Blame view
drivers/pci/intel-iommu.c
71.3 KB
ba3959276
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 |
/* * Copyright (c) 2006, Intel Corporation. * * This program is free software; you can redistribute it and/or modify it * under the terms and conditions of the GNU General Public License, * version 2, as published by the Free Software Foundation. * * This program is distributed in the hope it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for * more details. * * You should have received a copy of the GNU General Public License along with * this program; if not, write to the Free Software Foundation, Inc., 59 Temple * Place - Suite 330, Boston, MA 02111-1307 USA. * |
98bcef56c
|
17 18 19 20 |
* Copyright (C) 2006-2008 Intel Corporation * Author: Ashok Raj <ashok.raj@intel.com> * Author: Shaohua Li <shaohua.li@intel.com> * Author: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com> |
5b6985ce8
|
21 |
* Author: Fenghua Yu <fenghua.yu@intel.com> |
ba3959276
|
22 23 24 25 |
*/ #include <linux/init.h> #include <linux/bitmap.h> |
5e0d2a6fc
|
26 |
#include <linux/debugfs.h> |
ba3959276
|
27 28 29 |
#include <linux/slab.h> #include <linux/irq.h> #include <linux/interrupt.h> |
ba3959276
|
30 31 32 33 34 |
#include <linux/spinlock.h> #include <linux/pci.h> #include <linux/dmar.h> #include <linux/dma-mapping.h> #include <linux/mempool.h> |
5e0d2a6fc
|
35 |
#include <linux/timer.h> |
387179464
|
36 |
#include <linux/iova.h> |
5d450806e
|
37 |
#include <linux/iommu.h> |
387179464
|
38 |
#include <linux/intel-iommu.h> |
ba3959276
|
39 |
#include <asm/cacheflush.h> |
46a7fa270
|
40 |
#include <asm/iommu.h> |
ba3959276
|
41 |
#include "pci.h" |
5b6985ce8
|
42 43 |
#define ROOT_SIZE VTD_PAGE_SIZE #define CONTEXT_SIZE VTD_PAGE_SIZE |
ba3959276
|
44 45 46 47 48 49 50 51 |
#define IS_GFX_DEVICE(pdev) ((pdev->class >> 16) == PCI_BASE_CLASS_DISPLAY) #define IS_ISA_DEVICE(pdev) ((pdev->class >> 8) == PCI_CLASS_BRIDGE_ISA) #define IOAPIC_RANGE_START (0xfee00000) #define IOAPIC_RANGE_END (0xfeefffff) #define IOVA_START_ADDR (0x1000) #define DEFAULT_DOMAIN_ADDRESS_WIDTH 48 |
ba3959276
|
52 |
#define DOMAIN_MAX_ADDR(gaw) ((((u64)1) << gaw) - 1) |
f27be03b2
|
53 54 55 |
#define IOVA_PFN(addr) ((addr) >> PAGE_SHIFT) #define DMA_32BIT_PFN IOVA_PFN(DMA_32BIT_MASK) #define DMA_64BIT_PFN IOVA_PFN(DMA_64BIT_MASK) |
5e0d2a6fc
|
56 |
|
d9630fe94
|
57 58 |
/* global iommu list, set NULL for ignored DMAR units */ static struct intel_iommu **g_iommus; |
9af88143b
|
59 |
static int rwbf_quirk; |
46b08e1a7
|
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 |
/* * 0: Present * 1-11: Reserved * 12-63: Context Ptr (12 - (haw-1)) * 64-127: Reserved */ struct root_entry { u64 val; u64 rsvd1; }; #define ROOT_ENTRY_NR (VTD_PAGE_SIZE/sizeof(struct root_entry)) static inline bool root_present(struct root_entry *root) { return (root->val & 1); } static inline void set_root_present(struct root_entry *root) { root->val |= 1; } static inline void set_root_value(struct root_entry *root, unsigned long value) { root->val |= value & VTD_PAGE_MASK; } static inline struct context_entry * get_context_addr_from_root(struct root_entry *root) { return (struct context_entry *) (root_present(root)?phys_to_virt( root->val & VTD_PAGE_MASK) : NULL); } |
7a8fc25e0
|
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 |
/* * low 64 bits: * 0: present * 1: fault processing disable * 2-3: translation type * 12-63: address space root * high 64 bits: * 0-2: address width * 3-6: aval * 8-23: domain id */ struct context_entry { u64 lo; u64 hi; }; |
c07e7d217
|
107 108 109 110 111 112 113 114 115 116 117 118 119 120 |
static inline bool context_present(struct context_entry *context) { return (context->lo & 1); } static inline void context_set_present(struct context_entry *context) { context->lo |= 1; } static inline void context_set_fault_enable(struct context_entry *context) { context->lo &= (((u64)-1) << 2) | 1; } |
7a8fc25e0
|
121 |
#define CONTEXT_TT_MULTI_LEVEL 0 |
c07e7d217
|
122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 |
static inline void context_set_translation_type(struct context_entry *context, unsigned long value) { context->lo &= (((u64)-1) << 4) | 3; context->lo |= (value & 3) << 2; } static inline void context_set_address_root(struct context_entry *context, unsigned long value) { context->lo |= value & VTD_PAGE_MASK; } static inline void context_set_address_width(struct context_entry *context, unsigned long value) { context->hi |= value & 7; } static inline void context_set_domain_id(struct context_entry *context, unsigned long value) { context->hi |= (value & ((1 << 16) - 1)) << 8; } static inline void context_clear_entry(struct context_entry *context) { context->lo = 0; context->hi = 0; } |
7a8fc25e0
|
153 |
|
622ba12a4
|
154 155 156 157 158 159 160 161 162 163 164 |
/* * 0: readable * 1: writable * 2-6: reserved * 7: super page * 8-11: available * 12-63: Host physcial address */ struct dma_pte { u64 val; }; |
622ba12a4
|
165 |
|
19c239ce3
|
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 |
static inline void dma_clear_pte(struct dma_pte *pte) { pte->val = 0; } static inline void dma_set_pte_readable(struct dma_pte *pte) { pte->val |= DMA_PTE_READ; } static inline void dma_set_pte_writable(struct dma_pte *pte) { pte->val |= DMA_PTE_WRITE; } static inline void dma_set_pte_prot(struct dma_pte *pte, unsigned long prot) { pte->val = (pte->val & ~3) | (prot & 3); } static inline u64 dma_pte_addr(struct dma_pte *pte) { return (pte->val & VTD_PAGE_MASK); } static inline void dma_set_pte_addr(struct dma_pte *pte, u64 addr) { pte->val |= (addr & VTD_PAGE_MASK); } static inline bool dma_pte_present(struct dma_pte *pte) { return (pte->val & 3) != 0; } |
622ba12a4
|
200 |
|
3b5410e73
|
201 |
/* devices under the same p2p bridge are owned in one domain */ |
cdc7b8372
|
202 |
#define DOMAIN_FLAG_P2P_MULTIPLE_DEVICES (1 << 0) |
3b5410e73
|
203 |
|
1ce28feb2
|
204 205 206 207 |
/* domain represents a virtual machine, more than one devices * across iommus may be owned in one domain, e.g. kvm guest. */ #define DOMAIN_FLAG_VIRTUAL_MACHINE (1 << 1) |
99126f7ce
|
208 209 |
struct dmar_domain { int id; /* domain id */ |
8c11e798e
|
210 |
unsigned long iommu_bmp; /* bitmap of iommus this domain uses*/ |
99126f7ce
|
211 212 213 214 215 216 217 218 219 220 |
struct list_head devices; /* all devices' list */ struct iova_domain iovad; /* iova's that belong to this domain */ struct dma_pte *pgd; /* virtual address */ spinlock_t mapping_lock; /* page table lock */ int gaw; /* max guest address width */ /* adjusted guest address width, 0 is level 2 30-bit */ int agaw; |
3b5410e73
|
221 |
int flags; /* flags to find out type of domain */ |
8e604097d
|
222 223 |
int iommu_coherency;/* indicate coherency of iommu access */ |
c7151a8df
|
224 225 |
int iommu_count; /* reference count of iommu */ spinlock_t iommu_lock; /* protect iommu set in domain */ |
fe40f1e02
|
226 |
u64 max_addr; /* maximum mapped address */ |
99126f7ce
|
227 |
}; |
a647dacbb
|
228 229 230 231 232 233 234 235 236 |
