xref: /openbmc/linux/drivers/iommu/intel/dmar.c (revision c6acb1e7)
1 // SPDX-License-Identifier: GPL-2.0-only
2 /*
3  * Copyright (c) 2006, Intel Corporation.
4  *
5  * Copyright (C) 2006-2008 Intel Corporation
6  * Author: Ashok Raj <ashok.raj@intel.com>
7  * Author: Shaohua Li <shaohua.li@intel.com>
8  * Author: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
9  *
10  * This file implements early detection/parsing of Remapping Devices
11  * reported to OS through BIOS via DMA remapping reporting (DMAR) ACPI
12  * tables.
13  *
14  * These routines are used by both DMA-remapping and Interrupt-remapping
15  */
16 
17 #define pr_fmt(fmt)     "DMAR: " fmt
18 
19 #include <linux/pci.h>
20 #include <linux/dmar.h>
21 #include <linux/iova.h>
22 #include <linux/intel-iommu.h>
23 #include <linux/timer.h>
24 #include <linux/irq.h>
25 #include <linux/interrupt.h>
26 #include <linux/tboot.h>
27 #include <linux/dmi.h>
28 #include <linux/slab.h>
29 #include <linux/iommu.h>
30 #include <linux/numa.h>
31 #include <linux/limits.h>
32 #include <asm/irq_remapping.h>
33 #include <asm/iommu_table.h>
34 #include <trace/events/intel_iommu.h>
35 
36 #include "../irq_remapping.h"
37 #include "perf.h"
38 
39 typedef int (*dmar_res_handler_t)(struct acpi_dmar_header *, void *);
40 struct dmar_res_callback {
41 	dmar_res_handler_t	cb[ACPI_DMAR_TYPE_RESERVED];
42 	void			*arg[ACPI_DMAR_TYPE_RESERVED];
43 	bool			ignore_unhandled;
44 	bool			print_entry;
45 };
46 
47 /*
48  * Assumptions:
49  * 1) The hotplug framework guarentees that DMAR unit will be hot-added
50  *    before IO devices managed by that unit.
51  * 2) The hotplug framework guarantees that DMAR unit will be hot-removed
52  *    after IO devices managed by that unit.
53  * 3) Hotplug events are rare.
54  *
55  * Locking rules for DMA and interrupt remapping related global data structures:
56  * 1) Use dmar_global_lock in process context
57  * 2) Use RCU in interrupt context
58  */
59 DECLARE_RWSEM(dmar_global_lock);
60 LIST_HEAD(dmar_drhd_units);
61 
62 struct acpi_table_header * __initdata dmar_tbl;
63 static int dmar_dev_scope_status = 1;
64 static unsigned long dmar_seq_ids[BITS_TO_LONGS(DMAR_UNITS_SUPPORTED)];
65 
66 static int alloc_iommu(struct dmar_drhd_unit *drhd);
67 static void free_iommu(struct intel_iommu *iommu);
68 
69 extern const struct iommu_ops intel_iommu_ops;
70 
71 static void dmar_register_drhd_unit(struct dmar_drhd_unit *drhd)
72 {
73 	/*
74 	 * add INCLUDE_ALL at the tail, so scan the list will find it at
75 	 * the very end.
76 	 */
77 	if (drhd->include_all)
78 		list_add_tail_rcu(&drhd->list, &dmar_drhd_units);
79 	else
80 		list_add_rcu(&drhd->list, &dmar_drhd_units);
81 }
82 
83 void *dmar_alloc_dev_scope(void *start, void *end, int *cnt)
84 {
85 	struct acpi_dmar_device_scope *scope;
86 
87 	*cnt = 0;
88 	while (start < end) {
89 		scope = start;
90 		if (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_NAMESPACE ||
91 		    scope->entry_type == ACPI_DMAR_SCOPE_TYPE_ENDPOINT ||
92 		    scope->entry_type == ACPI_DMAR_SCOPE_TYPE_BRIDGE)
93 			(*cnt)++;
94 		else if (scope->entry_type != ACPI_DMAR_SCOPE_TYPE_IOAPIC &&
95 			scope->entry_type != ACPI_DMAR_SCOPE_TYPE_HPET) {
96 			pr_warn("Unsupported device scope\n");
97 		}
98 		start += scope->length;
99 	}
100 	if (*cnt == 0)
101 		return NULL;
102 
103 	return kcalloc(*cnt, sizeof(struct dmar_dev_scope), GFP_KERNEL);
104 }
105 
106 void dmar_free_dev_scope(struct dmar_dev_scope **devices, int *cnt)
107 {
108 	int i;
109 	struct device *tmp_dev;
110 
111 	if (*devices && *cnt) {
112 		for_each_active_dev_scope(*devices, *cnt, i, tmp_dev)
113 			put_device(tmp_dev);
114 		kfree(*devices);
115 	}
116 
117 	*devices = NULL;
118 	*cnt = 0;
119 }
120 
121 /* Optimize out kzalloc()/kfree() for normal cases */
122 static char dmar_pci_notify_info_buf[64];
123 
124 static struct dmar_pci_notify_info *
125 dmar_alloc_pci_notify_info(struct pci_dev *dev, unsigned long event)
126 {
127 	int level = 0;
128 	size_t size;
129 	struct pci_dev *tmp;
130 	struct dmar_pci_notify_info *info;
131 
132 	BUG_ON(dev->is_virtfn);
133 
134 	/*
135 	 * Ignore devices that have a domain number higher than what can
136 	 * be looked up in DMAR, e.g. VMD subdevices with domain 0x10000
137 	 */
138 	if (pci_domain_nr(dev->bus) > U16_MAX)
139 		return NULL;
140 
141 	/* Only generate path[] for device addition event */
142 	if (event == BUS_NOTIFY_ADD_DEVICE)
143 		for (tmp = dev; tmp; tmp = tmp->bus->self)
144 			level++;
145 
146 	size = struct_size(info, path, level);
147 	if (size <= sizeof(dmar_pci_notify_info_buf)) {
148 		info = (struct dmar_pci_notify_info *)dmar_pci_notify_info_buf;
149 	} else {
150 		info = kzalloc(size, GFP_KERNEL);
151 		if (!info) {
152 			if (dmar_dev_scope_status == 0)
153 				dmar_dev_scope_status = -ENOMEM;
154 			return NULL;
155 		}
156 	}
157 
158 	info->event = event;
159 	info->dev = dev;
160 	info->seg = pci_domain_nr(dev->bus);
161 	info->level = level;
162 	if (event == BUS_NOTIFY_ADD_DEVICE) {
163 		for (tmp = dev; tmp; tmp = tmp->bus->self) {
164 			level--;
165 			info->path[level].bus = tmp->bus->number;
166 			info->path[level].device = PCI_SLOT(tmp->devfn);
167 			info->path[level].function = PCI_FUNC(tmp->devfn);
168 			if (pci_is_root_bus(tmp->bus))
169 				info->bus = tmp->bus->number;
170 		}
171 	}
172 
173 	return info;
174 }
175 
176 static inline void dmar_free_pci_notify_info(struct dmar_pci_notify_info *info)
177 {
178 	if ((void *)info != dmar_pci_notify_info_buf)
179 		kfree(info);
180 }
181 
182 static bool dmar_match_pci_path(struct dmar_pci_notify_info *info, int bus,
183 				struct acpi_dmar_pci_path *path, int count)
184 {
185 	int i;
186 
187 	if (info->bus != bus)
188 		goto fallback;
189 	if (info->level != count)
190 		goto fallback;
191 
192 	for (i = 0; i < count; i++) {
193 		if (path[i].device != info->path[i].device ||
194 		    path[i].function != info->path[i].function)
195 			goto fallback;
196 	}
197 
198 	return true;
199 
200 fallback:
201 
202 	if (count != 1)
203 		return false;
204 
205 	i = info->level - 1;
206 	if (bus              == info->path[i].bus &&
207 	    path[0].device   == info->path[i].device &&
208 	    path[0].function == info->path[i].function) {
209 		pr_info(FW_BUG "RMRR entry for device %02x:%02x.%x is broken - applying workaround\n",
210 			bus, path[0].device, path[0].function);
211 		return true;
212 	}
213 
214 	return false;
215 }
216 
217 /* Return: > 0 if match found, 0 if no match found, < 0 if error happens */
218 int dmar_insert_dev_scope(struct dmar_pci_notify_info *info,
219 			  void *start, void*end, u16 segment,
220 			  struct dmar_dev_scope *devices,
221 			  int devices_cnt)
222 {
223 	int i, level;
224 	struct device *tmp, *dev = &info->dev->dev;
225 	struct acpi_dmar_device_scope *scope;
226 	struct acpi_dmar_pci_path *path;
227 
228 	if (segment != info->seg)
229 		return 0;
230 
231 	for (; start < end; start += scope->length) {
232 		scope = start;
233 		if (scope->entry_type != ACPI_DMAR_SCOPE_TYPE_ENDPOINT &&
234 		    scope->entry_type != ACPI_DMAR_SCOPE_TYPE_BRIDGE)
235 			continue;
236 
237 		path = (struct acpi_dmar_pci_path *)(scope + 1);
238 		level = (scope->length - sizeof(*scope)) / sizeof(*path);
239 		if (!dmar_match_pci_path(info, scope->bus, path, level))
240 			continue;
241 
242 		/*
243 		 * We expect devices with endpoint scope to have normal PCI
244 		 * headers, and devices with bridge scope to have bridge PCI
245 		 * headers.  However PCI NTB devices may be listed in the
246 		 * DMAR table with bridge scope, even though they have a
247 		 * normal PCI header.  NTB devices are identified by class
248 		 * "BRIDGE_OTHER" (0680h) - we don't declare a socpe mismatch
249 		 * for this special case.
250 		 */
251 		if ((scope->entry_type == ACPI_DMAR_SCOPE_TYPE_ENDPOINT &&
252 		     info->dev->hdr_type != PCI_HEADER_TYPE_NORMAL) ||
253 		    (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_BRIDGE &&
254 		     (info->dev->hdr_type == PCI_HEADER_TYPE_NORMAL &&
255 		      info->dev->class >> 16 != PCI_BASE_CLASS_BRIDGE))) {
256 			pr_warn("Device scope type does not match for %s\n",
257 				pci_name(info->dev));
258 			return -EINVAL;
259 		}
260 
261 		for_each_dev_scope(devices, devices_cnt, i, tmp)
262 			if (tmp == NULL) {
263 				devices[i].bus = info->dev->bus->number;
264 				devices[i].devfn = info->dev->devfn;
265 				rcu_assign_pointer(devices[i].dev,
266 						   get_device(dev));
267 				return 1;
268 			}
269 		BUG_ON(i >= devices_cnt);
270 	}
271 
272 	return 0;
273 }
274 
275 int dmar_remove_dev_scope(struct dmar_pci_notify_info *info, u16 segment,
276 			  struct dmar_dev_scope *devices, int count)
277 {
278 	int index;
279 	struct device *tmp;
280 
281 	if (info->seg != segment)
282 		return 0;
283 
284 	for_each_active_dev_scope(devices, count, index, tmp)
285 		if (tmp == &info->dev->dev) {
286 			RCU_INIT_POINTER(devices[index].dev, NULL);
287 			synchronize_rcu();
288 			put_device(tmp);
289 			return 1;
290 		}
291 
292 	return 0;
293 }
294 
295 static int dmar_pci_bus_add_dev(struct dmar_pci_notify_info *info)
296 {
297 	int ret = 0;
298 	struct dmar_drhd_unit *dmaru;
299 	struct acpi_dmar_hardware_unit *drhd;
300 
301 	for_each_drhd_unit(dmaru) {
302 		if (dmaru->include_all)
303 			continue;
304 
305 		drhd = container_of(dmaru->hdr,
306 				    struct acpi_dmar_hardware_unit, header);
307 		ret = dmar_insert_dev_scope(info, (void *)(drhd + 1),
308 				((void *)drhd) + drhd->header.length,
309 				dmaru->segment,
310 				dmaru->devices, dmaru->devices_cnt);
311 		if (ret)
312 			break;
313 	}
314 	if (ret >= 0)
315 		ret = dmar_iommu_notify_scope_dev(info);
316 	if (ret < 0 && dmar_dev_scope_status == 0)
317 		dmar_dev_scope_status = ret;
318 
319 	if (ret >= 0)
320 		intel_irq_remap_add_device(info);
321 
322 	return ret;
323 }
324 
325 static void  dmar_pci_bus_del_dev(struct dmar_pci_notify_info *info)
326 {
327 	struct dmar_drhd_unit *dmaru;
328 
329 	for_each_drhd_unit(dmaru)
330 		if (dmar_remove_dev_scope(info, dmaru->segment,
331 			dmaru->devices, dmaru->devices_cnt))
332 			break;
333 	dmar_iommu_notify_scope_dev(info);
334 }
335 
336 static inline void vf_inherit_msi_domain(struct pci_dev *pdev)
337 {
338 	struct pci_dev *physfn = pci_physfn(pdev);
339 
340 	dev_set_msi_domain(&pdev->dev, dev_get_msi_domain(&physfn->dev));
341 }
342 
343 static int dmar_pci_bus_notifier(struct notifier_block *nb,
344 				 unsigned long action, void *data)
345 {
346 	struct pci_dev *pdev = to_pci_dev(data);
347 	struct dmar_pci_notify_info *info;
348 
349 	/* Only care about add/remove events for physical functions.
