opal.c 20.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11
/*
 * PowerNV OPAL high level interfaces
 *
 * Copyright 2011 IBM Corp.
 *
 * 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; either version
 * 2 of the License, or (at your option) any later version.
 */

12
#define pr_fmt(fmt)	"opal: " fmt
13

14
#include <linux/printk.h>
15 16
#include <linux/types.h>
#include <linux/of.h>
R
Rob Herring 已提交
17
#include <linux/of_fdt.h>
18
#include <linux/of_platform.h>
19
#include <linux/interrupt.h>
20
#include <linux/notifier.h>
21
#include <linux/slab.h>
22
#include <linux/sched.h>
23
#include <linux/kobject.h>
24
#include <linux/delay.h>
25
#include <linux/memblock.h>
26 27
#include <linux/kthread.h>
#include <linux/freezer.h>
28 29

#include <asm/machdep.h>
30 31
#include <asm/opal.h>
#include <asm/firmware.h>
32
#include <asm/mce.h>
33 34 35

#include "powernv.h"

36 37 38
/* /sys/firmware/opal */
struct kobject *opal_kobj;

39 40 41
struct opal {
	u64 base;
	u64 entry;
42
	u64 size;
43 44
} opal;

45 46 47 48 49 50 51 52 53
struct mcheck_recoverable_range {
	u64 start_addr;
	u64 end_addr;
	u64 recover_addr;
};

static struct mcheck_recoverable_range *mc_recoverable_range;
static int mc_recoverable_range_len;

54
struct device_node *opal_node;
55
static DEFINE_SPINLOCK(opal_write_lock);
56
static struct atomic_notifier_head opal_msg_notifier_head[OPAL_MSG_TYPE_MAX];
57
static uint32_t opal_heartbeat;
58

59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
static void opal_reinit_cores(void)
{
	/* Do the actual re-init, This will clobber all FPRs, VRs, etc...
	 *
	 * It will preserve non volatile GPRs and HSPRG0/1. It will
	 * also restore HIDs and other SPRs to their original value
	 * but it might clobber a bunch.
	 */
#ifdef __BIG_ENDIAN__
	opal_reinit_cpus(OPAL_REINIT_CPUS_HILE_BE);
#else
	opal_reinit_cpus(OPAL_REINIT_CPUS_HILE_LE);
#endif
}

74 75 76
int __init early_init_dt_scan_opal(unsigned long node,
				   const char *uname, int depth, void *data)
{
77
	const void *basep, *entryp, *sizep;
78
	int basesz, entrysz, runtimesz;
79 80 81 82 83 84

	if (depth != 1 || strcmp(uname, "ibm,opal") != 0)
		return 0;

	basep  = of_get_flat_dt_prop(node, "opal-base-address", &basesz);
	entryp = of_get_flat_dt_prop(node, "opal-entry-address", &entrysz);
85
	sizep = of_get_flat_dt_prop(node, "opal-runtime-size", &runtimesz);
86

87
	if (!basep || !entryp || !sizep)
88 89 90 91
		return 1;

	opal.base = of_read_number(basep, basesz/4);
	opal.entry = of_read_number(entryp, entrysz/4);
92
	opal.size = of_read_number(sizep, runtimesz/4);
93

94
	pr_debug("OPAL Base  = 0x%llx (basep=%p basesz=%d)\n",
95
		 opal.base, basep, basesz);
96
	pr_debug("OPAL Entry = 0x%llx (entryp=%p basesz=%d)\n",
97
		 opal.entry, entryp, entrysz);
98
	pr_debug("OPAL Entry = 0x%llx (sizep=%p runtimesz=%d)\n",
99
		 opal.size, sizep, runtimesz);
100 101

	powerpc_firmware_features |= FW_FEATURE_OPAL;
102 103 104
	if (of_flat_dt_is_compatible(node, "ibm,opal-v3")) {
		powerpc_firmware_features |= FW_FEATURE_OPALv2;
		powerpc_firmware_features |= FW_FEATURE_OPALv3;
105
		pr_info("OPAL V3 detected !\n");
106
	} else if (of_flat_dt_is_compatible(node, "ibm,opal-v2")) {
107
		powerpc_firmware_features |= FW_FEATURE_OPALv2;
108
		pr_info("OPAL V2 detected !\n");
109
	} else {
110
		pr_info("OPAL V1 detected !\n");
111 112
	}