/* PCI domain-device relationship */ struct device_domain_info { struct list_head link; /* link to domain siblings */ struct list_head global; /* link to global list */ u8 bus; /* PCI bus numer */ u8 devfn; /* PCI devfn number */ struct pci_dev *dev; /* it's NULL for PCIE-to-PCI bridge */ struct dmar_domain *domain; /* pointer to domain */ }; |
5e0d2a6fc
|
237 238 239 |
static void flush_unmaps_timeout(unsigned long data); DEFINE_TIMER(unmap_timer, flush_unmaps_timeout, 0, 0); |
80b20dd85
|
240 241 242 243 244 245 246 247 |
#define HIGH_WATER_MARK 250 struct deferred_flush_tables { int next; struct iova *iova[HIGH_WATER_MARK]; struct dmar_domain *domain[HIGH_WATER_MARK]; }; static struct deferred_flush_tables *deferred_flush; |
5e0d2a6fc
|
248 |
/* bitmap for indexing intel_iommus */ |
5e0d2a6fc
|
249 250 251 252 253 254 255 |
static int g_num_of_iommus; static DEFINE_SPINLOCK(async_umap_flush_lock); static LIST_HEAD(unmaps_to_do); static int timer_on; static long list_size; |
5e0d2a6fc
|
256 |
|
ba3959276
|
257 |
static void domain_remove_dev_info(struct dmar_domain *domain); |
0cd5c3c80
|
258 259 260 261 262 |
#ifdef CONFIG_DMAR_DEFAULT_ON int dmar_disabled = 0; #else int dmar_disabled = 1; #endif /*CONFIG_DMAR_DEFAULT_ON*/ |
ba3959276
|
263 |
static int __initdata dmar_map_gfx = 1; |
7d3b03ce7
|
264 |
static int dmar_forcedac; |
5e0d2a6fc
|
265 |
static int intel_iommu_strict; |
ba3959276
|
266 267 268 269 |
#define DUMMY_DEVICE_DOMAIN_INFO ((struct device_domain_info *)(-1)) static DEFINE_SPINLOCK(device_domain_lock); static LIST_HEAD(device_domain_list); |
a8bcbb0de
|
270 |
static struct iommu_ops intel_iommu_ops; |
ba3959276
|
271 272 273 274 275 |
static int __init intel_iommu_setup(char *str) { if (!str) return -EINVAL; while (*str) { |
0cd5c3c80
|
276 277 278 279 280 |
if (!strncmp(str, "on", 2)) { dmar_disabled = 0; printk(KERN_INFO "Intel-IOMMU: enabled "); } else if (!strncmp(str, "off", 3)) { |
ba3959276
|
281 |
dmar_disabled = 1; |
0cd5c3c80
|
282 283 |
printk(KERN_INFO "Intel-IOMMU: disabled "); |
ba3959276
|
284 285 286 287 288 |
} else if (!strncmp(str, "igfx_off", 8)) { dmar_map_gfx = 0; printk(KERN_INFO "Intel-IOMMU: disable GFX device mapping "); |
7d3b03ce7
|
289 |
} else if (!strncmp(str, "forcedac", 8)) { |
5e0d2a6fc
|
290 |
printk(KERN_INFO |
7d3b03ce7
|
291 292 293 |
"Intel-IOMMU: Forcing DAC for PCI devices "); dmar_forcedac = 1; |
5e0d2a6fc
|
294 295 296 297 298 |
} else if (!strncmp(str, "strict", 6)) { printk(KERN_INFO "Intel-IOMMU: disable batched IOTLB flush "); intel_iommu_strict = 1; |
ba3959276
|
299 300 301 302 303 304 305 306 307 308 309 310 311 |
} str += strcspn(str, ","); while (*str == ',') str++; } return 0; } __setup("intel_iommu=", intel_iommu_setup); static struct kmem_cache *iommu_domain_cache; static struct kmem_cache *iommu_devinfo_cache; static struct kmem_cache *iommu_iova_cache; |
eb3fa7cb5
|
312 313 314 315 316 317 318 319 320 321 322 323 |
static inline void *iommu_kmem_cache_alloc(struct kmem_cache *cachep) { unsigned int flags; void *vaddr; /* trying to avoid low memory issues */ flags = current->flags & PF_MEMALLOC; current->flags |= PF_MEMALLOC; vaddr = kmem_cache_alloc(cachep, GFP_ATOMIC); current->flags &= (~PF_MEMALLOC | flags); return vaddr; } |
ba3959276
|
324 325 |
static inline void *alloc_pgtable_page(void) { |
eb3fa7cb5
|
326 327 328 329 330 331 332 333 334 |
unsigned int flags; void *vaddr; /* trying to avoid low memory issues */ flags = current->flags & PF_MEMALLOC; current->flags |= PF_MEMALLOC; vaddr = (void *)get_zeroed_page(GFP_ATOMIC); current->flags &= (~PF_MEMALLOC | flags); return vaddr; |
ba3959276
|
335 336 337 338 339 340 341 342 343 |
} static inline void free_pgtable_page(void *vaddr) { free_page((unsigned long)vaddr); } static inline void *alloc_domain_mem(void) { |
eb3fa7cb5
|
344 |
return iommu_kmem_cache_alloc(iommu_domain_cache); |
ba3959276
|
345 |
} |
387179464
|
346 |
static void free_domain_mem(void *vaddr) |
ba3959276
|
347 348 349 350 351 352 |
{ kmem_cache_free(iommu_domain_cache, vaddr); } static inline void * alloc_devinfo_mem(void) { |
eb3fa7cb5
|
353 |
return iommu_kmem_cache_alloc(iommu_devinfo_cache); |
ba3959276
|
354 355 356 357 358 359 360 361 362 |
} static inline void free_devinfo_mem(void *vaddr) { kmem_cache_free(iommu_devinfo_cache, vaddr); } struct iova *alloc_iova_mem(void) { |
eb3fa7cb5
|
363 |
return iommu_kmem_cache_alloc(iommu_iova_cache); |
ba3959276
|
364 365 366 367 368 369 |
} void free_iova_mem(struct iova *iova) { kmem_cache_free(iommu_iova_cache, iova); } |
1b5736839
|
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 |
static inline int width_to_agaw(int width); /* calculate agaw for each iommu. * "SAGAW" may be different across iommus, use a default agaw, and * get a supported less agaw for iommus that don't support the default agaw. */ int iommu_calculate_agaw(struct intel_iommu *iommu) { unsigned long sagaw; int agaw = -1; sagaw = cap_sagaw(iommu->cap); for (agaw = width_to_agaw(DEFAULT_DOMAIN_ADDRESS_WIDTH); agaw >= 0; agaw--) { if (test_bit(agaw, &sagaw)) break; } return agaw; } |
8c11e798e
|
391 392 393 394 |
/* in native case, each domain is related to only one iommu */ static struct intel_iommu *domain_get_iommu(struct dmar_domain *domain) { int iommu_id; |
1ce28feb2
|
395 |
BUG_ON(domain->flags & DOMAIN_FLAG_VIRTUAL_MACHINE); |
8c11e798e
|
396 397 398 399 400 401 |
iommu_id = find_first_bit(&domain->iommu_bmp, g_num_of_iommus); if (iommu_id < 0 || iommu_id >= g_num_of_iommus) return NULL; return g_iommus[iommu_id]; } |
8e604097d
|
402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 |
/* "Coherency" capability may be different across iommus */ static void domain_update_iommu_coherency(struct dmar_domain *domain) { int i; domain->iommu_coherency = 1; i = find_first_bit(&domain->iommu_bmp, g_num_of_iommus); for (; i < g_num_of_iommus; ) { if (!ecap_coherent(g_iommus[i]->ecap)) { domain->iommu_coherency = 0; break; } i = find_next_bit(&domain->iommu_bmp, g_num_of_iommus, i+1); } } |
c7151a8df
|
418 419 420 421 422 423 424 425 426 427 |
static struct intel_iommu *device_to_iommu(u8 bus, u8 devfn) { struct dmar_drhd_unit *drhd = NULL; int i; for_each_drhd_unit(drhd) { if (drhd->ignored) continue; for (i = 0; i < drhd->devices_cnt; i++) |
288e4877f
|
428 429 |
if (drhd->devices[i] && drhd->devices[i]->bus->number == bus && |
c7151a8df
|
430 431 432 433 434 435 436 437 438 |
drhd->devices[i]->devfn == devfn) return drhd->iommu; if (drhd->include_all) return drhd->iommu; } return NULL; } |
5331fe6f5
|
439 440 441 442 443 444 |
static void domain_flush_cache(struct dmar_domain *domain, void *addr, int size) { if (!domain->iommu_coherency) clflush_cache_range(addr, size); } |
ba3959276
|
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 |
/* Gets context entry for a given bus and devfn */ static struct context_entry * device_to_context_entry(struct intel_iommu *iommu, u8 bus, u8 devfn) { struct root_entry *root; struct context_entry *context; unsigned long phy_addr; unsigned long flags; spin_lock_irqsave(&iommu->lock, flags); root = &iommu->root_entry[bus]; context = get_context_addr_from_root(root); if (!context) { context = (struct context_entry *)alloc_pgtable_page(); if (!context) { spin_unlock_irqrestore(&iommu->lock, flags); return NULL; } |