350 	 * For VFs we actually do the lookup based on the corresponding
351 	 * PF in device_to_iommu() anyway. */
352 	if (pdev->is_virtfn) {
353 		/*
354 		 * Ensure that the VF device inherits the irq domain of the
355 		 * PF device. Ideally the device would inherit the domain
356 		 * from the bus, but DMAR can have multiple units per bus
357 		 * which makes this impossible. The VF 'bus' could inherit
358 		 * from the PF device, but that's yet another x86'sism to
359 		 * inflict on everybody else.
360 		 */
361 		if (action == BUS_NOTIFY_ADD_DEVICE)
362 			vf_inherit_msi_domain(pdev);
363 		return NOTIFY_DONE;
364 	}
365 
366 	if (action != BUS_NOTIFY_ADD_DEVICE &&
367 	    action != BUS_NOTIFY_REMOVED_DEVICE)
368 		return NOTIFY_DONE;
369 
370 	info = dmar_alloc_pci_notify_info(pdev, action);
371 	if (!info)
372 		return NOTIFY_DONE;
373 
374 	down_write(&dmar_global_lock);
375 	if (action == BUS_NOTIFY_ADD_DEVICE)
376 		dmar_pci_bus_add_dev(info);
377 	else if (action == BUS_NOTIFY_REMOVED_DEVICE)
378 		dmar_pci_bus_del_dev(info);
379 	up_write(&dmar_global_lock);
380 
381 	dmar_free_pci_notify_info(info);
382 
383 	return NOTIFY_OK;
384 }
385 
386 static struct notifier_block dmar_pci_bus_nb = {
387 	.notifier_call = dmar_pci_bus_notifier,
388 	.priority = INT_MIN,
389 };
390 
391 static struct dmar_drhd_unit *
392 dmar_find_dmaru(struct acpi_dmar_hardware_unit *drhd)
393 {
394 	struct dmar_drhd_unit *dmaru;
395 
396 	list_for_each_entry_rcu(dmaru, &dmar_drhd_units, list,
397 				dmar_rcu_check())
398 		if (dmaru->segment == drhd->segment &&
399 		    dmaru->reg_base_addr == drhd->address)
400 			return dmaru;
401 
402 	return NULL;
403 }
404 
405 /*
406  * dmar_parse_one_drhd - parses exactly one DMA remapping hardware definition
407  * structure which uniquely represent one DMA remapping hardware unit
408  * present in the platform
409  */
410 static int dmar_parse_one_drhd(struct acpi_dmar_header *header, void *arg)
411 {
412 	struct acpi_dmar_hardware_unit *drhd;
413 	struct dmar_drhd_unit *dmaru;
414 	int ret;
415 
416 	drhd = (struct acpi_dmar_hardware_unit *)header;
417 	dmaru = dmar_find_dmaru(drhd);
418 	if (dmaru)
419 		goto out;
420 
421 	dmaru = kzalloc(sizeof(*dmaru) + header->length, GFP_KERNEL);
422 	if (!dmaru)
423 		return -ENOMEM;
424 
425 	/*
426 	 * If header is allocated from slab by ACPI _DSM method, we need to
427 	 * copy the content because the memory buffer will be freed on return.
428 	 */
429 	dmaru->hdr = (void *)(dmaru + 1);
430 	memcpy(dmaru->hdr, header, header->length);
431 	dmaru->reg_base_addr = drhd->address;
432 	dmaru->segment = drhd->segment;
433 	dmaru->include_all = drhd->flags & 0x1; /* BIT0: INCLUDE_ALL */
434 	dmaru->devices = dmar_alloc_dev_scope((void *)(drhd + 1),
435 					      ((void *)drhd) + drhd->header.length,
436 					      &dmaru->devices_cnt);
437 	if (dmaru->devices_cnt && dmaru->devices == NULL) {
438 		kfree(dmaru);
439 		return -ENOMEM;
440 	}
441 
442 	ret = alloc_iommu(dmaru);
443 	if (ret) {
444 		dmar_free_dev_scope(&dmaru->devices,
445 				    &dmaru->devices_cnt);
446 		kfree(dmaru);
447 		return ret;
448 	}
449 	dmar_register_drhd_unit(dmaru);
450 
451 out:
452 	if (arg)
453 		(*(int *)arg)++;
454 
455 	return 0;
456 }
457 
458 static void dmar_free_drhd(struct dmar_drhd_unit *dmaru)
459 {
460 	if (dmaru->devices && dmaru->devices_cnt)
461 		dmar_free_dev_scope(&dmaru->devices, &dmaru->devices_cnt);
462 	if (dmaru->iommu)
463 		free_iommu(dmaru->iommu);
464 	kfree(dmaru);
465 }
466 
467 static int __init dmar_parse_one_andd(struct acpi_dmar_header *header,
468 				      void *arg)
469 {
470 	struct acpi_dmar_andd *andd = (void *)header;
471 
472 	/* Check for NUL termination within the designated length */
473 	if (strnlen(andd->device_name, header->length - 8) == header->length - 8) {
474 		pr_warn(FW_BUG
475 			   "Your BIOS is broken; ANDD object name is not NUL-terminated\n"
476 			   "BIOS vendor: %s; Ver: %s; Product Version: %s\n",
477 			   dmi_get_system_info(DMI_BIOS_VENDOR),
478 			   dmi_get_system_info(DMI_BIOS_VERSION),
479 			   dmi_get_system_info(DMI_PRODUCT_VERSION));
480 		add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
481 		return -EINVAL;
482 	}
483 	pr_info("ANDD device: %x name: %s\n", andd->device_number,
484 		andd->device_name);
485 
486 	return 0;
487 }
488 
489 #ifdef CONFIG_ACPI_NUMA
490 static int dmar_parse_one_rhsa(struct acpi_dmar_header *header, void *arg)
491 {
492 	struct acpi_dmar_rhsa *rhsa;
493 	struct dmar_drhd_unit *drhd;
494 
495 	rhsa = (struct acpi_dmar_rhsa *)header;
496 	for_each_drhd_unit(drhd) {
497 		if (drhd->reg_base_addr == rhsa->base_address) {
498 			int node = pxm_to_node(rhsa->proximity_domain);
499 
500 			if (!node_online(node))
501 				node = NUMA_NO_NODE;
502 			drhd->iommu->node = node;
503 			return 0;
504 		}
505 	}
506 	pr_warn(FW_BUG
507 		"Your BIOS is broken; RHSA refers to non-existent DMAR unit at %llx\n"
508 		"BIOS vendor: %s; Ver: %s; Product Version: %s\n",
509 		rhsa->base_address,
510 		dmi_get_system_info(DMI_BIOS_VENDOR),
511 		dmi_get_system_info(DMI_BIOS_VERSION),
512 		dmi_get_system_info(DMI_PRODUCT_VERSION));
513 	add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
514 
515 	return 0;
516 }
517 #else
518 #define	dmar_parse_one_rhsa		dmar_res_noop
519 #endif
520 
521 static void
522 dmar_table_print_dmar_entry(struct acpi_dmar_header *header)
523 {
524 	struct acpi_dmar_hardware_unit *drhd;
525 	struct acpi_dmar_reserved_memory *rmrr;
526 	struct acpi_dmar_atsr *atsr;
527 	struct acpi_dmar_rhsa *rhsa;
528 	struct acpi_dmar_satc *satc;
529 
530 	switch (header->type) {
531 	case ACPI_DMAR_TYPE_HARDWARE_UNIT:
532 		drhd = container_of(header, struct acpi_dmar_hardware_unit,
533 				    header);
534 		pr_info("DRHD base: %#016Lx flags: %#x\n",
535 			(unsigned long long)drhd->address, drhd->flags);
536 		break;
537 	case ACPI_DMAR_TYPE_RESERVED_MEMORY:
538 		rmrr = container_of(header, struct acpi_dmar_reserved_memory,
539 				    header);
540 		pr_info("RMRR base: %#016Lx end: %#016Lx\n",
541 			(unsigned long long)rmrr->base_address,
542 			(unsigned long long)rmrr->end_address);
543 		break;
544 	case ACPI_DMAR_TYPE_ROOT_ATS:
545 		atsr = container_of(header, struct acpi_dmar_atsr, header);
546 		pr_info("ATSR flags: %#x\n", atsr->flags);
547 		break;
548 	case ACPI_DMAR_TYPE_HARDWARE_AFFINITY:
549 		rhsa = container_of(header, struct acpi_dmar_rhsa, header);
550 		pr_info("RHSA base: %#016Lx proximity domain: %#x\n",
551 		       (unsigned long long)rhsa->base_address,
552 		       rhsa->proximity_domain);
553 		break;
554 	case ACPI_DMAR_TYPE_NAMESPACE:
555 		/* We don't print this here because we need to sanity-check
556 		   it first. So print it in dmar_parse_one_andd() instead. */
557 		break;
558 	case ACPI_DMAR_TYPE_SATC:
559 		satc = container_of(header, struct acpi_dmar_satc, header);
560 		pr_info("SATC flags: 0x%x\n", satc->flags);
561 		break;
562 	}
563 }
564 
565 /**
566  * dmar_table_detect - checks to see if the platform supports DMAR devices
567  */
568 static int __init dmar_table_detect(void)
569 {
570 	acpi_status status = AE_OK;
571 
572 	/* if we could find DMAR table, then there are DMAR devices */
573 	status = acpi_get_table(ACPI_SIG_DMAR, 0, &dmar_tbl);
574 
575 	if (ACPI_SUCCESS(status) && !dmar_tbl) {
576 		pr_warn("Unable to map DMAR\n");
577 		status = AE_NOT_FOUND;
578 	}
579 
580 	return ACPI_SUCCESS(status) ? 0 : -ENOENT;
581 }
582 
583 static int dmar_walk_remapping_entries(struct acpi_dmar_header *start,
584 				       size_t len, struct dmar_res_callback *cb)
585 {
586 	struct acpi_dmar_header *iter, *next;
587 	struct acpi_dmar_header *end = ((void *)start) + len;
588 
589 	for (iter = start; iter < end; iter = next) {
590 		next = (void *)iter + iter->length;
591 		if (iter->length == 0) {
592 			/* Avoid looping forever on bad ACPI tables */
593 			pr_debug(FW_BUG "Invalid 0-length structure\n");
594 			break;
595 		} else if (next > end) {
596 			/* Avoid passing table end */
597 			pr_warn(FW_BUG "Record passes table end\n");
598 			return -EINVAL;
599 		}
600 
601 		if (cb->print_entry)
602 			dmar_table_print_dmar_entry(iter);
603 
604 		if (iter->type >= ACPI_DMAR_TYPE_RESERVED) {
605 			/* continue for forward compatibility */
606 			pr_debug("Unknown DMAR structure type %d\n",
607 				 iter->type);
608 		} else if (cb->cb[iter->type]) {
609 			int ret;
610 
611 			ret = cb->cb[iter->type](iter, cb->arg[iter->type]);
612 			if (ret)
613 				return ret;
614 		} else if (!cb->ignore_unhandled) {
615 			pr_warn("No handler for DMAR structure type %d\n",
616 				iter->type);
617 			return -EINVAL;
618 		}
619 	}
620 
621 	return 0;
622 }
623 
624 static inline int dmar_walk_dmar_table(struct acpi_table_dmar *dmar,
625 				       struct dmar_res_callback *cb)
626 {
627 	return dmar_walk_remapping_entries((void *)(dmar + 1),
628 			dmar->header.length - sizeof(*dmar), cb);
629 }
630 
631 /**
632  * parse_dmar_table - parses the DMA reporting table
633  */
634 static int __init
635 parse_dmar_table(void)
636 {
637 	struct acpi_table_dmar *dmar;
638 	int drhd_count = 0;
639 	int ret;
640 	struct dmar_res_callback cb = {
641 		.print_entry = true,
642 		.ignore_unhandled = true,
643 		.arg[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &drhd_count,
644 		.cb[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &dmar_parse_one_drhd,
645 		.cb[ACPI_DMAR_TYPE_RESERVED_MEMORY] = &dmar_parse_one_rmrr,
646 		.cb[ACPI_DMAR_TYPE_ROOT_ATS] = &dmar_parse_one_atsr,
647 		.cb[ACPI_DMAR_TYPE_HARDWARE_AFFINITY] = &dmar_parse_one_rhsa,
648 		.cb[ACPI_DMAR_TYPE_NAMESPACE] = &dmar_parse_one_andd,
649 		.cb[ACPI_DMAR_TYPE_SATC] = &dmar_parse_one_satc,
650 	};
651 
652 	/*
653 	 * Do it again, earlier dmar_tbl mapping could be mapped with
654 	 * fixed map.