113 114 115 116 117 118 119
	/* Reinit all cores with the right endian */
	opal_reinit_cores();

	/* Restore some bits */
	if (cur_cpu_spec->cpu_restore)
		cur_cpu_spec->cpu_restore();

120 121 122
	return 1;
}

123 124 125
int __init early_init_dt_scan_recoverable_ranges(unsigned long node,
				   const char *uname, int depth, void *data)
{
126
	int i, psize, size;
127 128 129 130 131
	const __be32 *prop;

	if (depth != 1 || strcmp(uname, "ibm,opal") != 0)
		return 0;

132
	prop = of_get_flat_dt_prop(node, "mcheck-recoverable-ranges", &psize);
133 134 135 136 137 138

	if (!prop)
		return 1;

	pr_debug("Found machine check recoverable ranges.\n");

139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
	/*
	 * Calculate number of available entries.
	 *
	 * Each recoverable address range entry is (start address, len,
	 * recovery address), 2 cells each for start and recovery address,
	 * 1 cell for len, totalling 5 cells per entry.
	 */
	mc_recoverable_range_len = psize / (sizeof(*prop) * 5);

	/* Sanity check */
	if (!mc_recoverable_range_len)
		return 1;

	/* Size required to hold all the entries. */
	size = mc_recoverable_range_len *
			sizeof(struct mcheck_recoverable_range);

156 157 158 159 160 161 162 163 164
	/*
	 * Allocate a buffer to hold the MC recoverable ranges. We would be
	 * accessing them in real mode, hence it needs to be within
	 * RMO region.
	 */
	mc_recoverable_range =__va(memblock_alloc_base(size, __alignof__(u64),
							ppc64_rma_size));
	memset(mc_recoverable_range, 0, size);

165
	for (i = 0; i < mc_recoverable_range_len; i++) {
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181
		mc_recoverable_range[i].start_addr =
					of_read_number(prop + (i * 5) + 0, 2);
		mc_recoverable_range[i].end_addr =
					mc_recoverable_range[i].start_addr +
					of_read_number(prop + (i * 5) + 2, 1);
		mc_recoverable_range[i].recover_addr =
					of_read_number(prop + (i * 5) + 3, 2);

		pr_debug("Machine check recoverable range: %llx..%llx: %llx\n",
				mc_recoverable_range[i].start_addr,
				mc_recoverable_range[i].end_addr,
				mc_recoverable_range[i].recover_addr);
	}
	return 1;
}

182 183
static int __init opal_register_exception_handlers(void)
{
184
#ifdef __BIG_ENDIAN__
185 186 187 188 189
	u64 glue;

	if (!(powerpc_firmware_features & FW_FEATURE_OPAL))
		return -ENODEV;

190 191
	/* Hookup some exception handlers except machine check. We use the
	 * fwnmi area at 0x7000 to provide the glue space to OPAL
192 193
	 */
	glue = 0x7000;
194 195 196 197 198 199 200 201 202 203 204 205 206 207

	/*
	 * Check if we are running on newer firmware that exports
	 * OPAL_HANDLE_HMI token. If yes, then don't ask OPAL to patch
	 * the HMI interrupt and we catch it directly in Linux.
	 *
	 * For older firmware (i.e currently released POWER8 System Firmware
	 * as of today <= SV810_087), we fallback to old behavior and let OPAL
	 * patch the HMI vector and handle it inside OPAL firmware.
	 *
	 * For newer firmware (in development/yet to be released) we will
	 * start catching/handling HMI directly in Linux.
	 */
	if (!opal_check_token(OPAL_HANDLE_HMI)) {
208
		pr_info("Old firmware detected, OPAL handles HMIs.\n");
209 210 211 212 213 214
		opal_register_exception_handler(
				OPAL_HYPERVISOR_MAINTENANCE_HANDLER,
				0, glue);
		glue += 128;
	}

215
	opal_register_exception_handler(OPAL_SOFTPATCH_HANDLER, 0, glue);
216
#endif
217

218
	return 0;
219
}
220
machine_early_initcall(powernv, opal_register_exception_handlers);
221