5b6985ce8
|
463 |
__iommu_flush_cache(iommu, (void *)context, CONTEXT_SIZE); |
ba3959276
|
464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 |
phy_addr = virt_to_phys((void *)context); set_root_value(root, phy_addr); set_root_present(root); __iommu_flush_cache(iommu, root, sizeof(*root)); } spin_unlock_irqrestore(&iommu->lock, flags); return &context[devfn]; } static int device_context_mapped(struct intel_iommu *iommu, u8 bus, u8 devfn) { struct root_entry *root; struct context_entry *context; int ret; unsigned long flags; spin_lock_irqsave(&iommu->lock, flags); root = &iommu->root_entry[bus]; context = get_context_addr_from_root(root); if (!context) { ret = 0; goto out; } |
c07e7d217
|
487 |
ret = context_present(&context[devfn]); |
ba3959276
|
488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 |
out: spin_unlock_irqrestore(&iommu->lock, flags); return ret; } static void clear_context_table(struct intel_iommu *iommu, u8 bus, u8 devfn) { struct root_entry *root; struct context_entry *context; unsigned long flags; spin_lock_irqsave(&iommu->lock, flags); root = &iommu->root_entry[bus]; context = get_context_addr_from_root(root); if (context) { |
c07e7d217
|
503 |
context_clear_entry(&context[devfn]); |
ba3959276
|
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 |
__iommu_flush_cache(iommu, &context[devfn], \ sizeof(*context)); } spin_unlock_irqrestore(&iommu->lock, flags); } static void free_context_table(struct intel_iommu *iommu) { struct root_entry *root; int i; unsigned long flags; struct context_entry *context; spin_lock_irqsave(&iommu->lock, flags); if (!iommu->root_entry) { goto out; } for (i = 0; i < ROOT_ENTRY_NR; i++) { root = &iommu->root_entry[i]; context = get_context_addr_from_root(root); if (context) free_pgtable_page(context); } free_pgtable_page(iommu->root_entry); iommu->root_entry = NULL; out: spin_unlock_irqrestore(&iommu->lock, flags); } /* page table handling */ #define LEVEL_STRIDE (9) #define LEVEL_MASK (((u64)1 << LEVEL_STRIDE) - 1) static inline int agaw_to_level(int agaw) { return agaw + 2; } static inline int agaw_to_width(int agaw) { return 30 + agaw * LEVEL_STRIDE; } static inline int width_to_agaw(int width) { return (width - 30) / LEVEL_STRIDE; } static inline unsigned int level_to_offset_bits(int level) { return (12 + (level - 1) * LEVEL_STRIDE); } static inline int address_level_offset(u64 addr, int level) { return ((addr >> level_to_offset_bits(level)) & LEVEL_MASK); } static inline u64 level_mask(int level) { return ((u64)-1 << level_to_offset_bits(level)); } static inline u64 level_size(int level) { return ((u64)1 << level_to_offset_bits(level)); } static inline u64 align_to_level(u64 addr, int level) { return ((addr + level_size(level) - 1) & level_mask(level)); } static struct dma_pte * addr_to_dma_pte(struct dmar_domain *domain, u64 addr) { int addr_width = agaw_to_width(domain->agaw); struct dma_pte *parent, *pte = NULL; int level = agaw_to_level(domain->agaw); int offset; unsigned long flags; BUG_ON(!domain->pgd); addr &= (((u64)1) << addr_width) - 1; parent = domain->pgd; spin_lock_irqsave(&domain->mapping_lock, flags); while (level > 0) { void *tmp_page; offset = address_level_offset(addr, level); pte = &parent[offset]; if (level == 1) break; |
19c239ce3
|
599 |
if (!dma_pte_present(pte)) { |
ba3959276
|
600 601 602 603 604 605 606 |
tmp_page = alloc_pgtable_page(); if (!tmp_page) { spin_unlock_irqrestore(&domain->mapping_lock, flags); return NULL; } |
5331fe6f5
|
607 |
domain_flush_cache(domain, tmp_page, PAGE_SIZE); |
19c239ce3
|
608 |
dma_set_pte_addr(pte, virt_to_phys(tmp_page)); |
ba3959276
|
609 610 611 612 |
/* * high level table always sets r/w, last level page * table control read/write */ |
19c239ce3
|
613 614 |
dma_set_pte_readable(pte); dma_set_pte_writable(pte); |
5331fe6f5
|
615 |
domain_flush_cache(domain, pte, sizeof(*pte)); |
ba3959276
|
616 |
} |
19c239ce3
|
617 |
parent = phys_to_virt(dma_pte_addr(pte)); |
ba3959276
|
618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 |
level--; } spin_unlock_irqrestore(&domain->mapping_lock, flags); return pte; } /* return address's pte at specific level */ static struct dma_pte *dma_addr_level_pte(struct dmar_domain *domain, u64 addr, int level) { struct dma_pte *parent, *pte = NULL; int total = agaw_to_level(domain->agaw); int offset; parent = domain->pgd; while (level <= total) { offset = address_level_offset(addr, total); pte = &parent[offset]; if (level == total) return pte; |
19c239ce3
|
639 |
if (!dma_pte_present(pte)) |
ba3959276
|
640 |
break; |
19c239ce3
|
641 |
parent = phys_to_virt(dma_pte_addr(pte)); |
ba3959276
|
642 643 644 645 646 647 648 649 650 651 652 653 654 655 |
total--; } return NULL; } /* clear one page's page table */ static void dma_pte_clear_one(struct dmar_domain *domain, u64 addr) { struct dma_pte *pte = NULL; /* get last level pte */ pte = dma_addr_level_pte(domain, addr, 1); if (pte) { |
19c239ce3
|
656 |
dma_clear_pte(pte); |
5331fe6f5
|
657 |
domain_flush_cache(domain, pte, sizeof(*pte)); |
ba3959276
|
658 659 660 661 662 663 664 665 666 667 668 |
} } /* clear last level pte, a tlb flush should be followed */ static void dma_pte_clear_range(struct dmar_domain *domain, u64 start, u64 end) { int addr_width = agaw_to_width(domain->agaw); start &= (((u64)1) << addr_width) - 1; end &= (((u64)1) << addr_width) - 1; /* in case it's partial page */ |
5b6985ce8
|
669 670 |
start = PAGE_ALIGN(start); end &= PAGE_MASK; |
ba3959276
|
671 672 673 674 |
/* we don't need lock here, nobody else touches the iova range */ while (start < end) { dma_pte_clear_one(domain, start); |
5b6985ce8
|
675 |
start += VTD_PAGE_SIZE; |
ba3959276
|
676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 |
} } /* free page table pages. last level pte should already be cleared */ static void dma_pte_free_pagetable(struct dmar_domain *domain, u64 start, u64 end) { int addr_width = agaw_to_width(domain->agaw); struct dma_pte *pte; int total = agaw_to_level(domain->agaw); int level; u64 tmp; start &= (((u64)1) << addr_width) - 1; end &= (((u64)1) << addr_width) - 1; /* we don't need lock here, nobody else touches the iova range */ level = 2; while (level <= total) { tmp = align_to_level(start, level); if (tmp >= end || (tmp + level_size(level) > end)) return; while (tmp < end) { pte = dma_addr_level_pte(domain, tmp, level); if (pte) { free_pgtable_page( |
19c239ce3
|
703 704 |
phys_to_virt(dma_pte_addr(pte))); dma_clear_pte(pte); |
5331fe6f5
|
705 |
domain_flush_cache(domain, pte, sizeof(*pte)); |
ba3959276
|
706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 |
} tmp += level_size(level); } level++; } /* free pgd */ if (start == 0 && end >= ((((u64)1) << addr_width) - 1)) { free_pgtable_page(domain->pgd); domain->pgd = NULL; } } /* iommu handling */ static int iommu_alloc_root_entry(struct intel_iommu *iommu) { struct root_entry *root; unsigned long flags; root = (struct root_entry *)alloc_pgtable_page(); if (!root) return -ENOMEM; |