655 	 */
656 	dmar_table_detect();
657 
658 	/*
659 	 * ACPI tables may not be DMA protected by tboot, so use DMAR copy
660 	 * SINIT saved in SinitMleData in TXT heap (which is DMA protected)
661 	 */
662 	dmar_tbl = tboot_get_dmar_table(dmar_tbl);
663 
664 	dmar = (struct acpi_table_dmar *)dmar_tbl;
665 	if (!dmar)
666 		return -ENODEV;
667 
668 	if (dmar->width < PAGE_SHIFT - 1) {
669 		pr_warn("Invalid DMAR haw\n");
670 		return -EINVAL;
671 	}
672 
673 	pr_info("Host address width %d\n", dmar->width + 1);
674 	ret = dmar_walk_dmar_table(dmar, &cb);
675 	if (ret == 0 && drhd_count == 0)
676 		pr_warn(FW_BUG "No DRHD structure found in DMAR table\n");
677 
678 	return ret;
679 }
680 
681 static int dmar_pci_device_match(struct dmar_dev_scope devices[],
682 				 int cnt, struct pci_dev *dev)
683 {
684 	int index;
685 	struct device *tmp;
686 
687 	while (dev) {
688 		for_each_active_dev_scope(devices, cnt, index, tmp)
689 			if (dev_is_pci(tmp) && dev == to_pci_dev(tmp))
690 				return 1;
691 
692 		/* Check our parent */
693 		dev = dev->bus->self;
694 	}
695 
696 	return 0;
697 }
698 
699 struct dmar_drhd_unit *
700 dmar_find_matched_drhd_unit(struct pci_dev *dev)
701 {
702 	struct dmar_drhd_unit *dmaru;
703 	struct acpi_dmar_hardware_unit *drhd;
704 
705 	dev = pci_physfn(dev);
706 
707 	rcu_read_lock();
708 	for_each_drhd_unit(dmaru) {
709 		drhd = container_of(dmaru->hdr,
710 				    struct acpi_dmar_hardware_unit,
711 				    header);
712 
713 		if (dmaru->include_all &&
714 		    drhd->segment == pci_domain_nr(dev->bus))
715 			goto out;
716 
717 		if (dmar_pci_device_match(dmaru->devices,
718 					  dmaru->devices_cnt, dev))
719 			goto out;
720 	}
721 	dmaru = NULL;
722 out:
723 	rcu_read_unlock();
724 
725 	return dmaru;
726 }
727 
728 static void __init dmar_acpi_insert_dev_scope(u8 device_number,
729 					      struct acpi_device *adev)
730 {
731 	struct dmar_drhd_unit *dmaru;
732 	struct acpi_dmar_hardware_unit *drhd;
733 	struct acpi_dmar_device_scope *scope;
734 	struct device *tmp;
735 	int i;
736 	struct acpi_dmar_pci_path *path;
737 
738 	for_each_drhd_unit(dmaru) {
739 		drhd = container_of(dmaru->hdr,
740 				    struct acpi_dmar_hardware_unit,
741 				    header);
742 
743 		for (scope = (void *)(drhd + 1);
744 		     (unsigned long)scope < ((unsigned long)drhd) + drhd->header.length;
745 		     scope = ((void *)scope) + scope->length) {
746 			if (scope->entry_type != ACPI_DMAR_SCOPE_TYPE_NAMESPACE)
747 				continue;
748 			if (scope->enumeration_id != device_number)
749 				continue;
750 
751 			path = (void *)(scope + 1);
752 			pr_info("ACPI device \"%s\" under DMAR at %llx as %02x:%02x.%d\n",
753 				dev_name(&adev->dev), dmaru->reg_base_addr,
754 				scope->bus, path->device, path->function);
755 			for_each_dev_scope(dmaru->devices, dmaru->devices_cnt, i, tmp)
756 				if (tmp == NULL) {
757 					dmaru->devices[i].bus = scope->bus;
758 					dmaru->devices[i].devfn = PCI_DEVFN(path->device,
759 									    path->function);
760 					rcu_assign_pointer(dmaru->devices[i].dev,
761 							   get_device(&adev->dev));
762 					return;
763 				}
764 			BUG_ON(i >= dmaru->devices_cnt);
765 		}
766 	}
767 	pr_warn("No IOMMU scope found for ANDD enumeration ID %d (%s)\n",
768 		device_number, dev_name(&adev->dev));
769 }
770 
771 static int __init dmar_acpi_dev_scope_init(void)
772 {
773 	struct acpi_dmar_andd *andd;
774 
775 	if (dmar_tbl == NULL)
776 		return -ENODEV;
777 
778 	for (andd = (void *)dmar_tbl + sizeof(struct acpi_table_dmar);
779 	     ((unsigned long)andd) < ((unsigned long)dmar_tbl) + dmar_tbl->length;
780 	     andd = ((void *)andd) + andd->header.length) {
781 		if (andd->header.type == ACPI_DMAR_TYPE_NAMESPACE) {
782 			acpi_handle h;
783 			struct acpi_device *adev;
784 
785 			if (!ACPI_SUCCESS(acpi_get_handle(ACPI_ROOT_OBJECT,
786 							  andd->device_name,
787 							  &h))) {
788 				pr_err("Failed to find handle for ACPI object %s\n",
789 				       andd->device_name);
790 				continue;
791 			}
792 			if (acpi_bus_get_device(h, &adev)) {
793 				pr_err("Failed to get device for ACPI object %s\n",
794 				       andd->device_name);
795 				continue;
796 			}
797 			dmar_acpi_insert_dev_scope(andd->device_number, adev);
798 		}
799 	}
800 	return 0;
801 }
802 
803 int __init dmar_dev_scope_init(void)
804 {
805 	struct pci_dev *dev = NULL;
806 	struct dmar_pci_notify_info *info;
807 
808 	if (dmar_dev_scope_status != 1)
809 		return dmar_dev_scope_status;
810 
811 	if (list_empty(&dmar_drhd_units)) {
812 		dmar_dev_scope_status = -ENODEV;
813 	} else {
814 		dmar_dev_scope_status = 0;
815 
816 		dmar_acpi_dev_scope_init();
817 
818 		for_each_pci_dev(dev) {
819 			if (dev->is_virtfn)
820 				continue;
821 
822 			info = dmar_alloc_pci_notify_info(dev,
823 					BUS_NOTIFY_ADD_DEVICE);
824 			if (!info) {
825 				return dmar_dev_scope_status;
826 			} else {
827 				dmar_pci_bus_add_dev(info);
828 				dmar_free_pci_notify_info(info);
829 			}
830 		}
831 	}
832 
833 	return dmar_dev_scope_status;
834 }
835 
836 void __init dmar_register_bus_notifier(void)
837 {
838 	bus_register_notifier(&pci_bus_type, &dmar_pci_bus_nb);
839 }
840 
841 
842 int __init dmar_table_init(void)
843 {
844 	static int dmar_table_initialized;
845 	int ret;
846 
847 	if (dmar_table_initialized == 0) {
848 		ret = parse_dmar_table();
849 		if (ret < 0) {
850 			if (ret != -ENODEV)
851 				pr_info("Parse DMAR table failure.\n");
852 		} else  if (list_empty(&dmar_drhd_units)) {
853 			pr_info("No DMAR devices found\n");
854 			ret = -ENODEV;
855 		}
856 
857 		if (ret < 0)
858 			dmar_table_initialized = ret;
859 		else
860 			dmar_table_initialized = 1;
861 	}
862 
863 	return dmar_table_initialized < 0 ? dmar_table_initialized : 0;
864 }
865 
866 static void warn_invalid_dmar(u64 addr, const char *message)
867 {
868 	pr_warn_once(FW_BUG
869 		"Your BIOS is broken; DMAR reported at address %llx%s!\n"
870 		"BIOS vendor: %s; Ver: %s; Product Version: %s\n",
871 		addr, message,
872 		dmi_get_system_info(DMI_BIOS_VENDOR),
873 		dmi_get_system_info(DMI_BIOS_VERSION),
874 		dmi_get_system_info(DMI_PRODUCT_VERSION));
875 	add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
876 }
877 
878 static int __ref
879 dmar_validate_one_drhd(struct acpi_dmar_header *entry, void *arg)
880 {
881 	struct acpi_dmar_hardware_unit *drhd;
882 	void __iomem *addr;
883 	u64 cap, ecap;
884 
885 	drhd = (void *)entry;
886 	if (!drhd->address) {
887 		warn_invalid_dmar(0, "");
888 		return -EINVAL;
889 	}
890 
891 	if (arg)
892 		addr = ioremap(drhd->address, VTD_PAGE_SIZE);
893 	else
894 		addr = early_ioremap(drhd->address, VTD_PAGE_SIZE);
895 	if (!addr) {
896 		pr_warn("Can't validate DRHD address: %llx\n", drhd->address);
897 		return -EINVAL;
898 	}
899 
900 	cap = dmar_readq(addr + DMAR_CAP_REG);
901 	ecap = dmar_readq(addr + DMAR_ECAP_REG);
902 
903 	if (arg)
904 		iounmap(addr);
905 	else
906 		early_iounmap(addr, VTD_PAGE_SIZE);
907 
908 	if (cap == (uint64_t)-1 && ecap == (uint64_t)-1) {
909 		warn_invalid_dmar(drhd->address, " returns all ones");
910 		return -EINVAL;
911 	}
912 
913 	return 0;
914 }
915 
916 int __init detect_intel_iommu(void)
917 {
918 	int ret;
919 	struct dmar_res_callback validate_drhd_cb = {
920 		.cb[ACPI_DMAR_TYPE_HARDWARE_UNIT] = &dmar_validate_one_drhd,
921 		.ignore_unhandled = true,
922 	};
923 
924 	down_write(&dmar_global_lock);
925 	ret = dmar_table_detect();
926 	if (!ret)
927 		ret = dmar_walk_dmar_table((struct acpi_table_dmar *)dmar_tbl,
928 					   &validate_drhd_cb);
929 	if (!ret && !no_iommu && !iommu_detected &&
930 	    (!dmar_disabled || dmar_platform_optin())) {
931 		iommu_detected = 1;
932 		/* Make sure ACS will be enabled */
933 		pci_request_acs();
934 	}
935 
936 #ifdef CONFIG_X86
937 	if (!ret) {
938 		x86_init.iommu.iommu_init = intel_iommu_init;
939 		x86_platform.iommu_shutdown = intel_iommu_shutdown;
940 	}
941 
942 #endif
943 
944 	if (dmar_tbl) {
945 		acpi_put_table(dmar_tbl);
946 		dmar_tbl = NULL;
947 	}
948 	up_write(&dmar_global_lock);
949 
950 	return ret ? ret : 1;
951 }
952 
953 static void unmap_iommu(struct intel_iommu *iommu)
954 {
955 	iounmap(iommu->reg);
956 	release_mem_region(iommu->reg_phys, iommu->reg_size);
957 }
958 
959 /**
960  * map_iommu: map the iommu's registers
961  * @iommu: the iommu to map
962  * @phys_addr: the physical address of the base resgister
963  *
964  * Memory map the iommu's registers.  Start w/ a single page, and
965  * possibly expand if that turns out to be insufficent.