222 223 224 225
/*
 * Opal message notifier based on message type. Allow subscribers to get
 * notified for specific messgae type.
 */
226
int opal_message_notifier_register(enum opal_msg_type msg_type,
227 228
					struct notifier_block *nb)
{
229 230
	if (!nb || msg_type >= OPAL_MSG_TYPE_MAX) {
		pr_warning("%s: Invalid arguments, msg_type:%d\n",
231 232 233
			   __func__, msg_type);
		return -EINVAL;
	}
234

235 236 237
	return atomic_notifier_chain_register(
				&opal_msg_notifier_head[msg_type], nb);
}
238
EXPORT_SYMBOL_GPL(opal_message_notifier_register);
239

240
int opal_message_notifier_unregister(enum opal_msg_type msg_type,
241 242 243 244 245
				     struct notifier_block *nb)
{
	return atomic_notifier_chain_unregister(
			&opal_msg_notifier_head[msg_type], nb);
}
246
EXPORT_SYMBOL_GPL(opal_message_notifier_unregister);
247

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262
static void opal_message_do_notify(uint32_t msg_type, void *msg)
{
	/* notify subscribers */
	atomic_notifier_call_chain(&opal_msg_notifier_head[msg_type],
					msg_type, msg);
}

static void opal_handle_message(void)
{
	s64 ret;
	/*
	 * TODO: pre-allocate a message buffer depending on opal-msg-size
	 * value in /proc/device-tree.
	 */
	static struct opal_msg msg;
263
	u32 type;
264 265 266 267 268 269 270 271

	ret = opal_get_msg(__pa(&msg), sizeof(msg));
	/* No opal message pending. */
	if (ret == OPAL_RESOURCE)
		return;

	/* check for errors. */
	if (ret) {
M
Masanari Iida 已提交
272
		pr_warning("%s: Failed to retrieve opal message, err=%lld\n",
273 274 275 276
				__func__, ret);
		return;
	}

277 278
	type = be32_to_cpu(msg.msg_type);

279
	/* Sanity check */
280
	if (type >= OPAL_MSG_TYPE_MAX) {
281
		pr_warning("%s: Unknown message type: %u\n", __func__, type);
282 283
		return;
	}
284
	opal_message_do_notify(type, (void *)&msg);
285 286
}

287
static irqreturn_t opal_message_notify(int irq, void *data)
288
{
289 290
	opal_handle_message();
	return IRQ_HANDLED;
291 292 293 294
}

static int __init opal_message_init(void)
{
295
	int ret, i, irq;
296 297 298 299

	for (i = 0; i < OPAL_MSG_TYPE_MAX; i++)
		ATOMIC_INIT_NOTIFIER_HEAD(&opal_msg_notifier_head[i]);

300 301 302 303 304 305 306 307 308
	irq = opal_event_request(ilog2(OPAL_EVENT_MSG_PENDING));
	if (!irq) {
		pr_err("%s: Can't register OPAL event irq (%d)\n",
		       __func__, irq);
		return irq;
	}

	ret = request_irq(irq, opal_message_notify,
			IRQ_TYPE_LEVEL_HIGH, "opal-msg", NULL);
309
	if (ret) {
310
		pr_err("%s: Can't request OPAL event irq (%d)\n",
311 312 313
		       __func__, ret);
		return ret;
	}
314

315 316 317
	return 0;
}

318 319
int opal_get_chars(uint32_t vtermno, char *buf, int count)
{
320 321
	s64 rc;
	__be64 evt, len;
322 323

	if (!opal.entry)
324
		return -ENODEV;
325
	opal_poll_events(&evt);
326
	if ((be64_to_cpu(evt) & OPAL_EVENT_CONSOLE_INPUT) == 0)
327
		return 0;
328
	len = cpu_to_be64(count);
329
	rc = opal_console_read(vtermno, &len, buf);
330
	if (rc == OPAL_SUCCESS)
331
		return be64_to_cpu(len);
332 333 334 335 336 337
	return 0;
}

int opal_put_chars(uint32_t vtermno, const char *data, int total_len)
{
	int written = 0;
338
	__be64 olen;
339
	s64 len, rc;
340
	unsigned long flags;
341
	__be64 evt;
342 343