5b6985ce8
|
727 |
__iommu_flush_cache(iommu, root, ROOT_SIZE); |
ba3959276
|
728 729 730 731 732 733 734 |
spin_lock_irqsave(&iommu->lock, flags); iommu->root_entry = root; spin_unlock_irqrestore(&iommu->lock, flags); return 0; } |
ba3959276
|
735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 |
static void iommu_set_root_entry(struct intel_iommu *iommu) { void *addr; u32 cmd, sts; unsigned long flag; addr = iommu->root_entry; spin_lock_irqsave(&iommu->register_lock, flag); dmar_writeq(iommu->reg + DMAR_RTADDR_REG, virt_to_phys(addr)); cmd = iommu->gcmd | DMA_GCMD_SRTP; writel(cmd, iommu->reg + DMAR_GCMD_REG); /* Make sure hardware complete it */ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (sts & DMA_GSTS_RTPS), sts); spin_unlock_irqrestore(&iommu->register_lock, flag); } static void iommu_flush_write_buffer(struct intel_iommu *iommu) { u32 val; unsigned long flag; |
9af88143b
|
760 |
if (!rwbf_quirk && !cap_rwbf(iommu->cap)) |
ba3959276
|
761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 |
return; val = iommu->gcmd | DMA_GCMD_WBF; spin_lock_irqsave(&iommu->register_lock, flag); writel(val, iommu->reg + DMAR_GCMD_REG); /* Make sure hardware complete it */ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (!(val & DMA_GSTS_WBFS)), val); spin_unlock_irqrestore(&iommu->register_lock, flag); } /* return value determine if we need a write buffer flush */ static int __iommu_flush_context(struct intel_iommu *iommu, u16 did, u16 source_id, u8 function_mask, u64 type, int non_present_entry_flush) { u64 val = 0; unsigned long flag; /* * In the non-present entry flush case, if hardware doesn't cache * non-present entry we do nothing and if hardware cache non-present * entry, we flush entries of domain 0 (the domain id is used to cache * any non-present entries) */ if (non_present_entry_flush) { if (!cap_caching_mode(iommu->cap)) return 1; else did = 0; } switch (type) { case DMA_CCMD_GLOBAL_INVL: val = DMA_CCMD_GLOBAL_INVL; break; case DMA_CCMD_DOMAIN_INVL: val = DMA_CCMD_DOMAIN_INVL|DMA_CCMD_DID(did); break; case DMA_CCMD_DEVICE_INVL: val = DMA_CCMD_DEVICE_INVL|DMA_CCMD_DID(did) | DMA_CCMD_SID(source_id) | DMA_CCMD_FM(function_mask); break; default: BUG(); } val |= DMA_CCMD_ICC; spin_lock_irqsave(&iommu->register_lock, flag); dmar_writeq(iommu->reg + DMAR_CCMD_REG, val); /* Make sure hardware complete it */ IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, (!(val & DMA_CCMD_ICC)), val); spin_unlock_irqrestore(&iommu->register_lock, flag); |
4d235ba6c
|
819 |
/* flush context entry will implicitly flush write buffer */ |
ba3959276
|
820 821 |
return 0; } |
ba3959276
|
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 |
/* return value determine if we need a write buffer flush */ static int __iommu_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr, unsigned int size_order, u64 type, int non_present_entry_flush) { int tlb_offset = ecap_iotlb_offset(iommu->ecap); u64 val = 0, val_iva = 0; unsigned long flag; /* * In the non-present entry flush case, if hardware doesn't cache * non-present entry we do nothing and if hardware cache non-present * entry, we flush entries of domain 0 (the domain id is used to cache * any non-present entries) */ if (non_present_entry_flush) { if (!cap_caching_mode(iommu->cap)) return 1; else did = 0; } switch (type) { case DMA_TLB_GLOBAL_FLUSH: /* global flush doesn't need set IVA_REG */ val = DMA_TLB_GLOBAL_FLUSH|DMA_TLB_IVT; break; case DMA_TLB_DSI_FLUSH: val = DMA_TLB_DSI_FLUSH|DMA_TLB_IVT|DMA_TLB_DID(did); break; case DMA_TLB_PSI_FLUSH: val = DMA_TLB_PSI_FLUSH|DMA_TLB_IVT|DMA_TLB_DID(did); /* Note: always flush non-leaf currently */ val_iva = size_order | addr; break; default: BUG(); } /* Note: set drain read/write */ #if 0 /* * This is probably to be super secure.. Looks like we can * ignore it without any impact. */ if (cap_read_drain(iommu->cap)) val |= DMA_TLB_READ_DRAIN; #endif if (cap_write_drain(iommu->cap)) val |= DMA_TLB_WRITE_DRAIN; spin_lock_irqsave(&iommu->register_lock, flag); /* Note: Only uses first TLB reg currently */ if (val_iva) dmar_writeq(iommu->reg + tlb_offset, val_iva); dmar_writeq(iommu->reg + tlb_offset + 8, val); /* Make sure hardware complete it */ IOMMU_WAIT_OP(iommu, tlb_offset + 8, dmar_readq, (!(val & DMA_TLB_IVT)), val); spin_unlock_irqrestore(&iommu->register_lock, flag); /* check IOTLB invalidation granularity */ if (DMA_TLB_IAIG(val) == 0) printk(KERN_ERR"IOMMU: flush IOTLB failed "); if (DMA_TLB_IAIG(val) != DMA_TLB_IIRG(type)) pr_debug("IOMMU: tlb flush request %Lx, actual %Lx ", |
5b6985ce8
|
891 892 |
(unsigned long long)DMA_TLB_IIRG(type), (unsigned long long)DMA_TLB_IAIG(val)); |
4d235ba6c
|
893 |
/* flush iotlb entry will implicitly flush write buffer */ |
ba3959276
|
894 895 |
return 0; } |
ba3959276
|
896 897 898 |
static int iommu_flush_iotlb_psi(struct intel_iommu *iommu, u16 did, u64 addr, unsigned int pages, int non_present_entry_flush) { |
f76aec76e
|
899 |
unsigned int mask; |
ba3959276
|
900 |
|
5b6985ce8
|
901 |
BUG_ON(addr & (~VTD_PAGE_MASK)); |
ba3959276
|
902 903 904 905 |
BUG_ON(pages == 0); /* Fallback to domain selective flush if no PSI support */ if (!cap_pgsel_inv(iommu->cap)) |
a77b67d40
|
906 907 908 |
return iommu->flush.flush_iotlb(iommu, did, 0, 0, DMA_TLB_DSI_FLUSH, non_present_entry_flush); |
ba3959276
|
909 910 911 912 913 |
/* * PSI requires page size to be 2 ^ x, and the base address is naturally * aligned to the size */ |
f76aec76e
|
914 |
mask = ilog2(__roundup_pow_of_two(pages)); |
ba3959276
|
915 |
/* Fallback to domain selective flush if size is too big */ |
f76aec76e
|
916 |
if (mask > cap_max_amask_val(iommu->cap)) |
a77b67d40
|
917 918 |
return iommu->flush.flush_iotlb(iommu, did, 0, 0, DMA_TLB_DSI_FLUSH, non_present_entry_flush); |
ba3959276
|
919 |
|
a77b67d40
|
920 921 922 |
return iommu->flush.flush_iotlb(iommu, did, addr, mask, DMA_TLB_PSI_FLUSH, non_present_entry_flush); |
ba3959276
|
923 |
} |
f8bab7351
|
924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 |
static void iommu_disable_protect_mem_regions(struct intel_iommu *iommu) { u32 pmen; unsigned long flags; spin_lock_irqsave(&iommu->register_lock, flags); pmen = readl(iommu->reg + DMAR_PMEN_REG); pmen &= ~DMA_PMEN_EPM; writel(pmen, iommu->reg + DMAR_PMEN_REG); /* wait for the protected region status bit to clear */ IOMMU_WAIT_OP(iommu, DMAR_PMEN_REG, readl, !(pmen & DMA_PMEN_PRS), pmen); spin_unlock_irqrestore(&iommu->register_lock, flags); } |
ba3959276
|
940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 |
static int iommu_enable_translation(struct intel_iommu *iommu) { u32 sts; unsigned long flags; spin_lock_irqsave(&iommu->register_lock, flags); writel(iommu->gcmd|DMA_GCMD_TE, iommu->reg + DMAR_GCMD_REG); /* Make sure hardware complete it */ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (sts & DMA_GSTS_TES), sts); iommu->gcmd |= DMA_GCMD_TE; spin_unlock_irqrestore(&iommu->register_lock, flags); return 0; } static int iommu_disable_translation(struct intel_iommu *iommu) { u32 sts; unsigned long flag; spin_lock_irqsave(&iommu->register_lock, flag); iommu->gcmd &= ~DMA_GCMD_TE; writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG); /* Make sure hardware complete it */ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (!(sts & DMA_GSTS_TES)), sts); spin_unlock_irqrestore(&iommu->register_lock, flag); return 0; } |