966  */
967 static int map_iommu(struct intel_iommu *iommu, u64 phys_addr)
968 {
969 	int map_size, err=0;
970 
971 	iommu->reg_phys = phys_addr;
972 	iommu->reg_size = VTD_PAGE_SIZE;
973 
974 	if (!request_mem_region(iommu->reg_phys, iommu->reg_size, iommu->name)) {
975 		pr_err("Can't reserve memory\n");
976 		err = -EBUSY;
977 		goto out;
978 	}
979 
980 	iommu->reg = ioremap(iommu->reg_phys, iommu->reg_size);
981 	if (!iommu->reg) {
982 		pr_err("Can't map the region\n");
983 		err = -ENOMEM;
984 		goto release;
985 	}
986 
987 	iommu->cap = dmar_readq(iommu->reg + DMAR_CAP_REG);
988 	iommu->ecap = dmar_readq(iommu->reg + DMAR_ECAP_REG);
989 
990 	if (iommu->cap == (uint64_t)-1 && iommu->ecap == (uint64_t)-1) {
991 		err = -EINVAL;
992 		warn_invalid_dmar(phys_addr, " returns all ones");
993 		goto unmap;
994 	}
995 	if (ecap_vcs(iommu->ecap))
996 		iommu->vccap = dmar_readq(iommu->reg + DMAR_VCCAP_REG);
997 
998 	/* the registers might be more than one page */
999 	map_size = max_t(int, ecap_max_iotlb_offset(iommu->ecap),
1000 			 cap_max_fault_reg_offset(iommu->cap));
1001 	map_size = VTD_PAGE_ALIGN(map_size);
1002 	if (map_size > iommu->reg_size) {
1003 		iounmap(iommu->reg);
1004 		release_mem_region(iommu->reg_phys, iommu->reg_size);
1005 		iommu->reg_size = map_size;
1006 		if (!request_mem_region(iommu->reg_phys, iommu->reg_size,
1007 					iommu->name)) {
1008 			pr_err("Can't reserve memory\n");
1009 			err = -EBUSY;
1010 			goto out;
1011 		}
1012 		iommu->reg = ioremap(iommu->reg_phys, iommu->reg_size);
1013 		if (!iommu->reg) {
1014 			pr_err("Can't map the region\n");
1015 			err = -ENOMEM;
1016 			goto release;
1017 		}
1018 	}
1019 	err = 0;
1020 	goto out;
1021 
1022 unmap:
1023 	iounmap(iommu->reg);
1024 release:
1025 	release_mem_region(iommu->reg_phys, iommu->reg_size);
1026 out:
1027 	return err;
1028 }
1029 
1030 static int dmar_alloc_seq_id(struct intel_iommu *iommu)
1031 {
1032 	iommu->seq_id = find_first_zero_bit(dmar_seq_ids,
1033 					    DMAR_UNITS_SUPPORTED);
1034 	if (iommu->seq_id >= DMAR_UNITS_SUPPORTED) {
1035 		iommu->seq_id = -1;
1036 	} else {
1037 		set_bit(iommu->seq_id, dmar_seq_ids);
1038 		sprintf(iommu->name, "dmar%d", iommu->seq_id);
1039 	}
1040 
1041 	return iommu->seq_id;
1042 }
1043 
1044 static void dmar_free_seq_id(struct intel_iommu *iommu)
1045 {
1046 	if (iommu->seq_id >= 0) {
1047 		clear_bit(iommu->seq_id, dmar_seq_ids);
1048 		iommu->seq_id = -1;
1049 	}
1050 }
1051 
1052 static int alloc_iommu(struct dmar_drhd_unit *drhd)
1053 {
1054 	struct intel_iommu *iommu;
1055 	u32 ver, sts;
1056 	int agaw = -1;
1057 	int msagaw = -1;
1058 	int err;
1059 
1060 	if (!drhd->reg_base_addr) {
1061 		warn_invalid_dmar(0, "");
1062 		return -EINVAL;
1063 	}
1064 
1065 	iommu = kzalloc(sizeof(*iommu), GFP_KERNEL);
1066 	if (!iommu)
1067 		return -ENOMEM;
1068 
1069 	if (dmar_alloc_seq_id(iommu) < 0) {
1070 		pr_err("Failed to allocate seq_id\n");
1071 		err = -ENOSPC;
1072 		goto error;
1073 	}
1074 
1075 	err = map_iommu(iommu, drhd->reg_base_addr);
1076 	if (err) {
1077 		pr_err("Failed to map %s\n", iommu->name);
1078 		goto error_free_seq_id;
1079 	}
1080 
1081 	err = -EINVAL;
1082 	if (cap_sagaw(iommu->cap) == 0) {
1083 		pr_info("%s: No supported address widths. Not attempting DMA translation.\n",
1084 			iommu->name);
1085 		drhd->ignored = 1;
1086 	}
1087 
1088 	if (!drhd->ignored) {
1089 		agaw = iommu_calculate_agaw(iommu);
1090 		if (agaw < 0) {
1091 			pr_err("Cannot get a valid agaw for iommu (seq_id = %d)\n",
1092 			       iommu->seq_id);
1093 			drhd->ignored = 1;
1094 		}
1095 	}
1096 	if (!drhd->ignored) {
1097 		msagaw = iommu_calculate_max_sagaw(iommu);
1098 		if (msagaw < 0) {
1099 			pr_err("Cannot get a valid max agaw for iommu (seq_id = %d)\n",
1100 			       iommu->seq_id);
1101 			drhd->ignored = 1;
1102 			agaw = -1;
1103 		}
1104 	}
1105 	iommu->agaw = agaw;
1106 	iommu->msagaw = msagaw;
1107 	iommu->segment = drhd->segment;
1108 
1109 	iommu->node = NUMA_NO_NODE;
1110 
1111 	ver = readl(iommu->reg + DMAR_VER_REG);
1112 	pr_info("%s: reg_base_addr %llx ver %d:%d cap %llx ecap %llx\n",
1113 		iommu->name,
1114 		(unsigned long long)drhd->reg_base_addr,
1115 		DMAR_VER_MAJOR(ver), DMAR_VER_MINOR(ver),
1116 		(unsigned long long)iommu->cap,
1117 		(unsigned long long)iommu->ecap);
1118 
1119 	/* Reflect status in gcmd */
1120 	sts = readl(iommu->reg + DMAR_GSTS_REG);
1121 	if (sts & DMA_GSTS_IRES)
1122 		iommu->gcmd |= DMA_GCMD_IRE;
1123 	if (sts & DMA_GSTS_TES)
1124 		iommu->gcmd |= DMA_GCMD_TE;
1125 	if (sts & DMA_GSTS_QIES)
1126 		iommu->gcmd |= DMA_GCMD_QIE;
1127 
1128 	raw_spin_lock_init(&iommu->register_lock);
1129 
1130 	/*
1131 	 * This is only for hotplug; at boot time intel_iommu_enabled won't
1132 	 * be set yet. When intel_iommu_init() runs, it registers the units
1133 	 * present at boot time, then sets intel_iommu_enabled.
1134 	 */
1135 	if (intel_iommu_enabled && !drhd->ignored) {
1136 		err = iommu_device_sysfs_add(&iommu->iommu, NULL,
1137 					     intel_iommu_groups,
1138 					     "%s", iommu->name);
1139 		if (err)
1140 			goto err_unmap;
1141 
1142 		err = iommu_device_register(&iommu->iommu, &intel_iommu_ops, NULL);
1143 		if (err)
1144 			goto err_sysfs;
1145 	}
1146 
1147 	drhd->iommu = iommu;
1148 	iommu->drhd = drhd;
1149 
1150 	return 0;
1151 
1152 err_sysfs:
1153 	iommu_device_sysfs_remove(&iommu->iommu);
1154 err_unmap:
1155 	unmap_iommu(iommu);
1156 error_free_seq_id:
1157 	dmar_free_seq_id(iommu);
1158 error:
1159 	kfree(iommu);
1160 	return err;
1161 }
1162 
1163 static void free_iommu(struct intel_iommu *iommu)
1164 {
1165 	if (intel_iommu_enabled && !iommu->drhd->ignored) {
1166 		iommu_device_unregister(&iommu->iommu);
1167 		iommu_device_sysfs_remove(&iommu->iommu);
1168 	}
1169 
1170 	if (iommu->irq) {
1171 		if (iommu->pr_irq) {
1172 			free_irq(iommu->pr_irq, iommu);
1173 			dmar_free_hwirq(iommu->pr_irq);
1174 			iommu->pr_irq = 0;
1175 		}
1176 		free_irq(iommu->irq, iommu);
1177 		dmar_free_hwirq(iommu->irq);
1178 		iommu->irq = 0;
1179 	}
1180 
1181 	if (iommu->qi) {
1182 		free_page((unsigned long)iommu->qi->desc);
1183 		kfree(iommu->qi->desc_status);
1184 		kfree(iommu->qi);
1185 	}
1186 
1187 	if (iommu->reg)
1188 		unmap_iommu(iommu);
1189 
1190 	dmar_free_seq_id(iommu);
1191 	kfree(iommu);
1192 }
1193 
1194 /*
1195  * Reclaim all the submitted descriptors which have completed its work.
1196  */
1197 static inline void reclaim_free_desc(struct q_inval *qi)
1198 {
1199 	while (qi->desc_status[qi->free_tail] == QI_DONE ||
1200 	       qi->desc_status[qi->free_tail] == QI_ABORT) {
1201 		qi->desc_status[qi->free_tail] = QI_FREE;
1202 		qi->free_tail = (qi->free_tail + 1) % QI_LENGTH;
1203 		qi->free_cnt++;
1204 	}
1205 }
1206 
1207 static const char *qi_type_string(u8 type)
1208 {
1209 	switch (type) {
1210 	case QI_CC_TYPE:
1211 		return "Context-cache Invalidation";
1212 	case QI_IOTLB_TYPE:
1213 		return "IOTLB Invalidation";
1214 	case QI_DIOTLB_TYPE:
1215 		return "Device-TLB Invalidation";
1216 	case QI_IEC_TYPE:
1217 		return "Interrupt Entry Cache Invalidation";
1218 	case QI_IWD_TYPE:
1219 		return "Invalidation Wait";
1220 	case QI_EIOTLB_TYPE:
1221 		return "PASID-based IOTLB Invalidation";
1222 	case QI_PC_TYPE:
1223 		return "PASID-cache Invalidation";
1224 	case QI_DEIOTLB_TYPE:
1225 		return "PASID-based Device-TLB Invalidation";
1226 	case QI_PGRP_RESP_TYPE:
1227 		return "Page Group Response";
1228 	default:
1229 		return "UNKNOWN";
1230 	}
1231 }
1232 
1233 static void qi_dump_fault(struct intel_iommu *iommu, u32 fault)
1234 {
1235 	unsigned int head = dmar_readl(iommu->reg + DMAR_IQH_REG);
1236 	u64 iqe_err = dmar_readq(iommu->reg + DMAR_IQER_REG);
1237 	struct qi_desc *desc = iommu->qi->desc + head;
1238 
1239 	if (fault & DMA_FSTS_IQE)
1240 		pr_err("VT-d detected Invalidation Queue Error: Reason %llx",
1241 		       DMAR_IQER_REG_IQEI(iqe_err));
1242 	if (fault & DMA_FSTS_ITE)
1243 		pr_err("VT-d detected Invalidation Time-out Error: SID %llx",
1244 		       DMAR_IQER_REG_ITESID(iqe_err));
1245 	if (fault & DMA_FSTS_ICE)
1246 		pr_err("VT-d detected Invalidation Completion Error: SID %llx",
1247 		       DMAR_IQER_REG_ICESID(iqe_err));
1248 
1249 	pr_err("QI HEAD: %s qw0 = 0x%llx, qw1 = 0x%llx\n",
1250 	       qi_type_string(desc->qw0 & 0xf),
1251 	       (unsigned long long)desc->qw0,
1252 	       (unsigned long long)desc->qw1);
1253 
1254 	head = ((head >> qi_shift(iommu)) + QI_LENGTH - 1) % QI_LENGTH;
1255 	head <<= qi_shift(iommu);
1256 	desc = iommu->qi->desc + head;
1257 
1258 	pr_err("QI PRIOR: %s qw0 = 0x%llx, qw1 = 0x%llx\n",
1259 	       qi_type_string(desc->qw0 & 0xf),
1260 	       (unsigned long long)desc->qw0,
1261 	       (unsigned long long)desc->qw1);
1262 }
1263 
1264 static int qi_check_fault(struct intel_iommu *iommu, int index, int wait_index)
1265 {
1266 	u32 fault;
1267 	int head, tail;
1268 	struct q_inval *qi = iommu->qi;
1269 	int shift = qi_shift(iommu);
1270 
1271 	if (qi->desc_status[wait_index] == QI_ABORT)
1272 		return -EAGAIN;
1273 
1274 	fault = readl(iommu->reg + DMAR_FSTS_REG);
1275 	if (fault & (DMA_FSTS_IQE | DMA_FSTS_ITE | DMA_FSTS_ICE))
1276 		qi_dump_fault(iommu, fault);
1277 
1278 	/*
1279 	 * If IQE happens, the head points to the descriptor associated
1280 	 * with the error. No new descriptors are fetched until the IQE
1281 	 * is cleared.