	if (!opal.entry)
344
		return -ENODEV;
345 346 347

	/* We want put_chars to be atomic to avoid mangling of hvsi
	 * packets. To do that, we first test for room and return
348 349 350 351 352
	 * -EAGAIN if there isn't enough.
	 *
	 * Unfortunately, opal_console_write_buffer_space() doesn't
	 * appear to work on opal v1, so we just assume there is
	 * enough room and be done with it
353 354
	 */
	spin_lock_irqsave(&opal_write_lock, flags);
355
	if (firmware_has_feature(FW_FEATURE_OPALv2)) {
356 357
		rc = opal_console_write_buffer_space(vtermno, &olen);
		len = be64_to_cpu(olen);
358 359 360 361 362
		if (rc || len < total_len) {
			spin_unlock_irqrestore(&opal_write_lock, flags);
			/* Closed -> drop characters */
			if (rc)
				return total_len;
363
			opal_poll_events(NULL);
364 365
			return -EAGAIN;
		}
366 367 368 369 370
	}

	/* We still try to handle partial completions, though they
	 * should no longer happen.
	 */
371
	rc = OPAL_BUSY;
372 373
	while(total_len > 0 && (rc == OPAL_BUSY ||
				rc == OPAL_BUSY_EVENT || rc == OPAL_SUCCESS)) {
374 375 376
		olen = cpu_to_be64(total_len);
		rc = opal_console_write(vtermno, &olen, data);
		len = be64_to_cpu(olen);
377 378 379 380 381 382 383

		/* Closed or other error drop */
		if (rc != OPAL_SUCCESS && rc != OPAL_BUSY &&
		    rc != OPAL_BUSY_EVENT) {
			written = total_len;
			break;
		}
384 385 386 387 388 389 390 391 392 393 394 395
		if (rc == OPAL_SUCCESS) {
			total_len -= len;
			data += len;
			written += len;
		}
		/* This is a bit nasty but we need that for the console to
		 * flush when there aren't any interrupts. We will clean
		 * things a bit later to limit that to synchronous path
		 * such as the kernel console and xmon/udbg
		 */
		do
			opal_poll_events(&evt);
396 397
		while(rc == OPAL_SUCCESS &&
			(be64_to_cpu(evt) & OPAL_EVENT_CONSOLE_OUTPUT));
398 399 400 401 402
	}
	spin_unlock_irqrestore(&opal_write_lock, flags);
	return written;
}

403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
static int opal_recover_mce(struct pt_regs *regs,
					struct machine_check_event *evt)
{
	int recovered = 0;
	uint64_t ea = get_mce_fault_addr(evt);

	if (!(regs->msr & MSR_RI)) {
		/* If MSR_RI isn't set, we cannot recover */
		recovered = 0;
	} else if (evt->disposition == MCE_DISPOSITION_RECOVERED) {
		/* Platform corrected itself */
		recovered = 1;
	} else if (ea && !is_kernel_addr(ea)) {
		/*
		 * Faulting address is not in kernel text. We should be fine.
		 * We need to find which process uses this address.
		 * For now, kill the task if we have received exception when
		 * in userspace.
		 *
		 * TODO: Queue up this address for hwpoisioning later.
		 */
		if (user_mode(regs) && !is_global_init(current)) {
			_exception(SIGBUS, regs, BUS_MCEERR_AR, regs->nip);
			recovered = 1;
		} else
			recovered = 0;
	} else if (user_mode(regs) && !is_global_init(current) &&
		evt->severity == MCE_SEV_ERROR_SYNC) {
		/*
		 * If we have received a synchronous error when in userspace
		 * kill the task.
		 */
		_exception(SIGBUS, regs, BUS_MCEERR_AR, regs->nip);
		recovered = 1;
	}
	return recovered;
}

441 442
int opal_machine_check(struct pt_regs *regs)
{
443
	struct machine_check_event evt;
444

445 446
	if (!get_mce_event(&evt, MCE_EVENT_RELEASE))
		return 0;
447 448

	/* Print things out */
449
	if (evt.version != MCE_V1) {
450 451 452 453
		pr_err("Machine Check Exception, Unknown event version %d !\n",
		       evt.version);
		return 0;
	}
454
	machine_check_print_event_info(&evt);
455