3460a6d9c
|
973 |
|
ba3959276
|
974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 |
static int iommu_init_domains(struct intel_iommu *iommu) { unsigned long ndomains; unsigned long nlongs; ndomains = cap_ndoms(iommu->cap); pr_debug("Number of Domains supportd <%ld> ", ndomains); nlongs = BITS_TO_LONGS(ndomains); /* TBD: there might be 64K domains, * consider other allocation for future chip */ iommu->domain_ids = kcalloc(nlongs, sizeof(unsigned long), GFP_KERNEL); if (!iommu->domain_ids) { printk(KERN_ERR "Allocating domain id array failed "); return -ENOMEM; } iommu->domains = kcalloc(ndomains, sizeof(struct dmar_domain *), GFP_KERNEL); if (!iommu->domains) { printk(KERN_ERR "Allocating domain array failed "); kfree(iommu->domain_ids); return -ENOMEM; } |
e61d98d8d
|
1001 |
spin_lock_init(&iommu->lock); |
ba3959276
|
1002 1003 1004 1005 1006 1007 1008 1009 |
/* * if Caching mode is set, then invalid translations are tagged * with domainid 0. Hence we need to pre-allocate it. */ if (cap_caching_mode(iommu->cap)) set_bit(0, iommu->domain_ids); return 0; } |
ba3959276
|
1010 |
|
ba3959276
|
1011 1012 |
static void domain_exit(struct dmar_domain *domain); |
5e98c4b1d
|
1013 |
static void vm_domain_exit(struct dmar_domain *domain); |
e61d98d8d
|
1014 1015 |
void free_dmar_iommu(struct intel_iommu *iommu) |
ba3959276
|
1016 1017 1018 |
{ struct dmar_domain *domain; int i; |
c7151a8df
|
1019 |
unsigned long flags; |
ba3959276
|
1020 |
|
ba3959276
|
1021 1022 1023 1024 |
i = find_first_bit(iommu->domain_ids, cap_ndoms(iommu->cap)); for (; i < cap_ndoms(iommu->cap); ) { domain = iommu->domains[i]; clear_bit(i, iommu->domain_ids); |
c7151a8df
|
1025 1026 |
spin_lock_irqsave(&domain->iommu_lock, flags); |
5e98c4b1d
|
1027 1028 1029 1030 1031 1032 |
if (--domain->iommu_count == 0) { if (domain->flags & DOMAIN_FLAG_VIRTUAL_MACHINE) vm_domain_exit(domain); else domain_exit(domain); } |
c7151a8df
|
1033 |
spin_unlock_irqrestore(&domain->iommu_lock, flags); |
ba3959276
|
1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 |
i = find_next_bit(iommu->domain_ids, cap_ndoms(iommu->cap), i+1); } if (iommu->gcmd & DMA_GCMD_TE) iommu_disable_translation(iommu); if (iommu->irq) { set_irq_data(iommu->irq, NULL); /* This will mask the irq */ free_irq(iommu->irq, iommu); destroy_irq(iommu->irq); } kfree(iommu->domains); kfree(iommu->domain_ids); |
d9630fe94
|
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 |
g_iommus[iommu->seq_id] = NULL; /* if all iommus are freed, free g_iommus */ for (i = 0; i < g_num_of_iommus; i++) { if (g_iommus[i]) break; } if (i == g_num_of_iommus) kfree(g_iommus); |
ba3959276
|
1060 1061 |
/* free context mapping */ free_context_table(iommu); |
ba3959276
|
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 |
} static struct dmar_domain * iommu_alloc_domain(struct intel_iommu *iommu) { unsigned long num; unsigned long ndomains; struct dmar_domain *domain; unsigned long flags; domain = alloc_domain_mem(); if (!domain) return NULL; ndomains = cap_ndoms(iommu->cap); spin_lock_irqsave(&iommu->lock, flags); num = find_first_zero_bit(iommu->domain_ids, ndomains); if (num >= ndomains) { spin_unlock_irqrestore(&iommu->lock, flags); free_domain_mem(domain); printk(KERN_ERR "IOMMU: no free domain ids "); return NULL; } set_bit(num, iommu->domain_ids); domain->id = num; |
8c11e798e
|
1089 1090 |
memset(&domain->iommu_bmp, 0, sizeof(unsigned long)); set_bit(iommu->seq_id, &domain->iommu_bmp); |
d71a2f33a
|
1091 |
domain->flags = 0; |
ba3959276
|
1092 1093 1094 1095 1096 1097 1098 1099 1100 |
iommu->domains[num] = domain; spin_unlock_irqrestore(&iommu->lock, flags); return domain; } static void iommu_free_domain(struct dmar_domain *domain) { unsigned long flags; |
8c11e798e
|
1101 1102 1103 |
struct intel_iommu *iommu; iommu = domain_get_iommu(domain); |
ba3959276
|
1104 |
|
8c11e798e
|
1105 1106 1107 |
spin_lock_irqsave(&iommu->lock, flags); clear_bit(domain->id, iommu->domain_ids); spin_unlock_irqrestore(&iommu->lock, flags); |
ba3959276
|
1108 1109 1110 |
} static struct iova_domain reserved_iova_list; |
8a443df40
|
1111 1112 |
static struct lock_class_key reserved_alloc_key; static struct lock_class_key reserved_rbtree_key; |
ba3959276
|
1113 1114 1115 1116 1117 1118 1119 |
static void dmar_init_reserved_ranges(void) { struct pci_dev *pdev = NULL; struct iova *iova; int i; u64 addr, size; |
f661197e0
|
1120 |
init_iova_domain(&reserved_iova_list, DMA_32BIT_PFN); |
ba3959276
|
1121 |
|
8a443df40
|
1122 1123 1124 1125 |
lockdep_set_class(&reserved_iova_list.iova_alloc_lock, &reserved_alloc_key); lockdep_set_class(&reserved_iova_list.iova_rbtree_lock, &reserved_rbtree_key); |
ba3959276
|
1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 |
/* IOAPIC ranges shouldn't be accessed by DMA */ iova = reserve_iova(&reserved_iova_list, IOVA_PFN(IOAPIC_RANGE_START), IOVA_PFN(IOAPIC_RANGE_END)); if (!iova) printk(KERN_ERR "Reserve IOAPIC range failed "); /* Reserve all PCI MMIO to avoid peer-to-peer access */ for_each_pci_dev(pdev) { struct resource *r; for (i = 0; i < PCI_NUM_RESOURCES; i++) { r = &pdev->resource[i]; if (!r->flags || !(r->flags & IORESOURCE_MEM)) continue; addr = r->start; |
5b6985ce8
|
1142 |
addr &= PAGE_MASK; |
ba3959276
|
1143 |
size = r->end - addr; |
5b6985ce8
|
1144 |
size = PAGE_ALIGN(size); |
ba3959276
|
1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 |
iova = reserve_iova(&reserved_iova_list, IOVA_PFN(addr), IOVA_PFN(size + addr) - 1); if (!iova) printk(KERN_ERR "Reserve iova failed "); } } } static void domain_reserve_special_ranges(struct dmar_domain *domain) { copy_reserved_iova(&reserved_iova_list, &domain->iovad); } static inline int guestwidth_to_adjustwidth(int gaw) { int agaw; int r = (gaw - 12) % 9; if (r == 0) agaw = gaw; else agaw = gaw + 9 - r; if (agaw > 64) agaw = 64; return agaw; } static int domain_init(struct dmar_domain *domain, int guest_width) { struct intel_iommu *iommu; int adjust_width, agaw; unsigned long sagaw; |
f661197e0
|
1179 |
init_iova_domain(&domain->iovad, DMA_32BIT_PFN); |
ba3959276
|
1180 |
spin_lock_init(&domain->mapping_lock); |
c7151a8df
|
1181 |
spin_lock_init(&domain->iommu_lock); |
ba3959276
|
1182 1183 1184 1185 |
domain_reserve_special_ranges(domain); /* calculate AGAW */ |
8c11e798e
|
1186 |
iommu = domain_get_iommu(domain); |
ba3959276
|
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 |
if (guest_width > cap_mgaw(iommu->cap)) guest_width = cap_mgaw(iommu->cap); domain->gaw = guest_width; adjust_width = guestwidth_to_adjustwidth(guest_width); agaw = width_to_agaw(adjust_width); sagaw = cap_sagaw(iommu->cap); if (!test_bit(agaw, &sagaw)) { /* hardware doesn't support it, choose a bigger one */ pr_debug("IOMMU: hardware doesn't support agaw %d ", agaw); agaw = find_next_bit(&sagaw, 5, agaw); if (agaw >= 5) return -ENODEV; } domain->agaw = agaw; INIT_LIST_HEAD(&domain->devices); |
