struct ctlr_info *h)
{
	if (hpsa_send_host_reset(h, RAID_CTLR_LUNID,
		HPSA_RESET_TYPE_CONTROLLER)) {
		dev_warn(&h->pdev->dev, "Resetting array controller failed.\n");
		return -EIO;
	}

	dev_info(&h->pdev->dev, "Waiting for board to soft reset.\n");
	if (hpsa_wait_for_board_state(h->pdev, h->vaddr, BOARD_NOT_READY)) {
		dev_warn(&h->pdev->dev, "Soft reset had no effect.\n");
		return -1;
	}

	dev_info(&h->pdev->dev, "Board reset, awaiting READY status.\n");
	if (hpsa_wait_for_board_state(h->pdev, h->vaddr, BOARD_READY)) {
		dev_warn(&h->pdev->dev, "Board failed to become ready "
			"after soft reset.\n");
		return -1;
	}

	return 0;
}

static void hpsa_undo_allocations_after_kdump_soft_reset(struct ctlr_info *h)
{
	free_irq(h->intr[h->intr_mode], h);
#ifdef CONFIG_PCI_MSI
	if (h->msix_vector)
		pci_disable_msix(h->pdev);
	else if (h->msi_vector)
		pci_disable_msi(h->pdev);
#endif /* CONFIG_PCI_MSI */
	hpsa_free_sg_chain_blocks(h);
	hpsa_free_cmd_pool(h);
	kfree(h->blockFetchTable);
	pci_free_consistent(h->pdev, h->reply_pool_size,
		h->reply_pool, h->reply_pool_dhandle);
	if (h->vaddr)
		iounmap(h->vaddr);
	if (h->transtable)
		iounmap(h->transtable);
	if (h->cfgtable)
		iounmap(h->cfgtable);
	pci_release_regions(h->pdev);
	kfree(h);
}

4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268
static void remove_ctlr_from_lockup_detector_list(struct ctlr_info *h)
{
	assert_spin_locked(&lockup_detector_lock);
	if (!hpsa_lockup_detector)
		return;
	if (h->lockup_detected)
		return; /* already stopped the lockup detector */
	list_del(&h->lockup_list);
}

/* Called when controller lockup detected. */
static void fail_all_cmds_on_list(struct ctlr_info *h, struct list_head *list)
{
	struct CommandList *c = NULL;

	assert_spin_locked(&h->lock);
	/* Mark all outstanding commands as failed and complete them. */
	while (!list_empty(list)) {
		c = list_entry(list->next, struct CommandList, list);
		c->err_info->CommandStatus = CMD_HARDWARE_ERR;
		finish_cmd(c, c->Header.Tag.lower);
	}
}

static void controller_lockup_detected(struct ctlr_info *h)
{
	unsigned long flags;

	assert_spin_locked(&lockup_detector_lock);
	remove_ctlr_from_lockup_detector_list(h);
	h->access.set_intr_mask(h, HPSA_INTR_OFF);
	spin_lock_irqsave(&h->lock, flags);
	h->lockup_detected = readl(h->vaddr + SA5_SCRATCHPAD_OFFSET);
	spin_unlock_irqrestore(&h->lock, flags);
	dev_warn(&h->pdev->dev, "Controller lockup detected: 0x%08x\n",
			h->lockup_detected);
	pci_disable_device(h->pdev);
	spin_lock_irqsave(&h->lock, flags);
	fail_all_cmds_on_list(h, &h->cmpQ);
	fail_all_cmds_on_list(h, &h->reqQ);
	spin_unlock_irqrestore(&h->lock, flags);
}

#define HEARTBEAT_SAMPLE_INTERVAL (10 * HZ)
#define HEARTBEAT_CHECK_MINIMUM_INTERVAL (HEARTBEAT_SAMPLE_INTERVAL / 2)

static void detect_controller_lockup(struct ctlr_info *h)
{
	u64 now;
	u32 heartbeat;
	unsigned long flags;

	assert_spin_locked(&lockup_detector_lock);
	now = get_jiffies_64();
	/* If we've received an interrupt recently, we're ok. */
	if (time_after64(h->last_intr_timestamp +
				(HEARTBEAT_CHECK_MINIMUM_INTERVAL), now))
		return;