456 457 458
	if (opal_recover_mce(regs, &evt))
		return 1;
	return 0;
459 460
}

461 462 463
/* Early hmi handler called in real mode. */
int opal_hmi_exception_early(struct pt_regs *regs)
{
464 465 466 467 468 469 470 471 472 473 474 475
	s64 rc;

	/*
	 * call opal hmi handler. Pass paca address as token.
	 * The return value OPAL_SUCCESS is an indication that there is
	 * an HMI event generated waiting to pull by Linux.
	 */
	rc = opal_handle_hmi();
	if (rc == OPAL_SUCCESS) {
		local_paca->hmi_event_available = 1;
		return 1;
	}
476 477 478 479 480 481
	return 0;
}

/* HMI exception handler called in virtual mode during check_irq_replay. */
int opal_handle_hmi_exception(struct pt_regs *regs)
{
482 483 484 485 486 487 488 489 490 491 492 493 494
	s64 rc;
	__be64 evt = 0;

	/*
	 * Check if HMI event is available.
	 * if Yes, then call opal_poll_events to pull opal messages and
	 * process them.
	 */
	if (!local_paca->hmi_event_available)
		return 0;

	local_paca->hmi_event_available = 0;
	rc = opal_poll_events(&evt);
A
Alistair Popple 已提交
495
	if (rc == OPAL_SUCCESS && evt)
496
		opal_handle_events(be64_to_cpu(evt));
497 498

	return 1;
499 500
}

501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532
static uint64_t find_recovery_address(uint64_t nip)
{
	int i;

	for (i = 0; i < mc_recoverable_range_len; i++)
		if ((nip >= mc_recoverable_range[i].start_addr) &&
		    (nip < mc_recoverable_range[i].end_addr))
		    return mc_recoverable_range[i].recover_addr;
	return 0;
}

bool opal_mce_check_early_recovery(struct pt_regs *regs)
{
	uint64_t recover_addr = 0;

	if (!opal.base || !opal.size)
		goto out;

	if ((regs->nip >= opal.base) &&
			(regs->nip <= (opal.base + opal.size)))
		recover_addr = find_recovery_address(regs->nip);

	/*
	 * Setup regs->nip to rfi into fixup address.
	 */
	if (recover_addr)
		regs->nip = recover_addr;

out:
	return !!recover_addr;
}

533 534 535 536 537 538 539 540 541 542 543
static int opal_sysfs_init(void)
{
	opal_kobj = kobject_create_and_add("opal", firmware_kobj);
	if (!opal_kobj) {
		pr_warn("kobject_create_and_add opal failed\n");
		return -ENOMEM;
	}

	return 0;
}

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
static ssize_t symbol_map_read(struct file *fp, struct kobject *kobj,
			       struct bin_attribute *bin_attr,
			       char *buf, loff_t off, size_t count)
{
	return memory_read_from_buffer(buf, count, &off, bin_attr->private,
				       bin_attr->size);
}

static BIN_ATTR_RO(symbol_map, 0);

static void opal_export_symmap(void)
{
	const __be64 *syms;
	unsigned int size;
	struct device_node *fw;
	int rc;

	fw = of_find_node_by_path("/ibm,opal/firmware");
	if (!fw)
		return;
	syms = of_get_property(fw, "symbol-map", &size);
	if (!syms || size != 2 * sizeof(__be64))
		return;

	/* Setup attributes */
	bin_attr_symbol_map.private = __va(be64_to_cpu(syms[0]));
	bin_attr_symbol_map.size = be64_to_cpu(syms[1]);

	rc = sysfs_create_bin_file(opal_kobj, &bin_attr_symbol_map);
	if (rc)
		pr_warn("Error %d creating OPAL symbols file\n", rc);
}

577 578 579 580 581 582
static void __init opal_dump_region_init(void)
{
	void *addr;
	uint64_t size;
	int rc;

583 584 585
	if (!opal_check_token(OPAL_REGISTER_DUMP_REGION))
		return;

586 587
	/* Register kernel log buffer */
	addr = log_buf_addr_get();
588 589 590
	if (addr == NULL)
		return;

591
	size = log_buf_len_get();
592 593 594
	if (size == 0)
		return;