8e604097d
|
1203 1204 1205 1206 |
if (ecap_coherent(iommu->ecap)) domain->iommu_coherency = 1; else domain->iommu_coherency = 0; |
c7151a8df
|
1207 |
domain->iommu_count = 1; |
ba3959276
|
1208 1209 1210 1211 |
/* always allocate the top pgd */ domain->pgd = (struct dma_pte *)alloc_pgtable_page(); if (!domain->pgd) return -ENOMEM; |
5b6985ce8
|
1212 |
__iommu_flush_cache(iommu, domain->pgd, PAGE_SIZE); |
ba3959276
|
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 |
return 0; } static void domain_exit(struct dmar_domain *domain) { u64 end; /* Domain 0 is reserved, so dont process it */ if (!domain) return; domain_remove_dev_info(domain); /* destroy iovas */ put_iova_domain(&domain->iovad); end = DOMAIN_MAX_ADDR(domain->gaw); |
5b6985ce8
|
1228 |
end = end & (~PAGE_MASK); |
ba3959276
|
1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 |
/* clear ptes */ dma_pte_clear_range(domain, 0, end); /* free page tables */ dma_pte_free_pagetable(domain, 0, end); iommu_free_domain(domain); free_domain_mem(domain); } static int domain_context_mapping_one(struct dmar_domain *domain, u8 bus, u8 devfn) { struct context_entry *context; |
ba3959276
|
1244 |
unsigned long flags; |
5331fe6f5
|
1245 |
struct intel_iommu *iommu; |
ea6606b02
|
1246 1247 1248 1249 1250 |
struct dma_pte *pgd; unsigned long num; unsigned long ndomains; int id; int agaw; |
ba3959276
|
1251 1252 1253 1254 1255 |
pr_debug("Set context mapping for %02x:%02x.%d ", bus, PCI_SLOT(devfn), PCI_FUNC(devfn)); BUG_ON(!domain->pgd); |
5331fe6f5
|
1256 1257 1258 1259 |
iommu = device_to_iommu(bus, devfn); if (!iommu) return -ENODEV; |
ba3959276
|
1260 1261 1262 1263 |
context = device_to_context_entry(iommu, bus, devfn); if (!context) return -ENOMEM; spin_lock_irqsave(&iommu->lock, flags); |
c07e7d217
|
1264 |
if (context_present(context)) { |
ba3959276
|
1265 1266 1267 |
spin_unlock_irqrestore(&iommu->lock, flags); return 0; } |
ea6606b02
|
1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 |
id = domain->id; pgd = domain->pgd; if (domain->flags & DOMAIN_FLAG_VIRTUAL_MACHINE) { int found = 0; /* find an available domain id for this device in iommu */ ndomains = cap_ndoms(iommu->cap); num = find_first_bit(iommu->domain_ids, ndomains); for (; num < ndomains; ) { if (iommu->domains[num] == domain) { id = num; found = 1; break; } num = find_next_bit(iommu->domain_ids, cap_ndoms(iommu->cap), num+1); } if (found == 0) { num = find_first_zero_bit(iommu->domain_ids, ndomains); if (num >= ndomains) { spin_unlock_irqrestore(&iommu->lock, flags); printk(KERN_ERR "IOMMU: no free domain ids "); return -EFAULT; } set_bit(num, iommu->domain_ids); iommu->domains[num] = domain; id = num; } /* Skip top levels of page tables for * iommu which has less agaw than default. */ for (agaw = domain->agaw; agaw != iommu->agaw; agaw--) { pgd = phys_to_virt(dma_pte_addr(pgd)); if (!dma_pte_present(pgd)) { spin_unlock_irqrestore(&iommu->lock, flags); return -ENOMEM; } } } context_set_domain_id(context, id); context_set_address_width(context, iommu->agaw); context_set_address_root(context, virt_to_phys(pgd)); |
c07e7d217
|
1316 1317 1318 |
context_set_translation_type(context, CONTEXT_TT_MULTI_LEVEL); context_set_fault_enable(context); context_set_present(context); |
5331fe6f5
|
1319 |
domain_flush_cache(domain, context, sizeof(*context)); |
ba3959276
|
1320 1321 |
/* it's a non-present to present mapping */ |
a77b67d40
|
1322 1323 1324 |
if (iommu->flush.flush_context(iommu, domain->id, (((u16)bus) << 8) | devfn, DMA_CCMD_MASK_NOBIT, DMA_CCMD_DEVICE_INVL, 1)) |
ba3959276
|
1325 1326 |
iommu_flush_write_buffer(iommu); else |
a77b67d40
|
1327 |
iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_DSI_FLUSH, 0); |
ba3959276
|
1328 |
spin_unlock_irqrestore(&iommu->lock, flags); |
c7151a8df
|
1329 1330 1331 1332 1333 1334 1335 |
spin_lock_irqsave(&domain->iommu_lock, flags); if (!test_and_set_bit(iommu->seq_id, &domain->iommu_bmp)) { domain->iommu_count++; domain_update_iommu_coherency(domain); } spin_unlock_irqrestore(&domain->iommu_lock, flags); |
ba3959276
|
1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 |
return 0; } static int domain_context_mapping(struct dmar_domain *domain, struct pci_dev *pdev) { int ret; struct pci_dev *tmp, *parent; ret = domain_context_mapping_one(domain, pdev->bus->number, pdev->devfn); if (ret) return ret; /* dependent device mapping */ tmp = pci_find_upstream_pcie_bridge(pdev); if (!tmp) return 0; /* Secondary interface's bus number and devfn 0 */ parent = pdev->bus->self; while (parent != tmp) { ret = domain_context_mapping_one(domain, parent->bus->number, parent->devfn); if (ret) return ret; parent = parent->bus->self; } if (tmp->is_pcie) /* this is a PCIE-to-PCI bridge */ return domain_context_mapping_one(domain, tmp->subordinate->number, 0); else /* this is a legacy PCI bridge */ return domain_context_mapping_one(domain, tmp->bus->number, tmp->devfn); } |
5331fe6f5
|
1370 |
static int domain_context_mapped(struct pci_dev *pdev) |
ba3959276
|
1371 1372 1373 |
{ int ret; struct pci_dev *tmp, *parent; |
5331fe6f5
|
1374 1375 1376 1377 1378 |
struct intel_iommu *iommu; iommu = device_to_iommu(pdev->bus->number, pdev->devfn); if (!iommu) return -ENODEV; |
ba3959276
|
1379 |
|
8c11e798e
|
1380 |
ret = device_context_mapped(iommu, |
ba3959276
|
1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 |
pdev->bus->number, pdev->devfn); if (!ret) return ret; /* dependent device mapping */ tmp = pci_find_upstream_pcie_bridge(pdev); if (!tmp) return ret; /* Secondary interface's bus number and devfn 0 */ parent = pdev->bus->self; while (parent != tmp) { |
8c11e798e
|
1391 |
ret = device_context_mapped(iommu, parent->bus->number, |
ba3959276
|
1392 1393 1394 1395 1396 1397 |
parent->devfn); if (!ret) return ret; parent = parent->bus->self; } if (tmp->is_pcie) |
8c11e798e
|
1398 |
return device_context_mapped(iommu, |
ba3959276
|
1399 1400 |
tmp->subordinate->number, 0); else |
8c11e798e
|
1401 |
return device_context_mapped(iommu, |
ba3959276
|
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 |
tmp->bus->number, tmp->devfn); } static int domain_page_mapping(struct dmar_domain *domain, dma_addr_t iova, u64 hpa, size_t size, int prot) { u64 start_pfn, end_pfn; struct dma_pte *pte; int index; |
5b6985ce8
|
1412 1413 1414 |
int addr_width = agaw_to_width(domain->agaw); hpa &= (((u64)1) << addr_width) - 1; |
ba3959276
|
1415 1416 1417 |
if ((prot & (DMA_PTE_READ|DMA_PTE_WRITE)) == 0) return -EINVAL; |
5b6985ce8
|
1418 1419 1420 |
iova &= PAGE_MASK; start_pfn = ((u64)hpa) >> VTD_PAGE_SHIFT; end_pfn = (VTD_PAGE_ALIGN(((u64)hpa) + size)) >> VTD_PAGE_SHIFT; |
ba3959276
|
1421 1422 |
index = 0; while (start_pfn < end_pfn) { |
5b6985ce8
|
1423 |
pte = addr_to_dma_pte(domain, iova + VTD_PAGE_SIZE * index); |
ba3959276
|
1424 1425 1426 1427 1428 |
if (!pte) return -ENOMEM; /* We don't need lock here, nobody else * touches the iova range */ |
19c239ce3
|
1429 1430 1431 |
BUG_ON(dma_pte_addr(pte)); dma_set_pte_addr(pte, start_pfn << VTD_PAGE_SHIFT); dma_set_pte_prot(pte, prot); |