	/*
	 * If we've already checked the heartbeat recently, we're ok.
	 * This could happen if someone sends us a signal. We
	 * otherwise don't care about signals in this thread.
	 */
	if (time_after64(h->last_heartbeat_timestamp +
				(HEARTBEAT_CHECK_MINIMUM_INTERVAL), now))
		return;

	/* If heartbeat has not changed since we last looked, we're not ok. */
	spin_lock_irqsave(&h->lock, flags);
	heartbeat = readl(&h->cfgtable->HeartBeat);
	spin_unlock_irqrestore(&h->lock, flags);
	if (h->last_heartbeat == heartbeat) {
		controller_lockup_detected(h);
		return;
	}

	/* We're ok. */
	h->last_heartbeat = heartbeat;
	h->last_heartbeat_timestamp = now;
}

static int detect_controller_lockup_thread(void *notused)
{
	struct ctlr_info *h;
	unsigned long flags;

	while (1) {
		struct list_head *this, *tmp;

		schedule_timeout_interruptible(HEARTBEAT_SAMPLE_INTERVAL);
		if (kthread_should_stop())
			break;
		spin_lock_irqsave(&lockup_detector_lock, flags);
		list_for_each_safe(this, tmp, &hpsa_ctlr_list) {
			h = list_entry(this, struct ctlr_info, lockup_list);
			detect_controller_lockup(h);
		}
		spin_unlock_irqrestore(&lockup_detector_lock, flags);
	}
	return 0;
}

static void add_ctlr_to_lockup_detector_list(struct ctlr_info *h)
{
	unsigned long flags;

	spin_lock_irqsave(&lockup_detector_lock, flags);
	list_add_tail(&h->lockup_list, &hpsa_ctlr_list);
	spin_unlock_irqrestore(&lockup_detector_lock, flags);
}

static void start_controller_lockup_detector(struct ctlr_info *h)
{
	/* Start the lockup detector thread if not already started */
	if (!hpsa_lockup_detector) {
		spin_lock_init(&lockup_detector_lock);
		hpsa_lockup_detector =
			kthread_run(detect_controller_lockup_thread,
4269
						NULL, HPSA);
4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286
	}
	if (!hpsa_lockup_detector) {
		dev_warn(&h->pdev->dev,
			"Could not start lockup detector thread\n");
		return;
	}
	add_ctlr_to_lockup_detector_list(h);
}

static void stop_controller_lockup_detector(struct ctlr_info *h)
{
	unsigned long flags;

	spin_lock_irqsave(&lockup_detector_lock, flags);
	remove_ctlr_from_lockup_detector_list(h);
	/* If the list of ctlr's to monitor is empty, stop the thread */
	if (list_empty(&hpsa_ctlr_list)) {
4287
		spin_unlock_irqrestore(&lockup_detector_lock, flags);
4288
		kthread_stop(hpsa_lockup_detector);
4289
		spin_lock_irqsave(&lockup_detector_lock, flags);
4290 4291 4292 4293 4294
		hpsa_lockup_detector = NULL;
	}
	spin_unlock_irqrestore(&lockup_detector_lock, flags);
}

4295 4296 4297
static int __devinit hpsa_init_one(struct pci_dev *pdev,
				    const struct pci_device_id *ent)
{
4298
	int dac, rc;
4299
	struct ctlr_info *h;
4300 4301
	int try_soft_reset = 0;
	unsigned long flags;
4302 4303 4304 4305

	if (number_of_controllers == 0)
		printk(KERN_INFO DRIVER_NAME "\n");

4306
	rc = hpsa_init_reset_devices(pdev);
4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319
	if (rc) {
		if (rc != -ENOTSUPP)
			return rc;
		/* If the reset fails in a particular way (it has no way to do
		 * a proper hard reset, so returns -ENOTSUPP) we can try to do
		 * a soft reset once we get the controller configured up to the
		 * point that it can accept a command.
		 */
		try_soft_reset = 1;
		rc = 0;
	}

reinit_after_soft_reset:
4320

4321 4322 4323 4324 4325 4326
	/* Command structures must be aligned on a 32-byte boundary because
	 * the 5 lower bits of the address are used by the hardware. and by
	 * the driver.  See comments in hpsa.h for more info.
	 */
#define COMMANDLIST_ALIGNMENT 32
	BUILD_BUG_ON(sizeof(struct CommandList) % COMMANDLIST_ALIGNMENT);
4327 4328
	h = kzalloc(sizeof(*h), GFP_KERNEL);
	if (!h)
4329
		return -ENOMEM;
4330