595 596 597 598 599 600 601 602 603
	rc = opal_register_dump_region(OPAL_DUMP_REGION_LOG_BUF,
				       __pa(addr), size);
	/* Don't warn if this is just an older OPAL that doesn't
	 * know about that call
	 */
	if (rc && rc != OPAL_UNSUPPORTED)
		pr_warn("DUMP: Failed to register kernel log buffer. "
			"rc = %d\n", rc);
}
604

605 606
static void opal_pdev_init(struct device_node *opal_node,
		const char *compatible)
607 608 609 610
{
	struct device_node *np;

	for_each_child_of_node(opal_node, np)
611
		if (of_device_is_compatible(np, compatible))
612 613 614
			of_platform_device_create(np, NULL, NULL);
}

615 616 617 618 619 620 621 622
static void opal_i2c_create_devs(void)
{
	struct device_node *np;

	for_each_compatible_node(np, NULL, "ibm,opal-i2c")
		of_platform_device_create(np, NULL, NULL);
}

623 624
static int kopald(void *unused)
{
625 626
	__be64 events;

627 628 629
	set_freezable();
	do {
		try_to_freeze();
630 631
		opal_poll_events(&events);
		opal_handle_events(be64_to_cpu(events));
632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648
		msleep_interruptible(opal_heartbeat);
	} while (!kthread_should_stop());

	return 0;
}

static void opal_init_heartbeat(void)
{
	/* Old firwmware, we assume the HVC heartbeat is sufficient */
	if (of_property_read_u32(opal_node, "ibm,heartbeat-ms",
				 &opal_heartbeat) != 0)
		opal_heartbeat = 0;

	if (opal_heartbeat)
		kthread_run(kopald, NULL, "kopald");
}

649 650 651
static int __init opal_init(void)
{
	struct device_node *np, *consoles;
652
	int rc;
653 654 655

	opal_node = of_find_node_by_path("/ibm,opal");
	if (!opal_node) {
656
		pr_warn("Device node not found\n");
657 658
		return -ENODEV;
	}
659

660 661 662
	/* Initialise OPAL events */
	opal_event_init();

663
	/* Register OPAL consoles if any ports */
664 665 666 667
	if (firmware_has_feature(FW_FEATURE_OPALv2))
		consoles = of_find_node_by_path("/ibm,opal/consoles");
	else
		consoles = of_node_get(opal_node);
668 669 670 671 672 673 674
	if (consoles) {
		for_each_child_of_node(consoles, np) {
			if (strcmp(np->name, "serial"))
				continue;
			of_platform_device_create(np, NULL, NULL);
		}
		of_node_put(consoles);
675
	}
676

677 678 679 680 681 682 683 684 685 686 687 688
	/* Initialise OPAL messaging system */
	opal_message_init();

	/* Initialise OPAL asynchronous completion interface */
	opal_async_comp_init();

	/* Initialise OPAL sensor interface */
	opal_sensor_init();

	/* Initialise OPAL hypervisor maintainence interrupt handling */
	opal_hmi_handler_init();

689 690 691
	/* Create i2c platform devices */
	opal_i2c_create_devs();

692 693 694
	/* Setup a heatbeat thread if requested by OPAL */
	opal_init_heartbeat();

695 696
	/* Create "opal" kobject under /sys/firmware */
	rc = opal_sysfs_init();
697
	if (rc == 0) {
698 699
		/* Export symbol map to userspace */
		opal_export_symmap();
700 701
		/* Setup dump region interface */
		opal_dump_region_init();
702 703
		/* Setup error log interface */
		rc = opal_elog_init();
704
		/* Setup code update interface */
705
		opal_flash_update_init();
706 707
		/* Setup platform dump extract interface */
		opal_platform_dump_init();
708 709
		/* Setup system parameters interface */
		opal_sys_param_init();
710 711
		/* Setup message log interface. */
		opal_msglog_init();
712
	}
713

714 715 716
	/* Initialize platform devices: IPMI backend & flash interface */
	opal_pdev_init(opal_node, "ibm,opal-ipmi");
	opal_pdev_init(opal_node, "ibm,opal-flash");
717