1282 	 */
1283 	if (fault & DMA_FSTS_IQE) {
1284 		head = readl(iommu->reg + DMAR_IQH_REG);
1285 		if ((head >> shift) == index) {
1286 			struct qi_desc *desc = qi->desc + head;
1287 
1288 			/*
1289 			 * desc->qw2 and desc->qw3 are either reserved or
1290 			 * used by software as private data. We won't print
1291 			 * out these two qw's for security consideration.
1292 			 */
1293 			memcpy(desc, qi->desc + (wait_index << shift),
1294 			       1 << shift);
1295 			writel(DMA_FSTS_IQE, iommu->reg + DMAR_FSTS_REG);
1296 			pr_info("Invalidation Queue Error (IQE) cleared\n");
1297 			return -EINVAL;
1298 		}
1299 	}
1300 
1301 	/*
1302 	 * If ITE happens, all pending wait_desc commands are aborted.
1303 	 * No new descriptors are fetched until the ITE is cleared.
1304 	 */
1305 	if (fault & DMA_FSTS_ITE) {
1306 		head = readl(iommu->reg + DMAR_IQH_REG);
1307 		head = ((head >> shift) - 1 + QI_LENGTH) % QI_LENGTH;
1308 		head |= 1;
1309 		tail = readl(iommu->reg + DMAR_IQT_REG);
1310 		tail = ((tail >> shift) - 1 + QI_LENGTH) % QI_LENGTH;
1311 
1312 		writel(DMA_FSTS_ITE, iommu->reg + DMAR_FSTS_REG);
1313 		pr_info("Invalidation Time-out Error (ITE) cleared\n");
1314 
1315 		do {
1316 			if (qi->desc_status[head] == QI_IN_USE)
1317 				qi->desc_status[head] = QI_ABORT;
1318 			head = (head - 2 + QI_LENGTH) % QI_LENGTH;
1319 		} while (head != tail);
1320 
1321 		if (qi->desc_status[wait_index] == QI_ABORT)
1322 			return -EAGAIN;
1323 	}
1324 
1325 	if (fault & DMA_FSTS_ICE) {
1326 		writel(DMA_FSTS_ICE, iommu->reg + DMAR_FSTS_REG);
1327 		pr_info("Invalidation Completion Error (ICE) cleared\n");
1328 	}
1329 
1330 	return 0;
1331 }
1332 
1333 /*
1334  * Function to submit invalidation descriptors of all types to the queued
1335  * invalidation interface(QI). Multiple descriptors can be submitted at a
1336  * time, a wait descriptor will be appended to each submission to ensure
1337  * hardware has completed the invalidation before return. Wait descriptors
1338  * can be part of the submission but it will not be polled for completion.
1339  */
1340 int qi_submit_sync(struct intel_iommu *iommu, struct qi_desc *desc,
1341 		   unsigned int count, unsigned long options)
1342 {
1343 	struct q_inval *qi = iommu->qi;
1344 	s64 devtlb_start_ktime = 0;
1345 	s64 iotlb_start_ktime = 0;
1346 	s64 iec_start_ktime = 0;
1347 	struct qi_desc wait_desc;
1348 	int wait_index, index;
1349 	unsigned long flags;
1350 	int offset, shift;
1351 	int rc, i;
1352 	u64 type;
1353 
1354 	if (!qi)
1355 		return 0;
1356 
1357 	type = desc->qw0 & GENMASK_ULL(3, 0);
1358 
1359 	if ((type == QI_IOTLB_TYPE || type == QI_EIOTLB_TYPE) &&
1360 	    dmar_latency_enabled(iommu, DMAR_LATENCY_INV_IOTLB))
1361 		iotlb_start_ktime = ktime_to_ns(ktime_get());
1362 
1363 	if ((type == QI_DIOTLB_TYPE || type == QI_DEIOTLB_TYPE) &&
1364 	    dmar_latency_enabled(iommu, DMAR_LATENCY_INV_DEVTLB))
1365 		devtlb_start_ktime = ktime_to_ns(ktime_get());
1366 
1367 	if (type == QI_IEC_TYPE &&
1368 	    dmar_latency_enabled(iommu, DMAR_LATENCY_INV_IEC))
1369 		iec_start_ktime = ktime_to_ns(ktime_get());
1370 
1371 restart:
1372 	rc = 0;
1373 
1374 	raw_spin_lock_irqsave(&qi->q_lock, flags);
1375 	/*
1376 	 * Check if we have enough empty slots in the queue to submit,
1377 	 * the calculation is based on:
1378 	 * # of desc + 1 wait desc + 1 space between head and tail
1379 	 */
1380 	while (qi->free_cnt < count + 2) {
1381 		raw_spin_unlock_irqrestore(&qi->q_lock, flags);
1382 		cpu_relax();
1383 		raw_spin_lock_irqsave(&qi->q_lock, flags);
1384 	}
1385 
1386 	index = qi->free_head;
1387 	wait_index = (index + count) % QI_LENGTH;
1388 	shift = qi_shift(iommu);
1389 
1390 	for (i = 0; i < count; i++) {
1391 		offset = ((index + i) % QI_LENGTH) << shift;
1392 		memcpy(qi->desc + offset, &desc[i], 1 << shift);
1393 		qi->desc_status[(index + i) % QI_LENGTH] = QI_IN_USE;
1394 		trace_qi_submit(iommu, desc[i].qw0, desc[i].qw1,
1395 				desc[i].qw2, desc[i].qw3);
1396 	}
1397 	qi->desc_status[wait_index] = QI_IN_USE;
1398 
1399 	wait_desc.qw0 = QI_IWD_STATUS_DATA(QI_DONE) |
1400 			QI_IWD_STATUS_WRITE | QI_IWD_TYPE;
1401 	if (options & QI_OPT_WAIT_DRAIN)
1402 		wait_desc.qw0 |= QI_IWD_PRQ_DRAIN;
1403 	wait_desc.qw1 = virt_to_phys(&qi->desc_status[wait_index]);
1404 	wait_desc.qw2 = 0;
1405 	wait_desc.qw3 = 0;
1406 
1407 	offset = wait_index << shift;
1408 	memcpy(qi->desc + offset, &wait_desc, 1 << shift);
1409 
1410 	qi->free_head = (qi->free_head + count + 1) % QI_LENGTH;
1411 	qi->free_cnt -= count + 1;
1412 
1413 	/*
1414 	 * update the HW tail register indicating the presence of
1415 	 * new descriptors.
1416 	 */
1417 	writel(qi->free_head << shift, iommu->reg + DMAR_IQT_REG);
1418 
1419 	while (qi->desc_status[wait_index] != QI_DONE) {
1420 		/*
1421 		 * We will leave the interrupts disabled, to prevent interrupt
1422 		 * context to queue another cmd while a cmd is already submitted
1423 		 * and waiting for completion on this cpu. This is to avoid
1424 		 * a deadlock where the interrupt context can wait indefinitely
1425 		 * for free slots in the queue.
1426 		 */
1427 		rc = qi_check_fault(iommu, index, wait_index);
1428 		if (rc)
1429 			break;
1430 
1431 		raw_spin_unlock(&qi->q_lock);
1432 		cpu_relax();
1433 		raw_spin_lock(&qi->q_lock);
1434 	}
1435 
1436 	for (i = 0; i < count; i++)
1437 		qi->desc_status[(index + i) % QI_LENGTH] = QI_DONE;
1438 
1439 	reclaim_free_desc(qi);
1440 	raw_spin_unlock_irqrestore(&qi->q_lock, flags);
1441 
1442 	if (rc == -EAGAIN)
1443 		goto restart;
1444 
1445 	if (iotlb_start_ktime)
1446 		dmar_latency_update(iommu, DMAR_LATENCY_INV_IOTLB,
1447 				ktime_to_ns(ktime_get()) - iotlb_start_ktime);
1448 
1449 	if (devtlb_start_ktime)
1450 		dmar_latency_update(iommu, DMAR_LATENCY_INV_DEVTLB,
1451 				ktime_to_ns(ktime_get()) - devtlb_start_ktime);
1452 
1453 	if (iec_start_ktime)
1454 		dmar_latency_update(iommu, DMAR_LATENCY_INV_IEC,
1455 				ktime_to_ns(ktime_get()) - iec_start_ktime);
1456 
1457 	return rc;
1458 }
1459 
1460 /*
1461  * Flush the global interrupt entry cache.
1462  */
1463 void qi_global_iec(struct intel_iommu *iommu)
1464 {
1465 	struct qi_desc desc;
1466 
1467 	desc.qw0 = QI_IEC_TYPE;
1468 	desc.qw1 = 0;
1469 	desc.qw2 = 0;
1470 	desc.qw3 = 0;
1471 
1472 	/* should never fail */
1473 	qi_submit_sync(iommu, &desc, 1, 0);
1474 }
1475 
1476 void qi_flush_context(struct intel_iommu *iommu, u16 did, u16 sid, u8 fm,
1477 		      u64 type)
1478 {
1479 	struct qi_desc desc;
1480 
1481 	desc.qw0 = QI_CC_FM(fm) | QI_CC_SID(sid) | QI_CC_DID(did)
1482 			| QI_CC_GRAN(type) | QI_CC_TYPE;
1483 	desc.qw1 = 0;
1484 	desc.qw2 = 0;
1485 	desc.qw3 = 0;
1486 
1487 	qi_submit_sync(iommu, &desc, 1, 0);
1488 }
1489 
1490 void qi_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr,
1491 		    unsigned int size_order, u64 type)
1492 {
1493 	u8 dw = 0, dr = 0;
1494 
1495 	struct qi_desc desc;
1496 	int ih = 0;
1497 
1498 	if (cap_write_drain(iommu->cap))
1499 		dw = 1;
1500 
1501 	if (cap_read_drain(iommu->cap))
1502 		dr = 1;
1503 
1504 	desc.qw0 = QI_IOTLB_DID(did) | QI_IOTLB_DR(dr) | QI_IOTLB_DW(dw)
1505 		| QI_IOTLB_GRAN(type) | QI_IOTLB_TYPE;
1506 	desc.qw1 = QI_IOTLB_ADDR(addr) | QI_IOTLB_IH(ih)
1507 		| QI_IOTLB_AM(size_order);
1508 	desc.qw2 = 0;
1509 	desc.qw3 = 0;
1510 
1511 	qi_submit_sync(iommu, &desc, 1, 0);
1512 }
1513 
1514 void qi_flush_dev_iotlb(struct intel_iommu *iommu, u16 sid, u16 pfsid,
1515 			u16 qdep, u64 addr, unsigned mask)
1516 {
1517 	struct qi_desc desc;
1518 
1519 	if (mask) {
1520 		addr |= (1ULL << (VTD_PAGE_SHIFT + mask - 1)) - 1;
1521 		desc.qw1 = QI_DEV_IOTLB_ADDR(addr) | QI_DEV_IOTLB_SIZE;
1522 	} else
1523 		desc.qw1 = QI_DEV_IOTLB_ADDR(addr);
1524 
1525 	if (qdep >= QI_DEV_IOTLB_MAX_INVS)
1526 		qdep = 0;
1527 
1528 	desc.qw0 = QI_DEV_IOTLB_SID(sid) | QI_DEV_IOTLB_QDEP(qdep) |
1529 		   QI_DIOTLB_TYPE | QI_DEV_IOTLB_PFSID(pfsid);
1530 	desc.qw2 = 0;
1531 	desc.qw3 = 0;
1532 
1533 	qi_submit_sync(iommu, &desc, 1, 0);
1534 }
1535 
1536 /* PASID-based IOTLB invalidation */
1537 void qi_flush_piotlb(struct intel_iommu *iommu, u16 did, u32 pasid, u64 addr,
1538 		     unsigned long npages, bool ih)
1539 {
1540 	struct qi_desc desc = {.qw2 = 0, .qw3 = 0};
1541 
1542 	/*
1543 	 * npages == -1 means a PASID-selective invalidation, otherwise,
1544 	 * a positive value for Page-selective-within-PASID invalidation.