4331
	h->pdev = pdev;
4332
	h->intr_mode = hpsa_simple_mode ? SIMPLE_MODE_INT : PERF_MODE_INT;
4333 4334
	INIT_LIST_HEAD(&h->cmpQ);
	INIT_LIST_HEAD(&h->reqQ);
4335 4336
	spin_lock_init(&h->lock);
	spin_lock_init(&h->scan_lock);
4337
	rc = hpsa_pci_init(h);
4338
	if (rc != 0)
4339 4340
		goto clean1;

4341
	sprintf(h->devname, HPSA "%d", number_of_controllers);
4342 4343 4344 4345
	h->ctlr = number_of_controllers;
	number_of_controllers++;

	/* configure PCI DMA stuff */
4346 4347
	rc = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
	if (rc == 0) {
4348
		dac = 1;
4349 4350 4351 4352 4353 4354 4355 4356
	} else {
		rc = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
		if (rc == 0) {
			dac = 0;
		} else {
			dev_err(&pdev->dev, "no suitable DMA available\n");
			goto clean1;
		}
4357 4358 4359 4360
	}

	/* make sure the board interrupts are off */
	h->access.set_intr_mask(h, HPSA_INTR_OFF);
4361

4362
	if (hpsa_request_irq(h, do_hpsa_intr_msi, do_hpsa_intr_intx))
4363
		goto clean2;
4364 4365
	dev_info(&pdev->dev, "%s: <0x%x> at IRQ %d%s using DAC\n",
	       h->devname, pdev->device,
4366
	       h->intr[h->intr_mode], dac ? "" : " not");
4367
	if (hpsa_allocate_cmd_pool(h))
4368
		goto clean4;
4369 4370
	if (hpsa_allocate_sg_chain_blocks(h))
		goto clean4;
4371 4372
	init_waitqueue_head(&h->scan_wait_queue);
	h->scan_finished = 1; /* no scan currently in progress */
4373 4374

	pci_set_drvdata(pdev, h);
4375 4376 4377
	h->ndevices = 0;
	h->scsi_host = NULL;
	spin_lock_init(&h->devlock);
4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433
	hpsa_put_ctlr_into_performant_mode(h);

	/* At this point, the controller is ready to take commands.
	 * Now, if reset_devices and the hard reset didn't work, try
	 * the soft reset and see if that works.
	 */
	if (try_soft_reset) {

		/* This is kind of gross.  We may or may not get a completion
		 * from the soft reset command, and if we do, then the value
		 * from the fifo may or may not be valid.  So, we wait 10 secs
		 * after the reset throwing away any completions we get during
		 * that time.  Unregister the interrupt handler and register
		 * fake ones to scoop up any residual completions.
		 */
		spin_lock_irqsave(&h->lock, flags);
		h->access.set_intr_mask(h, HPSA_INTR_OFF);
		spin_unlock_irqrestore(&h->lock, flags);
		free_irq(h->intr[h->intr_mode], h);
		rc = hpsa_request_irq(h, hpsa_msix_discard_completions,
					hpsa_intx_discard_completions);
		if (rc) {
			dev_warn(&h->pdev->dev, "Failed to request_irq after "
				"soft reset.\n");
			goto clean4;
		}

		rc = hpsa_kdump_soft_reset(h);
		if (rc)
			/* Neither hard nor soft reset worked, we're hosed. */
			goto clean4;

		dev_info(&h->pdev->dev, "Board READY.\n");
		dev_info(&h->pdev->dev,
			"Waiting for stale completions to drain.\n");
		h->access.set_intr_mask(h, HPSA_INTR_ON);
		msleep(10000);
		h->access.set_intr_mask(h, HPSA_INTR_OFF);

		rc = controller_reset_failed(h->cfgtable);
		if (rc)
			dev_info(&h->pdev->dev,
				"Soft reset appears to have failed.\n");