5331fe6f5
|
1432 |
domain_flush_cache(domain, pte, sizeof(*pte)); |
ba3959276
|
1433 1434 1435 1436 1437 |
start_pfn++; index++; } return 0; } |
c7151a8df
|
1438 |
static void iommu_detach_dev(struct intel_iommu *iommu, u8 bus, u8 devfn) |
ba3959276
|
1439 |
{ |
c7151a8df
|
1440 1441 |
if (!iommu) return; |
8c11e798e
|
1442 1443 1444 |
clear_context_table(iommu, bus, devfn); iommu->flush.flush_context(iommu, 0, 0, 0, |
a77b67d40
|
1445 |
DMA_CCMD_GLOBAL_INVL, 0); |
8c11e798e
|
1446 |
iommu->flush.flush_iotlb(iommu, 0, 0, 0, |
a77b67d40
|
1447 |
DMA_TLB_GLOBAL_FLUSH, 0); |
ba3959276
|
1448 1449 1450 1451 1452 1453 |
} static void domain_remove_dev_info(struct dmar_domain *domain) { struct device_domain_info *info; unsigned long flags; |
c7151a8df
|
1454 |
struct intel_iommu *iommu; |
ba3959276
|
1455 1456 1457 1458 1459 1460 1461 1462 |
spin_lock_irqsave(&device_domain_lock, flags); while (!list_empty(&domain->devices)) { info = list_entry(domain->devices.next, struct device_domain_info, link); list_del(&info->link); list_del(&info->global); if (info->dev) |
358dd8ac5
|
1463 |
info->dev->dev.archdata.iommu = NULL; |
ba3959276
|
1464 |
spin_unlock_irqrestore(&device_domain_lock, flags); |
c7151a8df
|
1465 1466 |
iommu = device_to_iommu(info->bus, info->devfn); iommu_detach_dev(iommu, info->bus, info->devfn); |
ba3959276
|
1467 1468 1469 1470 1471 1472 1473 1474 1475 |
free_devinfo_mem(info); spin_lock_irqsave(&device_domain_lock, flags); } spin_unlock_irqrestore(&device_domain_lock, flags); } /* * find_domain |
358dd8ac5
|
1476 |
* Note: we use struct pci_dev->dev.archdata.iommu stores the info |
ba3959276
|
1477 |
*/ |
387179464
|
1478 |
static struct dmar_domain * |
ba3959276
|
1479 1480 1481 1482 1483 |
find_domain(struct pci_dev *pdev) { struct device_domain_info *info; /* No lock here, assumes no domain exit in normal case */ |
358dd8ac5
|
1484 |
info = pdev->dev.archdata.iommu; |
ba3959276
|
1485 1486 1487 1488 |
if (info) return info->domain; return NULL; } |
ba3959276
|
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 |
/* domain is initialized */ static struct dmar_domain *get_domain_for_dev(struct pci_dev *pdev, int gaw) { struct dmar_domain *domain, *found = NULL; struct intel_iommu *iommu; struct dmar_drhd_unit *drhd; struct device_domain_info *info, *tmp; struct pci_dev *dev_tmp; unsigned long flags; int bus = 0, devfn = 0; domain = find_domain(pdev); if (domain) return domain; dev_tmp = pci_find_upstream_pcie_bridge(pdev); if (dev_tmp) { if (dev_tmp->is_pcie) { bus = dev_tmp->subordinate->number; devfn = 0; } else { bus = dev_tmp->bus->number; devfn = dev_tmp->devfn; } spin_lock_irqsave(&device_domain_lock, flags); list_for_each_entry(info, &device_domain_list, global) { if (info->bus == bus && info->devfn == devfn) { found = info->domain; break; } } spin_unlock_irqrestore(&device_domain_lock, flags); /* pcie-pci bridge already has a domain, uses it */ if (found) { domain = found; goto found_domain; } } /* Allocate new domain for the device */ drhd = dmar_find_matched_drhd_unit(pdev); if (!drhd) { printk(KERN_ERR "IOMMU: can't find DMAR for device %s ", pci_name(pdev)); return NULL; } iommu = drhd->iommu; domain = iommu_alloc_domain(iommu); if (!domain) goto error; if (domain_init(domain, gaw)) { domain_exit(domain); goto error; } /* register pcie-to-pci device */ if (dev_tmp) { info = alloc_devinfo_mem(); if (!info) { domain_exit(domain); goto error; } info->bus = bus; info->devfn = devfn; info->dev = NULL; info->domain = domain; /* This domain is shared by devices under p2p bridge */ |
3b5410e73
|
1559 |
domain->flags |= DOMAIN_FLAG_P2P_MULTIPLE_DEVICES; |
ba3959276
|
1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 |
/* pcie-to-pci bridge already has a domain, uses it */ found = NULL; spin_lock_irqsave(&device_domain_lock, flags); list_for_each_entry(tmp, &device_domain_list, global) { if (tmp->bus == bus && tmp->devfn == devfn) { found = tmp->domain; break; } } if (found) { free_devinfo_mem(info); domain_exit(domain); domain = found; } else { list_add(&info->link, &domain->devices); list_add(&info->global, &device_domain_list); } spin_unlock_irqrestore(&device_domain_lock, flags); } found_domain: info = alloc_devinfo_mem(); if (!info) goto error; info->bus = pdev->bus->number; info->devfn = pdev->devfn; info->dev = pdev; info->domain = domain; spin_lock_irqsave(&device_domain_lock, flags); /* somebody is fast */ found = find_domain(pdev); if (found != NULL) { spin_unlock_irqrestore(&device_domain_lock, flags); if (found != domain) { domain_exit(domain); domain = found; } free_devinfo_mem(info); return domain; } list_add(&info->link, &domain->devices); list_add(&info->global, &device_domain_list); |
358dd8ac5
|
1603 |
pdev->dev.archdata.iommu = info; |
ba3959276
|
1604 1605 1606 1607 1608 1609 |
spin_unlock_irqrestore(&device_domain_lock, flags); return domain; error: /* recheck it here, maybe others set it */ return find_domain(pdev); } |
5b6985ce8
|
1610 1611 1612 |
static int iommu_prepare_identity_map(struct pci_dev *pdev, unsigned long long start, unsigned long long end) |
ba3959276
|
1613 1614 1615 |
{ struct dmar_domain *domain; unsigned long size; |
5b6985ce8
|
1616 |
unsigned long long base; |
ba3959276
|
1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 |
int ret; printk(KERN_INFO "IOMMU: Setting identity map for device %s [0x%Lx - 0x%Lx] ", pci_name(pdev), start, end); /* page table init */ domain = get_domain_for_dev(pdev, DEFAULT_DOMAIN_ADDRESS_WIDTH); if (!domain) return -ENOMEM; /* The address might not be aligned */ |
5b6985ce8
|
1629 |
base = start & PAGE_MASK; |
ba3959276
|
1630 |
size = end - base; |
5b6985ce8
|
1631 |
size = PAGE_ALIGN(size); |
ba3959276
|
1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 |
if (!reserve_iova(&domain->iovad, IOVA_PFN(base), IOVA_PFN(base + size) - 1)) { printk(KERN_ERR "IOMMU: reserve iova failed "); ret = -ENOMEM; goto error; } pr_debug("Mapping reserved region %lx@%llx for %s ", size, base, pci_name(pdev)); /* * RMRR range might have overlap with physical memory range, * clear it first */ dma_pte_clear_range(domain, base, base + size); ret = domain_page_mapping(domain, base, base, size, DMA_PTE_READ|DMA_PTE_WRITE); if (ret) goto error; /* context entry init */ ret = domain_context_mapping(domain, pdev); if (!ret) return 0; error: domain_exit(domain); return ret; } static inline int iommu_prepare_rmrr_dev(struct dmar_rmrr_unit *rmrr, struct pci_dev *pdev) { |
358dd8ac5
|
1667 |
if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO) |
ba3959276
|
1668 1669 1670 1671 |
return 0; return iommu_prepare_identity_map(pdev, rmrr->base_address, rmrr->end_address + 1); } |
e820482cd
|
1672 |
#ifdef CONFIG_DMAR_GFX_WA |
d52d53b8a
|
1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 |