1545 	 * 0 is not a valid input.
1546 	 */
1547 	if (WARN_ON(!npages)) {
1548 		pr_err("Invalid input npages = %ld\n", npages);
1549 		return;
1550 	}
1551 
1552 	if (npages == -1) {
1553 		desc.qw0 = QI_EIOTLB_PASID(pasid) |
1554 				QI_EIOTLB_DID(did) |
1555 				QI_EIOTLB_GRAN(QI_GRAN_NONG_PASID) |
1556 				QI_EIOTLB_TYPE;
1557 		desc.qw1 = 0;
1558 	} else {
1559 		int mask = ilog2(__roundup_pow_of_two(npages));
1560 		unsigned long align = (1ULL << (VTD_PAGE_SHIFT + mask));
1561 
1562 		if (WARN_ON_ONCE(!IS_ALIGNED(addr, align)))
1563 			addr = ALIGN_DOWN(addr, align);
1564 
1565 		desc.qw0 = QI_EIOTLB_PASID(pasid) |
1566 				QI_EIOTLB_DID(did) |
1567 				QI_EIOTLB_GRAN(QI_GRAN_PSI_PASID) |
1568 				QI_EIOTLB_TYPE;
1569 		desc.qw1 = QI_EIOTLB_ADDR(addr) |
1570 				QI_EIOTLB_IH(ih) |
1571 				QI_EIOTLB_AM(mask);
1572 	}
1573 
1574 	qi_submit_sync(iommu, &desc, 1, 0);
1575 }
1576 
1577 /* PASID-based device IOTLB Invalidate */
1578 void qi_flush_dev_iotlb_pasid(struct intel_iommu *iommu, u16 sid, u16 pfsid,
1579 			      u32 pasid,  u16 qdep, u64 addr, unsigned int size_order)
1580 {
1581 	unsigned long mask = 1UL << (VTD_PAGE_SHIFT + size_order - 1);
1582 	struct qi_desc desc = {.qw1 = 0, .qw2 = 0, .qw3 = 0};
1583 
1584 	desc.qw0 = QI_DEV_EIOTLB_PASID(pasid) | QI_DEV_EIOTLB_SID(sid) |
1585 		QI_DEV_EIOTLB_QDEP(qdep) | QI_DEIOTLB_TYPE |
1586 		QI_DEV_IOTLB_PFSID(pfsid);
1587 
1588 	/*
1589 	 * If S bit is 0, we only flush a single page. If S bit is set,
1590 	 * The least significant zero bit indicates the invalidation address
1591 	 * range. VT-d spec 6.5.2.6.
1592 	 * e.g. address bit 12[0] indicates 8KB, 13[0] indicates 16KB.
1593 	 * size order = 0 is PAGE_SIZE 4KB
1594 	 * Max Invs Pending (MIP) is set to 0 for now until we have DIT in
1595 	 * ECAP.
1596 	 */
1597 	if (!IS_ALIGNED(addr, VTD_PAGE_SIZE << size_order))
1598 		pr_warn_ratelimited("Invalidate non-aligned address %llx, order %d\n",
1599 				    addr, size_order);
1600 
1601 	/* Take page address */
1602 	desc.qw1 = QI_DEV_EIOTLB_ADDR(addr);
1603 
1604 	if (size_order) {
1605 		/*
1606 		 * Existing 0s in address below size_order may be the least
1607 		 * significant bit, we must set them to 1s to avoid having
1608 		 * smaller size than desired.
1609 		 */
1610 		desc.qw1 |= GENMASK_ULL(size_order + VTD_PAGE_SHIFT - 1,
1611 					VTD_PAGE_SHIFT);
1612 		/* Clear size_order bit to indicate size */
1613 		desc.qw1 &= ~mask;
1614 		/* Set the S bit to indicate flushing more than 1 page */
1615 		desc.qw1 |= QI_DEV_EIOTLB_SIZE;
1616 	}
1617 
1618 	qi_submit_sync(iommu, &desc, 1, 0);
1619 }
1620 
1621 void qi_flush_pasid_cache(struct intel_iommu *iommu, u16 did,
1622 			  u64 granu, u32 pasid)
1623 {
1624 	struct qi_desc desc = {.qw1 = 0, .qw2 = 0, .qw3 = 0};
1625 
1626 	desc.qw0 = QI_PC_PASID(pasid) | QI_PC_DID(did) |
1627 			QI_PC_GRAN(granu) | QI_PC_TYPE;
1628 	qi_submit_sync(iommu, &desc, 1, 0);
1629 }
1630 
1631 /*
1632  * Disable Queued Invalidation interface.
1633  */
1634 void dmar_disable_qi(struct intel_iommu *iommu)
1635 {
1636 	unsigned long flags;
1637 	u32 sts;
1638 	cycles_t start_time = get_cycles();
1639 
1640 	if (!ecap_qis(iommu->ecap))
1641 		return;
1642 
1643 	raw_spin_lock_irqsave(&iommu->register_lock, flags);
1644 
1645 	sts =  readl(iommu->reg + DMAR_GSTS_REG);
1646 	if (!(sts & DMA_GSTS_QIES))
1647 		goto end;
1648 
1649 	/*
1650 	 * Give a chance to HW to complete the pending invalidation requests.
1651 	 */
1652 	while ((readl(iommu->reg + DMAR_IQT_REG) !=
1653 		readl(iommu->reg + DMAR_IQH_REG)) &&
1654 		(DMAR_OPERATION_TIMEOUT > (get_cycles() - start_time)))
1655 		cpu_relax();
1656 
1657 	iommu->gcmd &= ~DMA_GCMD_QIE;
1658 	writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
1659 
1660 	IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl,
1661 		      !(sts & DMA_GSTS_QIES), sts);
1662 end:
1663 	raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
1664 }
1665 
1666 /*
1667  * Enable queued invalidation.
1668  */
1669 static void __dmar_enable_qi(struct intel_iommu *iommu)
1670 {
1671 	u32 sts;
1672 	unsigned long flags;
1673 	struct q_inval *qi = iommu->qi;
1674 	u64 val = virt_to_phys(qi->desc);
1675 
1676 	qi->free_head = qi->free_tail = 0;
1677 	qi->free_cnt = QI_LENGTH;
1678 
1679 	/*
1680 	 * Set DW=1 and QS=1 in IQA_REG when Scalable Mode capability
1681 	 * is present.
1682 	 */
1683 	if (ecap_smts(iommu->ecap))
1684 		val |= (1 << 11) | 1;
1685 
1686 	raw_spin_lock_irqsave(&iommu->register_lock, flags);
1687 
1688 	/* write zero to the tail reg */
1689 	writel(0, iommu->reg + DMAR_IQT_REG);
1690 
1691 	dmar_writeq(iommu->reg + DMAR_IQA_REG, val);
1692 
1693 	iommu->gcmd |= DMA_GCMD_QIE;
1694 	writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
1695 
1696 	/* Make sure hardware complete it */
1697 	IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (sts & DMA_GSTS_QIES), sts);
1698 
1699 	raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
1700 }
1701 
1702 /*
1703  * Enable Queued Invalidation interface. This is a must to support
1704  * interrupt-remapping. Also used by DMA-remapping, which replaces
1705  * register based IOTLB invalidation.
1706  */
1707 int dmar_enable_qi(struct intel_iommu *iommu)
1708 {
1709 	struct q_inval *qi;
1710 	struct page *desc_page;
1711 
1712 	if (!ecap_qis(iommu->ecap))
1713 		return -ENOENT;
1714 
1715 	/*
1716 	 * queued invalidation is already setup and enabled.
1717 	 */
1718 	if (iommu->qi)
1719 		return 0;
1720 
1721 	iommu->qi = kmalloc(sizeof(*qi), GFP_ATOMIC);
1722 	if (!iommu->qi)
1723 		return -ENOMEM;
1724 
1725 	qi = iommu->qi;
1726 
1727 	/*
1728 	 * Need two pages to accommodate 256 descriptors of 256 bits each
1729 	 * if the remapping hardware supports scalable mode translation.