		/* since the controller's reset, we have to go back and re-init
		 * everything.  Easiest to just forget what we've done and do it
		 * all over again.
		 */
		hpsa_undo_allocations_after_kdump_soft_reset(h);
		try_soft_reset = 0;
		if (rc)
			/* don't go to clean4, we already unallocated */
			return -ENODEV;

		goto reinit_after_soft_reset;
	}
4434 4435 4436 4437

	/* Turn the interrupts on so we can service requests */
	h->access.set_intr_mask(h, HPSA_INTR_ON);

4438
	hpsa_hba_inquiry(h);
4439
	hpsa_register_scsi(h);	/* hook ourselves into SCSI subsystem */
4440
	start_controller_lockup_detector(h);
4441 4442 4443
	return 1;

clean4:
4444
	hpsa_free_sg_chain_blocks(h);
4445
	hpsa_free_cmd_pool(h);
4446
	free_irq(h->intr[h->intr_mode], h);
4447 4448 4449
clean2:
clean1:
	kfree(h);
4450
	return rc;
4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488
}

static void hpsa_flush_cache(struct ctlr_info *h)
{
	char *flush_buf;
	struct CommandList *c;

	flush_buf = kzalloc(4, GFP_KERNEL);
	if (!flush_buf)
		return;

	c = cmd_special_alloc(h);
	if (!c) {
		dev_warn(&h->pdev->dev, "cmd_special_alloc returned NULL!\n");
		goto out_of_memory;
	}
	fill_cmd(c, HPSA_CACHE_FLUSH, h, flush_buf, 4, 0,
		RAID_CTLR_LUNID, TYPE_CMD);
	hpsa_scsi_do_simple_cmd_with_retry(h, c, PCI_DMA_TODEVICE);
	if (c->err_info->CommandStatus != 0)
		dev_warn(&h->pdev->dev,
			"error flushing cache on controller\n");
	cmd_special_free(h, c);
out_of_memory:
	kfree(flush_buf);
}

static void hpsa_shutdown(struct pci_dev *pdev)
{
	struct ctlr_info *h;

	h = pci_get_drvdata(pdev);
	/* Turn board interrupts off  and send the flush cache command
	 * sendcmd will turn off interrupt, and send the flush...
	 * To write all data in the battery backed cache to disks
	 */
	hpsa_flush_cache(h);
	h->access.set_intr_mask(h, HPSA_INTR_OFF);
4489
	free_irq(h->intr[h->intr_mode], h);
4490 4491 4492 4493 4494 4495 4496 4497
#ifdef CONFIG_PCI_MSI
	if (h->msix_vector)
		pci_disable_msix(h->pdev);
	else if (h->msi_vector)
		pci_disable_msi(h->pdev);
#endif				/* CONFIG_PCI_MSI */
}

4498 4499 4500 4501 4502 4503 4504 4505
static void __devexit hpsa_free_device_info(struct ctlr_info *h)
{
	int i;

	for (i = 0; i < h->ndevices; i++)
		kfree(h->dev[i]);
}

4506 4507 4508 4509 4510
static void __devexit hpsa_remove_one(struct pci_dev *pdev)
{
	struct ctlr_info *h;

	if (pci_get_drvdata(pdev) == NULL) {
4511
		dev_err(&pdev->dev, "unable to remove device\n");
4512 4513 4514
		return;
	}
	h = pci_get_drvdata(pdev);
4515
	stop_controller_lockup_detector(h);
4516 4517 4518
	hpsa_unregister_scsi(h);	/* unhook from SCSI subsystem */
	hpsa_shutdown(pdev);
	iounmap(h->vaddr);
4519 4520
	iounmap(h->transtable);
	iounmap(h->cfgtable);
4521
	hpsa_free_device_info(h);
4522
	hpsa_free_sg_chain_blocks(h);
4523 4524 4525 4526 4527 4528
	pci_free_consistent(h->pdev,
		h->nr_cmds * sizeof(struct CommandList),
		h->cmd_pool, h->cmd_pool_dhandle);
	pci_free_consistent(h->pdev,
		h->nr_cmds * sizeof(struct ErrorInfo),
		h->errinfo_pool, h->errinfo_pool_dhandle);
4529 4530
	pci_free_consistent(h->pdev, h->reply_pool_size,
		h->reply_pool, h->reply_pool_dhandle);
4531
	kfree(h->cmd_pool_bits);
4532
	kfree(h->blockFetchTable);
4533
	kfree(h->hba_inquiry_data);
4534
	pci_disable_device(pdev);
4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551
	pci_release_regions(pdev);
	pci_set_drvdata(pdev, NULL);
	kfree(h);
}