718 719
	return 0;
}
720
machine_subsys_initcall(powernv, opal_init);
721 722 723

void opal_shutdown(void)
{
724
	long rc = OPAL_BUSY;
725

726
	opal_event_shutdown();
727 728 729 730 731 732 733 734 735 736 737 738 739

	/*
	 * Then sync with OPAL which ensure anything that can
	 * potentially write to our memory has completed such
	 * as an ongoing dump retrieval
	 */
	while (rc == OPAL_BUSY || rc == OPAL_BUSY_EVENT) {
		rc = opal_sync_host_reboot();
		if (rc == OPAL_BUSY)
			opal_poll_events(NULL);
		else
			mdelay(10);
	}
740 741

	/* Unregister memory dump region */
742 743
	if (opal_check_token(OPAL_UNREGISTER_DUMP_REGION))
		opal_unregister_dump_region(OPAL_DUMP_REGION_LOG_BUF);
744
}
745 746 747

/* Export this so that test modules can use it */
EXPORT_SYMBOL_GPL(opal_invalid_call);
748 749
EXPORT_SYMBOL_GPL(opal_xscom_read);
EXPORT_SYMBOL_GPL(opal_xscom_write);
750 751
EXPORT_SYMBOL_GPL(opal_ipmi_send);
EXPORT_SYMBOL_GPL(opal_ipmi_recv);
752 753 754
EXPORT_SYMBOL_GPL(opal_flash_read);
EXPORT_SYMBOL_GPL(opal_flash_write);
EXPORT_SYMBOL_GPL(opal_flash_erase);
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

/* Convert a region of vmalloc memory to an opal sg list */
struct opal_sg_list *opal_vmalloc_to_sg_list(void *vmalloc_addr,
					     unsigned long vmalloc_size)
{
	struct opal_sg_list *sg, *first = NULL;
	unsigned long i = 0;

	sg = kzalloc(PAGE_SIZE, GFP_KERNEL);
	if (!sg)
		goto nomem;

	first = sg;

	while (vmalloc_size > 0) {
		uint64_t data = vmalloc_to_pfn(vmalloc_addr) << PAGE_SHIFT;
		uint64_t length = min(vmalloc_size, PAGE_SIZE);

		sg->entry[i].data = cpu_to_be64(data);
		sg->entry[i].length = cpu_to_be64(length);
		i++;

		if (i >= SG_ENTRIES_PER_NODE) {
			struct opal_sg_list *next;

			next = kzalloc(PAGE_SIZE, GFP_KERNEL);
			if (!next)
				goto nomem;

			sg->length = cpu_to_be64(
					i * sizeof(struct opal_sg_entry) + 16);
			i = 0;
			sg->next = cpu_to_be64(__pa(next));
			sg = next;
		}

		vmalloc_addr += length;
		vmalloc_size -= length;
	}

	sg->length = cpu_to_be64(i * sizeof(struct opal_sg_entry) + 16);

	return first;

nomem:
	pr_err("%s : Failed to allocate memory\n", __func__);
	opal_free_sg_list(first);
	return NULL;
}

void opal_free_sg_list(struct opal_sg_list *sg)
{
	while (sg) {
		uint64_t next = be64_to_cpu(sg->next);

		kfree(sg);

		if (next)
			sg = __va(next);
		else
			sg = NULL;
	}
}
818

819 820 821 822 823 824 825 826 827
int opal_error_code(int rc)
{
	switch (rc) {
	case OPAL_SUCCESS:		return 0;

	case OPAL_PARAMETER:		return -EINVAL;
	case OPAL_ASYNC_COMPLETION:	return -EINPROGRESS;
	case OPAL_BUSY_EVENT:		return -EBUSY;
	case OPAL_NO_MEM:		return -ENOMEM;
828
	case OPAL_PERMISSION:		return -EPERM;
829 830 831 832 833 834 835 836 837 838

	case OPAL_UNSUPPORTED:		return -EIO;
	case OPAL_HARDWARE:		return -EIO;
	case OPAL_INTERNAL_ERROR:	return -EIO;
	default:
		pr_err("%s: unexpected OPAL error %d\n", __func__, rc);
		return -EIO;
	}
}

839 840 841 842 843
EXPORT_SYMBOL_GPL(opal_poll_events);
EXPORT_SYMBOL_GPL(opal_rtc_read);
EXPORT_SYMBOL_GPL(opal_rtc_write);
EXPORT_SYMBOL_GPL(opal_tpo_read);
EXPORT_SYMBOL_GPL(opal_tpo_write);
844
EXPORT_SYMBOL_GPL(opal_i2c_request);