1730 	 */
1731 	desc_page = alloc_pages_node(iommu->node, GFP_ATOMIC | __GFP_ZERO,
1732 				     !!ecap_smts(iommu->ecap));
1733 	if (!desc_page) {
1734 		kfree(qi);
1735 		iommu->qi = NULL;
1736 		return -ENOMEM;
1737 	}
1738 
1739 	qi->desc = page_address(desc_page);
1740 
1741 	qi->desc_status = kcalloc(QI_LENGTH, sizeof(int), GFP_ATOMIC);
1742 	if (!qi->desc_status) {
1743 		free_page((unsigned long) qi->desc);
1744 		kfree(qi);
1745 		iommu->qi = NULL;
1746 		return -ENOMEM;
1747 	}
1748 
1749 	raw_spin_lock_init(&qi->q_lock);
1750 
1751 	__dmar_enable_qi(iommu);
1752 
1753 	return 0;
1754 }
1755 
1756 /* iommu interrupt handling. Most stuff are MSI-like. */
1757 
1758 enum faulttype {
1759 	DMA_REMAP,
1760 	INTR_REMAP,
1761 	UNKNOWN,
1762 };
1763 
1764 static const char *dma_remap_fault_reasons[] =
1765 {
1766 	"Software",
1767 	"Present bit in root entry is clear",
1768 	"Present bit in context entry is clear",
1769 	"Invalid context entry",
1770 	"Access beyond MGAW",
1771 	"PTE Write access is not set",
1772 	"PTE Read access is not set",
1773 	"Next page table ptr is invalid",
1774 	"Root table address invalid",
1775 	"Context table ptr is invalid",
1776 	"non-zero reserved fields in RTP",
1777 	"non-zero reserved fields in CTP",
1778 	"non-zero reserved fields in PTE",
1779 	"PCE for translation request specifies blocking",
1780 };
1781 
1782 static const char * const dma_remap_sm_fault_reasons[] = {
1783 	"SM: Invalid Root Table Address",
1784 	"SM: TTM 0 for request with PASID",
1785 	"SM: TTM 0 for page group request",
1786 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x33-0x37 */
1787 	"SM: Error attempting to access Root Entry",
1788 	"SM: Present bit in Root Entry is clear",
1789 	"SM: Non-zero reserved field set in Root Entry",
1790 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x3B-0x3F */
1791 	"SM: Error attempting to access Context Entry",
1792 	"SM: Present bit in Context Entry is clear",
1793 	"SM: Non-zero reserved field set in the Context Entry",
1794 	"SM: Invalid Context Entry",
1795 	"SM: DTE field in Context Entry is clear",
1796 	"SM: PASID Enable field in Context Entry is clear",
1797 	"SM: PASID is larger than the max in Context Entry",
1798 	"SM: PRE field in Context-Entry is clear",
1799 	"SM: RID_PASID field error in Context-Entry",
1800 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x49-0x4F */
1801 	"SM: Error attempting to access the PASID Directory Entry",
1802 	"SM: Present bit in Directory Entry is clear",
1803 	"SM: Non-zero reserved field set in PASID Directory Entry",
1804 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x53-0x57 */
1805 	"SM: Error attempting to access PASID Table Entry",
1806 	"SM: Present bit in PASID Table Entry is clear",
1807 	"SM: Non-zero reserved field set in PASID Table Entry",
1808 	"SM: Invalid Scalable-Mode PASID Table Entry",
1809 	"SM: ERE field is clear in PASID Table Entry",
1810 	"SM: SRE field is clear in PASID Table Entry",
1811 	"Unknown", "Unknown",/* 0x5E-0x5F */
1812 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x60-0x67 */
1813 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x68-0x6F */
1814 	"SM: Error attempting to access first-level paging entry",
1815 	"SM: Present bit in first-level paging entry is clear",
1816 	"SM: Non-zero reserved field set in first-level paging entry",
1817 	"SM: Error attempting to access FL-PML4 entry",
1818 	"SM: First-level entry address beyond MGAW in Nested translation",
1819 	"SM: Read permission error in FL-PML4 entry in Nested translation",
1820 	"SM: Read permission error in first-level paging entry in Nested translation",
1821 	"SM: Write permission error in first-level paging entry in Nested translation",
1822 	"SM: Error attempting to access second-level paging entry",
1823 	"SM: Read/Write permission error in second-level paging entry",
1824 	"SM: Non-zero reserved field set in second-level paging entry",
1825 	"SM: Invalid second-level page table pointer",
1826 	"SM: A/D bit update needed in second-level entry when set up in no snoop",
1827 	"Unknown", "Unknown", "Unknown", /* 0x7D-0x7F */
1828 	"SM: Address in first-level translation is not canonical",
1829 	"SM: U/S set 0 for first-level translation with user privilege",
1830 	"SM: No execute permission for request with PASID and ER=1",
1831 	"SM: Address beyond the DMA hardware max",
1832 	"SM: Second-level entry address beyond the max",
1833 	"SM: No write permission for Write/AtomicOp request",
1834 	"SM: No read permission for Read/AtomicOp request",
1835 	"SM: Invalid address-interrupt address",
1836 	"Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", "Unknown", /* 0x88-0x8F */
1837 	"SM: A/D bit update needed in first-level entry when set up in no snoop",
1838 };
1839 
1840 static const char *irq_remap_fault_reasons[] =
1841 {
1842 	"Detected reserved fields in the decoded interrupt-remapped request",
1843 	"Interrupt index exceeded the interrupt-remapping table size",
1844 	"Present field in the IRTE entry is clear",
1845 	"Error accessing interrupt-remapping table pointed by IRTA_REG",
1846 	"Detected reserved fields in the IRTE entry",
1847 	"Blocked a compatibility format interrupt request",
1848 	"Blocked an interrupt request due to source-id verification failure",
1849 };
1850 
1851 static const char *dmar_get_fault_reason(u8 fault_reason, int *fault_type)
1852 {
1853 	if (fault_reason >= 0x20 && (fault_reason - 0x20 <
1854 					ARRAY_SIZE(irq_remap_fault_reasons))) {
1855 		*fault_type = INTR_REMAP;
1856 		return irq_remap_fault_reasons[fault_reason - 0x20];
1857 	} else if (fault_reason >= 0x30 && (fault_reason - 0x30 <
1858 			ARRAY_SIZE(dma_remap_sm_fault_reasons))) {
1859 		*fault_type = DMA_REMAP;
1860 		return dma_remap_sm_fault_reasons[fault_reason - 0x30];
1861 	} else if (fault_reason < ARRAY_SIZE(dma_remap_fault_reasons)) {
1862 		*fault_type = DMA_REMAP;
1863 		return dma_remap_fault_reasons[fault_reason];
1864 	} else {
1865 		*fault_type = UNKNOWN;
1866 		return "Unknown";
1867 	}
1868 }
1869 
1870 
1871 static inline int dmar_msi_reg(struct intel_iommu *iommu, int irq)
1872 {
1873 	if (iommu->irq == irq)
1874 		return DMAR_FECTL_REG;
1875 	else if (iommu->pr_irq == irq)
1876 		return DMAR_PECTL_REG;
1877 	else
1878 		BUG();
1879 }
1880 
1881 void dmar_msi_unmask(struct irq_data *data)
1882 {
1883 	struct intel_iommu *iommu = irq_data_get_irq_handler_data(data);
1884 	int reg = dmar_msi_reg(iommu, data->irq);
1885 	unsigned long flag;
1886 
1887 	/* unmask it */
1888 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1889 	writel(0, iommu->reg + reg);
1890 	/* Read a reg to force flush the post write */
1891 	readl(iommu->reg + reg);
1892 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1893 }
1894 
1895 void dmar_msi_mask(struct irq_data *data)
1896 {
1897 	struct intel_iommu *iommu = irq_data_get_irq_handler_data(data);
1898 	int reg = dmar_msi_reg(iommu, data->irq);
1899 	unsigned long flag;
1900 
1901 	/* mask it */
1902 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1903 	writel(DMA_FECTL_IM, iommu->reg + reg);
1904 	/* Read a reg to force flush the post write */
1905 	readl(iommu->reg + reg);
1906 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1907 }
1908 
1909 void dmar_msi_write(int irq, struct msi_msg *msg)
1910 {
1911 	struct intel_iommu *iommu = irq_get_handler_data(irq);
1912 	int reg = dmar_msi_reg(iommu, irq);
1913 	unsigned long flag;
1914 
1915 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1916 	writel(msg->data, iommu->reg + reg + 4);
1917 	writel(msg->address_lo, iommu->reg + reg + 8);
1918 	writel(msg->address_hi, iommu->reg + reg + 12);
1919 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1920 }
1921 
1922 void dmar_msi_read(int irq, struct msi_msg *msg)
1923 {
1924 	struct intel_iommu *iommu = irq_get_handler_data(irq);
1925 	int reg = dmar_msi_reg(iommu, irq);
1926 	unsigned long flag;
1927 
1928 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1929 	msg->data = readl(iommu->reg + reg + 4);
1930 	msg->address_lo = readl(iommu->reg + reg + 8);
1931 	msg->address_hi = readl(iommu->reg + reg + 12);
1932 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
1933 }
1934 
1935 static int dmar_fault_do_one(struct intel_iommu *iommu, int type,
1936 		u8 fault_reason, u32 pasid, u16 source_id,
1937 		unsigned long long addr)
1938 {
1939 	const char *reason;
1940 	int fault_type;
1941 
1942 	reason = dmar_get_fault_reason(fault_reason, &fault_type);
1943 
1944 	if (fault_type == INTR_REMAP) {
1945 		pr_err("[INTR-REMAP] Request device [%02x:%02x.%d] fault index 0x%llx [fault reason 0x%02x] %s\n",
1946 		       source_id >> 8, PCI_SLOT(source_id & 0xFF),
1947 		       PCI_FUNC(source_id & 0xFF), addr >> 48,
1948 		       fault_reason, reason);
1949 
1950 		return 0;
1951 	}
1952 
1953 	if (pasid == INVALID_IOASID)
1954 		pr_err("[%s NO_PASID] Request device [%02x:%02x.%d] fault addr 0x%llx [fault reason 0x%02x] %s\n",
1955 		       type ? "DMA Read" : "DMA Write",
1956 		       source_id >> 8, PCI_SLOT(source_id & 0xFF),
1957 		       PCI_FUNC(source_id & 0xFF), addr,
1958 		       fault_reason, reason);
1959 	else
1960 		pr_err("[%s PASID 0x%x] Request device [%02x:%02x.%d] fault addr 0x%llx [fault reason 0x%02x] %s\n",
1961 		       type ? "DMA Read" : "DMA Write", pasid,
1962 		       source_id >> 8, PCI_SLOT(source_id & 0xFF),
1963 		       PCI_FUNC(source_id & 0xFF), addr,
1964 		       fault_reason, reason);
1965 
1966 	dmar_fault_dump_ptes(iommu, source_id, addr, pasid);
1967 
1968 	return 0;
1969 }
1970 
1971 #define PRIMARY_FAULT_REG_LEN (16)
1972 irqreturn_t dmar_fault(int irq, void *dev_id)
1973 {
1974 	struct intel_iommu *iommu = dev_id;
1975 	int reg, fault_index;
1976 	u32 fault_status;
1977 	unsigned long flag;
1978 	static DEFINE_RATELIMIT_STATE(rs,
1979 				      DEFAULT_RATELIMIT_INTERVAL,
1980 				      DEFAULT_RATELIMIT_BURST);
1981 
1982 	raw_spin_lock_irqsave(&iommu->register_lock, flag);
1983 	fault_status = readl(iommu->reg + DMAR_FSTS_REG);
1984 	if (fault_status && __ratelimit(&rs))
1985 		pr_err("DRHD: handling fault status reg %x\n", fault_status);
1986 
1987 	/* TBD: ignore advanced fault log currently */
1988 	if (!(fault_status & DMA_FSTS_PPF))
1989 		goto unlock_exit;
1990 
1991 	fault_index = dma_fsts_fault_record_index(fault_status);
1992 	reg = cap_fault_reg_offset(iommu->cap);
1993 	while (1) {
1994 		/* Disable printing, simply clear the fault when ratelimited */
1995 		bool ratelimited = !__ratelimit(&rs);
1996 		u8 fault_reason;
1997 		u16 source_id;
1998 		u64 guest_addr;
1999 		u32 pasid;
2000 		int type;
2001 		u32 data;
2002 		bool pasid_present;
2003 
2004 		/* highest 32 bits */
2005 		data = readl(iommu->reg + reg +
2006 				fault_index * PRIMARY_FAULT_REG_LEN + 12);
2007 		if (!(data & DMA_FRCD_F))
2008 			break;
2009 
2010 		if (!ratelimited) {
2011 			fault_reason = dma_frcd_fault_reason(data);
2012 			type = dma_frcd_type(data);
2013 
2014 			pasid = dma_frcd_pasid_value(data);
2015 			data = readl(iommu->reg + reg +
2016 				     fault_index * PRIMARY_FAULT_REG_LEN + 8);
2017 			source_id = dma_frcd_source_id(data);
2018 
2019 			pasid_present = dma_frcd_pasid_present(data);
2020 			guest_addr = dmar_readq(iommu->reg + reg +
2021 					fault_index * PRIMARY_FAULT_REG_LEN);
2022 			guest_addr = dma_frcd_page_addr(guest_addr);
2023 		}
2024 
2025 		/* clear the fault */
2026 		writel(DMA_FRCD_F, iommu->reg + reg +
2027 			fault_index * PRIMARY_FAULT_REG_LEN + 12);
2028 
2029 		raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
2030 
2031 		if (!ratelimited)
2032 			/* Using pasid -1 if pasid is not present */
2033 			dmar_fault_do_one(iommu, type, fault_reason,
2034 					  pasid_present ? pasid : INVALID_IOASID,
2035 					  source_id, guest_addr);
2036 
2037 		fault_index++;
2038 		if (fault_index >= cap_num_fault_regs(iommu->cap))
2039 			fault_index = 0;
2040 		raw_spin_lock_irqsave(&iommu->register_lock, flag);
2041 	}
2042 
2043 	writel(DMA_FSTS_PFO | DMA_FSTS_PPF | DMA_FSTS_PRO,
2044 	       iommu->reg + DMAR_FSTS_REG);
2045 
2046 unlock_exit:
2047 	raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
2048 	return IRQ_HANDLED;
2049 }
2050 
2051 int dmar_set_interrupt(struct intel_iommu *iommu)
2052 {
2053 	int irq, ret;
2054 
2055 	/*
2056 	 * Check if the fault interrupt is already initialized.
2057 	 */
2058 	if (iommu->irq)
2059 		return 0;
2060 
2061 	irq = dmar_alloc_hwirq(iommu->seq_id, iommu->node, iommu);
2062 	if (irq > 0) {
2063 		iommu->irq = irq;
2064 	} else {
2065 		pr_err("No free IRQ vectors\n");
2066 		return -EINVAL;
2067 	}
2068 
2069 	ret = request_irq(irq, dmar_fault, IRQF_NO_THREAD, iommu->name, iommu);
2070 	if (ret)
2071 		pr_err("Can't request irq\n");
2072 	return ret;
2073 }
2074 
2075 int __init enable_drhd_fault_handling(void)
2076 {
2077 	struct dmar_drhd_unit *drhd;
2078 	struct intel_iommu *iommu;
2079 
2080 	/*
2081 	 * Enable fault control interrupt.
2082 	 */
2083 	for_each_iommu(iommu, drhd) {
2084 		u32 fault_status;
2085 		int ret = dmar_set_interrupt(iommu);
2086 
2087 		if (ret) {
2088 			pr_err("DRHD %Lx: failed to enable fault, interrupt, ret %d\n",
2089 			       (unsigned long long)drhd->reg_base_addr, ret);
2090 			return -1;
2091 		}
2092 
2093 		/*
2094 		 * Clear any previous faults.