static int hpsa_suspend(__attribute__((unused)) struct pci_dev *pdev,
	__attribute__((unused)) pm_message_t state)
{
	return -ENOSYS;
}

static int hpsa_resume(__attribute__((unused)) struct pci_dev *pdev)
{
	return -ENOSYS;
}

static struct pci_driver hpsa_pci_driver = {
4552
	.name = HPSA,
4553 4554 4555 4556 4557 4558 4559 4560
	.probe = hpsa_init_one,
	.remove = __devexit_p(hpsa_remove_one),
	.id_table = hpsa_pci_device_id,	/* id_table */
	.shutdown = hpsa_shutdown,
	.suspend = hpsa_suspend,
	.resume = hpsa_resume,
};

4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597
/* Fill in bucket_map[], given nsgs (the max number of
 * scatter gather elements supported) and bucket[],
 * which is an array of 8 integers.  The bucket[] array
 * contains 8 different DMA transfer sizes (in 16
 * byte increments) which the controller uses to fetch
 * commands.  This function fills in bucket_map[], which
 * maps a given number of scatter gather elements to one of
 * the 8 DMA transfer sizes.  The point of it is to allow the
 * controller to only do as much DMA as needed to fetch the
 * command, with the DMA transfer size encoded in the lower
 * bits of the command address.
 */
static void  calc_bucket_map(int bucket[], int num_buckets,
	int nsgs, int *bucket_map)
{
	int i, j, b, size;

	/* even a command with 0 SGs requires 4 blocks */
#define MINIMUM_TRANSFER_BLOCKS 4
#define NUM_BUCKETS 8
	/* Note, bucket_map must have nsgs+1 entries. */
	for (i = 0; i <= nsgs; i++) {
		/* Compute size of a command with i SG entries */
		size = i + MINIMUM_TRANSFER_BLOCKS;
		b = num_buckets; /* Assume the biggest bucket */
		/* Find the bucket that is just big enough */
		for (j = 0; j < 8; j++) {
			if (bucket[j] >= size) {
				b = j;
				break;
			}
		}
		/* for a command with i SG entries, use bucket b. */
		bucket_map[i] = b;
	}
}

4598 4599
static __devinit void hpsa_enter_performant_mode(struct ctlr_info *h,
	u32 use_short_tags)
4600
{
4601 4602
	int i;
	unsigned long register_value;
4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613

	/* This is a bit complicated.  There are 8 registers on
	 * the controller which we write to to tell it 8 different
	 * sizes of commands which there may be.  It's a way of
	 * reducing the DMA done to fetch each command.  Encoded into
	 * each command's tag are 3 bits which communicate to the controller
	 * which of the eight sizes that command fits within.  The size of
	 * each command depends on how many scatter gather entries there are.
	 * Each SG entry requires 16 bytes.  The eight registers are programmed
	 * with the number of 16-byte blocks a command of that size requires.
	 * The smallest command possible requires 5 such 16 byte blocks.
4614
	 * the largest command possible requires SG_ENTRIES_IN_CMD + 4 16-byte
4615 4616 4617 4618 4619 4620
	 * blocks.  Note, this only extends to the SG entries contained
	 * within the command block, and does not extend to chained blocks
	 * of SG elements.   bft[] contains the eight values we write to
	 * the registers.  They are not evenly distributed, but have more
	 * sizes for small commands, and fewer sizes for larger commands.
	 */
4621 4622
	int bft[8] = {5, 6, 8, 10, 12, 20, 28, SG_ENTRIES_IN_CMD + 4};
	BUILD_BUG_ON(28 > SG_ENTRIES_IN_CMD + 4);
4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634
	/*  5 = 1 s/g entry or 4k
	 *  6 = 2 s/g entry or 8k
	 *  8 = 4 s/g entry or 16k
	 * 10 = 6 s/g entry or 24k
	 */

	h->reply_pool_wraparound = 1; /* spec: init to 1 */

	/* Controller spec: zero out this buffer. */
	memset(h->reply_pool, 0, h->reply_pool_size);
	h->reply_pool_head = h->reply_pool;

4635 4636 4637
	bft[7] = SG_ENTRIES_IN_CMD + 4;
	calc_bucket_map(bft, ARRAY_SIZE(bft),
				SG_ENTRIES_IN_CMD, h->blockFetchTable);
4638 4639 4640 4641 4642 4643 4644 4645 4646 4647
	for (i = 0; i < 8; i++)
		writel(bft[i], &h->transtable->BlockFetch[i]);

	/* size of controller ring buffer */
	writel(h->max_commands, &h->transtable->RepQSize);
	writel(1, &h->transtable->RepQCount);
	writel(0, &h->transtable->RepQCtrAddrLow32);
	writel(0, &h->transtable->RepQCtrAddrHigh32);
	writel(h->reply_pool_dhandle, &h->transtable->RepQAddr0Low32);
	writel(0, &h->transtable->RepQAddr0High32);
4648
	writel(CFGTBL_Trans_Performant | use_short_tags,
4649 4650
		&(h->cfgtable->HostWrite.TransportRequest));
	writel(CFGTBL_ChangeReq, h->vaddr + SA5_DOORBELL);
4651
	hpsa_wait_for_mode_change_ack(h);
4652 4653 4654 4655 4656 4657
	register_value = readl(&(h->cfgtable->TransportActive));
	if (!(register_value & CFGTBL_Trans_Performant)) {
		dev_warn(&h->pdev->dev, "unable to get board into"
					" performant mode\n");
		return;
	}
4658 4659 4660
	/* Change the access methods to the performant access methods */
	h->access = SA5_performant_access;
	h->transMethod = CFGTBL_Trans_Performant;
4661 4662 4663 4664 4665 4666
}

static __devinit void hpsa_put_ctlr_into_performant_mode(struct ctlr_info *h)
{
	u32 trans_support;

4667 4668 4669
	if (hpsa_simple_mode)
		return;

4670 4671 4672 4673
	trans_support = readl(&(h->cfgtable->TransportSupport));
	if (!(trans_support & PERFORMANT_MODE))
		return;

4674
	hpsa_get_max_perf_mode_cmds(h);
4675 4676 4677 4678 4679 4680
	/* Performant mode ring buffer and supporting data structures */
	h->reply_pool_size = h->max_commands * sizeof(u64);
	h->reply_pool = pci_alloc_consistent(h->pdev, h->reply_pool_size,
				&(h->reply_pool_dhandle));

	/* Need a block fetch table for performant mode */
4681
	h->blockFetchTable = kmalloc(((SG_ENTRIES_IN_CMD + 1) *
4682 4683 4684 4685 4686 4687
				sizeof(u32)), GFP_KERNEL);

	if ((h->reply_pool == NULL)
		|| (h->blockFetchTable == NULL))
		goto clean_up;

4688 4689
	hpsa_enter_performant_mode(h,
		trans_support & CFGTBL_Trans_use_short_tags);
4690 4691 4692 4693 4694 4695 4696 4697 4698 4699

	return;

clean_up:
	if (h->reply_pool)
		pci_free_consistent(h->pdev, h->reply_pool_size,
			h->reply_pool, h->reply_pool_dhandle);
	kfree(h->blockFetchTable);
}

4700 4701 4702 4703 4704 4705
/*
 *  This is it.  Register the PCI driver information for the cards we control
 *  the OS will call our registered routines when it finds one of our cards.
 */
static int __init hpsa_init(void)
{
M
Mike Miller 已提交
4706
	return pci_register_driver(&hpsa_pci_driver);
4707 4708 4709 4710 4711 4712 4713 4714 4715
}

static void __exit hpsa_cleanup(void)
{
	pci_unregister_driver(&hpsa_pci_driver);
}

module_init(hpsa_init);
module_exit(hpsa_cleanup);