2095 		 */
2096 		dmar_fault(iommu->irq, iommu);
2097 		fault_status = readl(iommu->reg + DMAR_FSTS_REG);
2098 		writel(fault_status, iommu->reg + DMAR_FSTS_REG);
2099 	}
2100 
2101 	return 0;
2102 }
2103 
2104 /*
2105  * Re-enable Queued Invalidation interface.
2106  */
2107 int dmar_reenable_qi(struct intel_iommu *iommu)
2108 {
2109 	if (!ecap_qis(iommu->ecap))
2110 		return -ENOENT;
2111 
2112 	if (!iommu->qi)
2113 		return -ENOENT;
2114 
2115 	/*
2116 	 * First disable queued invalidation.
2117 	 */
2118 	dmar_disable_qi(iommu);
2119 	/*
2120 	 * Then enable queued invalidation again. Since there is no pending
2121 	 * invalidation requests now, it's safe to re-enable queued
2122 	 * invalidation.
2123 	 */
2124 	__dmar_enable_qi(iommu);
2125 
2126 	return 0;
2127 }
2128 
2129 /*
2130  * Check interrupt remapping support in DMAR table description.
2131  */
2132 int __init dmar_ir_support(void)
2133 {
2134 	struct acpi_table_dmar *dmar;
2135 	dmar = (struct acpi_table_dmar *)dmar_tbl;
2136 	if (!dmar)
2137 		return 0;
2138 	return dmar->flags & 0x1;
2139 }
2140 
2141 /* Check whether DMAR units are in use */
2142 static inline bool dmar_in_use(void)
2143 {
2144 	return irq_remapping_enabled || intel_iommu_enabled;
2145 }
2146 
2147 static int __init dmar_free_unused_resources(void)
2148 {
2149 	struct dmar_drhd_unit *dmaru, *dmaru_n;
2150 
2151 	if (dmar_in_use())
2152 		return 0;
2153 
2154 	if (dmar_dev_scope_status != 1 && !list_empty(&dmar_drhd_units))
2155 		bus_unregister_notifier(&pci_bus_type, &dmar_pci_bus_nb);
2156 
2157 	down_write(&dmar_global_lock);
2158 	list_for_each_entry_safe(dmaru, dmaru_n, &dmar_drhd_units, list) {
2159 		list_del(&dmaru->list);
2160 		dmar_free_drhd(dmaru);
2161 	}
2162 	up_write(&dmar_global_lock);
2163 
2164 	return 0;
2165 }
2166 
2167 late_initcall(dmar_free_unused_resources);
2168 IOMMU_INIT_POST(detect_intel_iommu);
2169 
2170 /*
2171  * DMAR Hotplug Support
2172  * For more details, please refer to Intel(R) Virtualization Technology
2173  * for Directed-IO Architecture Specifiction, Rev 2.2, Section 8.8
2174  * "Remapping Hardware Unit Hot Plug".
2175  */
2176 static guid_t dmar_hp_guid =
2177 	GUID_INIT(0xD8C1A3A6, 0xBE9B, 0x4C9B,
2178 		  0x91, 0xBF, 0xC3, 0xCB, 0x81, 0xFC, 0x5D, 0xAF);
2179 
2180 /*
2181  * Currently there's only one revision and BIOS will not check the revision id,
2182  * so use 0 for safety.
2183  */
2184 #define	DMAR_DSM_REV_ID			0
2185 #define	DMAR_DSM_FUNC_DRHD		1
2186 #define	DMAR_DSM_FUNC_ATSR		2
2187 #define	DMAR_DSM_FUNC_RHSA		3
2188 #define	DMAR_DSM_FUNC_SATC		4
2189 
2190 static inline bool dmar_detect_dsm(acpi_handle handle, int func)
2191 {
2192 	return acpi_check_dsm(handle, &dmar_hp_guid, DMAR_DSM_REV_ID, 1 << func);
2193 }
2194 
2195 static int dmar_walk_dsm_resource(acpi_handle handle, int func,
2196 				  dmar_res_handler_t handler, void *arg)
2197 {
2198 	int ret = -ENODEV;
2199 	union acpi_object *obj;
2200 	struct acpi_dmar_header *start;
2201 	struct dmar_res_callback callback;
2202 	static int res_type[] = {
2203 		[DMAR_DSM_FUNC_DRHD] = ACPI_DMAR_TYPE_HARDWARE_UNIT,
2204 		[DMAR_DSM_FUNC_ATSR] = ACPI_DMAR_TYPE_ROOT_ATS,
2205 		[DMAR_DSM_FUNC_RHSA] = ACPI_DMAR_TYPE_HARDWARE_AFFINITY,
2206 		[DMAR_DSM_FUNC_SATC] = ACPI_DMAR_TYPE_SATC,
2207 	};
2208 
2209 	if (!dmar_detect_dsm(handle, func))
2210 		return 0;
2211 
2212 	obj = acpi_evaluate_dsm_typed(handle, &dmar_hp_guid, DMAR_DSM_REV_ID,
2213 				      func, NULL, ACPI_TYPE_BUFFER);
2214 	if (!obj)
2215 		return -ENODEV;
2216 
2217 	memset(&callback, 0, sizeof(callback));
2218 	callback.cb[res_type[func]] = handler;
2219 	callback.arg[res_type[func]] = arg;
2220 	start = (struct acpi_dmar_header *)obj->buffer.pointer;
2221 	ret = dmar_walk_remapping_entries(start, obj->buffer.length, &callback);
2222 
2223 	ACPI_FREE(obj);
2224 
2225 	return ret;
2226 }
2227 
2228 static int dmar_hp_add_drhd(struct acpi_dmar_header *header, void *arg)
2229 {
2230 	int ret;
2231 	struct dmar_drhd_unit *dmaru;
2232 
2233 	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
2234 	if (!dmaru)
2235 		return -ENODEV;
2236 
2237 	ret = dmar_ir_hotplug(dmaru, true);
2238 	if (ret == 0)
2239 		ret = dmar_iommu_hotplug(dmaru, true);
2240 
2241 	return ret;
2242 }
2243 
2244 static int dmar_hp_remove_drhd(struct acpi_dmar_header *header, void *arg)
2245 {
2246 	int i, ret;
2247 	struct device *dev;
2248 	struct dmar_drhd_unit *dmaru;
2249 
2250 	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
2251 	if (!dmaru)
2252 		return 0;
2253 
2254 	/*
2255 	 * All PCI devices managed by this unit should have been destroyed.
2256 	 */
2257 	if (!dmaru->include_all && dmaru->devices && dmaru->devices_cnt) {
2258 		for_each_active_dev_scope(dmaru->devices,
2259 					  dmaru->devices_cnt, i, dev)
2260 			return -EBUSY;
2261 	}
2262 
2263 	ret = dmar_ir_hotplug(dmaru, false);
2264 	if (ret == 0)
2265 		ret = dmar_iommu_hotplug(dmaru, false);
2266 
2267 	return ret;
2268 }
2269 
2270 static int dmar_hp_release_drhd(struct acpi_dmar_header *header, void *arg)
2271 {
2272 	struct dmar_drhd_unit *dmaru;
2273 
2274 	dmaru = dmar_find_dmaru((struct acpi_dmar_hardware_unit *)header);
2275 	if (dmaru) {
2276 		list_del_rcu(&dmaru->list);
2277 		synchronize_rcu();
2278 		dmar_free_drhd(dmaru);
2279 	}
2280 
2281 	return 0;
2282 }
2283 
2284 static int dmar_hotplug_insert(acpi_handle handle)
2285 {
2286 	int ret;
2287 	int drhd_count = 0;
2288 
2289 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2290 				     &dmar_validate_one_drhd, (void *)1);
2291 	if (ret)
2292 		goto out;
2293 
2294 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2295 				     &dmar_parse_one_drhd, (void *)&drhd_count);
2296 	if (ret == 0 && drhd_count == 0) {
2297 		pr_warn(FW_BUG "No DRHD structures in buffer returned by _DSM method\n");
2298 		goto out;
2299 	} else if (ret) {
2300 		goto release_drhd;
2301 	}
2302 
2303 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_RHSA,
2304 				     &dmar_parse_one_rhsa, NULL);
2305 	if (ret)
2306 		goto release_drhd;
2307 
2308 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2309 				     &dmar_parse_one_atsr, NULL);
2310 	if (ret)
2311 		goto release_atsr;
2312 
2313 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2314 				     &dmar_hp_add_drhd, NULL);
2315 	if (!ret)
2316 		return 0;
2317 
2318 	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2319 			       &dmar_hp_remove_drhd, NULL);
2320 release_atsr:
2321 	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2322 			       &dmar_release_one_atsr, NULL);
2323 release_drhd:
2324 	dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2325 			       &dmar_hp_release_drhd, NULL);
2326 out:
2327 	return ret;
2328 }
2329 
2330 static int dmar_hotplug_remove(acpi_handle handle)
2331 {
2332 	int ret;
2333 
2334 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2335 				     &dmar_check_one_atsr, NULL);
2336 	if (ret)
2337 		return ret;
2338 
2339 	ret = dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2340 				     &dmar_hp_remove_drhd, NULL);
2341 	if (ret == 0) {
2342 		WARN_ON(dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_ATSR,
2343 					       &dmar_release_one_atsr, NULL));
2344 		WARN_ON(dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2345 					       &dmar_hp_release_drhd, NULL));
2346 	} else {
2347 		dmar_walk_dsm_resource(handle, DMAR_DSM_FUNC_DRHD,
2348 				       &dmar_hp_add_drhd, NULL);
2349 	}
2350 
2351 	return ret;
2352 }
2353 
2354 static acpi_status dmar_get_dsm_handle(acpi_handle handle, u32 lvl,
2355 				       void *context, void **retval)
2356 {
2357 	acpi_handle *phdl = retval;
2358 
2359 	if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) {
2360 		*phdl = handle;
2361 		return AE_CTRL_TERMINATE;
2362 	}
2363 
2364 	return AE_OK;
2365 }
2366 
2367 static int dmar_device_hotplug(acpi_handle handle, bool insert)
2368 {
2369 	int ret;
2370 	acpi_handle tmp = NULL;
2371 	acpi_status status;
2372 
2373 	if (!dmar_in_use())
2374 		return 0;
2375 
2376 	if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) {
2377 		tmp = handle;
2378 	} else {
2379 		status = acpi_walk_namespace(ACPI_TYPE_DEVICE, handle,
2380 					     ACPI_UINT32_MAX,
2381 					     dmar_get_dsm_handle,
2382 					     NULL, NULL, &tmp);
2383 		if (ACPI_FAILURE(status)) {
2384 			pr_warn("Failed to locate _DSM method.\n");
2385 			return -ENXIO;
2386 		}
2387 	}
2388 	if (tmp == NULL)
2389 		return 0;
2390 
2391 	down_write(&dmar_global_lock);
2392 	if (insert)
2393 		ret = dmar_hotplug_insert(tmp);
2394 	else
2395 		ret = dmar_hotplug_remove(tmp);
2396 	up_write(&dmar_global_lock);
2397 
2398 	return ret;
2399 }
2400 
2401 int dmar_device_add(acpi_handle handle)
2402 {
2403 	return dmar_device_hotplug(handle, true);
2404 }
2405 
2406 int dmar_device_remove(acpi_handle handle)
2407 {
2408 	return dmar_device_hotplug(handle, false);
2409 }
2410 
2411 /*
2412  * dmar_platform_optin - Is %DMA_CTRL_PLATFORM_OPT_IN_FLAG set in DMAR table
2413  *
2414  * Returns true if the platform has %DMA_CTRL_PLATFORM_OPT_IN_FLAG set in
2415  * the ACPI DMAR table. This means that the platform boot firmware has made
2416  * sure no device can issue DMA outside of RMRR regions.
2417  */
2418 bool dmar_platform_optin(void)
2419 {
2420 	struct acpi_table_dmar *dmar;
2421 	acpi_status status;
2422 	bool ret;
2423 
2424 	status = acpi_get_table(ACPI_SIG_DMAR, 0,
2425 				(struct acpi_table_header **)&dmar);
2426 	if (ACPI_FAILURE(status))
2427 		return false;
2428 
2429 	ret = !!(dmar->flags & DMAR_PLATFORM_OPT_IN);
2430 	acpi_put_table((struct acpi_table_header *)dmar);
2431 
2432 	return ret;
2433 }
2434 EXPORT_SYMBOL_GPL(dmar_platform_optin);
2435