struct iommu_prepare_data { struct pci_dev *pdev; int ret; }; static int __init iommu_prepare_work_fn(unsigned long start_pfn, unsigned long end_pfn, void *datax) { struct iommu_prepare_data *data; data = (struct iommu_prepare_data *)datax; data->ret = iommu_prepare_identity_map(data->pdev, start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT); return data->ret; } static int __init iommu_prepare_with_active_regions(struct pci_dev *pdev) { int nid; struct iommu_prepare_data data; data.pdev = pdev; data.ret = 0; for_each_online_node(nid) { work_with_active_regions(nid, iommu_prepare_work_fn, &data); if (data.ret) return data.ret; } return data.ret; } |
e820482cd
|
1706 1707 1708 |
static void __init iommu_prepare_gfx_mapping(void) { struct pci_dev *pdev = NULL; |
e820482cd
|
1709 1710 1711 |
int ret; for_each_pci_dev(pdev) { |
358dd8ac5
|
1712 |
if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO || |
e820482cd
|
1713 1714 1715 1716 1717 |
!IS_GFX_DEVICE(pdev)) continue; printk(KERN_INFO "IOMMU: gfx device %s 1-1 mapping ", pci_name(pdev)); |
d52d53b8a
|
1718 1719 1720 1721 |
ret = iommu_prepare_with_active_regions(pdev); if (ret) printk(KERN_ERR "IOMMU: mapping reserved region failed "); |
e820482cd
|
1722 1723 |
} } |
2abd7e167
|
1724 1725 1726 1727 1728 |
#else /* !CONFIG_DMAR_GFX_WA */ static inline void iommu_prepare_gfx_mapping(void) { return; } |
e820482cd
|
1729 |
#endif |
49a0429e5
|
1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 |
#ifdef CONFIG_DMAR_FLOPPY_WA static inline void iommu_prepare_isa(void) { struct pci_dev *pdev; int ret; pdev = pci_get_class(PCI_CLASS_BRIDGE_ISA << 8, NULL); if (!pdev) return; printk(KERN_INFO "IOMMU: Prepare 0-16M unity mapping for LPC "); ret = iommu_prepare_identity_map(pdev, 0, 16*1024*1024); if (ret) |
1c35b8e53
|
1745 |
printk(KERN_ERR "IOMMU: Failed to create 0-64M identity map, " |
49a0429e5
|
1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 |
"floppy might not work "); } #else static inline void iommu_prepare_isa(void) { return; } #endif /* !CONFIG_DMAR_FLPY_WA */ |
519a05491
|
1756 |
static int __init init_dmars(void) |
ba3959276
|
1757 1758 1759 1760 1761 |
{ struct dmar_drhd_unit *drhd; struct dmar_rmrr_unit *rmrr; struct pci_dev *pdev; struct intel_iommu *iommu; |
9d783ba04
|
1762 |
int i, ret; |
ba3959276
|
1763 1764 1765 1766 1767 1768 1769 1770 |
/* * for each drhd * allocate root * initialize and program root entry to not present * endfor */ for_each_drhd_unit(drhd) { |
5e0d2a6fc
|
1771 1772 1773 1774 1775 1776 1777 |
g_num_of_iommus++; /* * lock not needed as this is only incremented in the single * threaded kernel __init code path all other access are read * only */ } |
d9630fe94
|
1778 1779 1780 1781 1782 1783 1784 1785 |
g_iommus = kcalloc(g_num_of_iommus, sizeof(struct intel_iommu *), GFP_KERNEL); if (!g_iommus) { printk(KERN_ERR "Allocating global iommu array failed "); ret = -ENOMEM; goto error; } |
80b20dd85
|
1786 1787 1788 |
deferred_flush = kzalloc(g_num_of_iommus * sizeof(struct deferred_flush_tables), GFP_KERNEL); if (!deferred_flush) { |
d9630fe94
|
1789 |
kfree(g_iommus); |
5e0d2a6fc
|
1790 1791 1792 |
ret = -ENOMEM; goto error; } |
5e0d2a6fc
|
1793 1794 1795 |
for_each_drhd_unit(drhd) { if (drhd->ignored) continue; |
1886e8a90
|
1796 1797 |
iommu = drhd->iommu; |
d9630fe94
|
1798 |
g_iommus[iommu->seq_id] = iommu; |
ba3959276
|
1799 |
|
e61d98d8d
|
1800 1801 1802 |
ret = iommu_init_domains(iommu); if (ret) goto error; |
ba3959276
|
1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 |
/* * TBD: * we could share the same root & context tables * amoung all IOMMU's. Need to Split it later. */ ret = iommu_alloc_root_entry(iommu); if (ret) { printk(KERN_ERR "IOMMU: allocate root entry failed "); goto error; } } |
1531a6a6b
|
1815 1816 1817 |
/* * Start from the sane iommu hardware state. */ |
a77b67d40
|
1818 1819 1820 1821 1822 |
for_each_drhd_unit(drhd) { if (drhd->ignored) continue; iommu = drhd->iommu; |
1531a6a6b
|
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 |
/* * If the queued invalidation is already initialized by us * (for example, while enabling interrupt-remapping) then * we got the things already rolling from a sane state. */ if (iommu->qi) continue; /* * Clear any previous faults. */ dmar_fault(-1, iommu); /* * Disable queued invalidation if supported and already enabled * before OS handover. */ dmar_disable_qi(iommu); } for_each_drhd_unit(drhd) { if (drhd->ignored) continue; iommu = drhd->iommu; |
a77b67d40
|
1848 1849 1850 1851 1852 1853 1854 1855 |
if (dmar_enable_qi(iommu)) { /* * Queued Invalidate not enabled, use Register Based * Invalidate */ iommu->flush.flush_context = __iommu_flush_context; iommu->flush.flush_iotlb = __iommu_flush_iotlb; printk(KERN_INFO "IOMMU 0x%Lx: using Register based " |
b4e0f9eb8
|
1856 1857 1858 |
"invalidation ", (unsigned long long)drhd->reg_base_addr); |
a77b67d40
|
1859 1860 1861 1862 |
} else { iommu->flush.flush_context = qi_flush_context; iommu->flush.flush_iotlb = qi_flush_iotlb; printk(KERN_INFO "IOMMU 0x%Lx: using Queued " |
b4e0f9eb8
|
1863 1864 1865 |
"invalidation ", (unsigned long long)drhd->reg_base_addr); |
a77b67d40
|
1866 1867 |
} } |
ba3959276
|
1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 |
/* * For each rmrr * for each dev attached to rmrr * do * locate drhd for dev, alloc domain for dev * allocate free domain * allocate page table entries for rmrr * if context not allocated for bus * allocate and init context * set present in root table for this bus * init context with domain, translation etc * endfor * endfor */ for_each_rmrr_units(rmrr) { |
ba3959276
|
1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 |
for (i = 0; i < rmrr->devices_cnt; i++) { pdev = rmrr->devices[i]; /* some BIOS lists non-exist devices in DMAR table */ if (!pdev) continue; ret = iommu_prepare_rmrr_dev(rmrr, pdev); if (ret) printk(KERN_ERR "IOMMU: mapping reserved region failed "); } } |
e820482cd
|
1895 |
iommu_prepare_gfx_mapping(); |
49a0429e5
|
1896 |
iommu_prepare_isa(); |
ba3959276
|
1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 |
/* * for each drhd * enable fault log * global invalidate context cache * global invalidate iotlb * enable translation */ for_each_drhd_unit(drhd) { if (drhd->ignored) continue; iommu = drhd->iommu; |
ba3959276
|
1908 1909 |
iommu_flush_write_buffer(iommu); |
3460a6d9c
|
1910 1911 1912 |
ret = dmar_set_interrupt(iommu); if (ret) goto error; |
ba3959276
|
1913 |
iommu_set_root_entry(iommu); |
a77b67d40
|
1914 1915 1916 1917 |
iommu->flush.flush_context(iommu, 0, 0, 0, DMA_CCMD_GLOBAL_INVL, 0); iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH, 0); |
f8bab7351
|
1918 |
iommu_disable_protect_mem_regions(iommu); |
ba3959276
|
1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 |
ret = iommu_enable_translation(iommu); if (ret) goto error; } return 0; error: for_each_drhd_unit(drhd) { if (drhd->ignored) continue; iommu = drhd->iommu; free_iommu(iommu); } |
d9630fe94
|
1932 |
kfree(g_iommus); |
ba3959276
|
1933 1934 1935 1936 1937 1938 |
return ret; } static inline u64 aligned_size(u64 host_addr, size_t size) { u64 addr; |
5b6985ce8
|
1939 1940 |
addr = (host_addr & (~PAGE_MASK)) + size; return PAGE_ALIGN(addr); |
ba3959276
|
1941 1942 1943 |
} struct iova * |
f76aec76e
|
1944 |
iommu_alloc_iova(struct dmar_domain *domain, size_t size, u64 end) |
ba3959276
|
1945 |
{ |
ba3959276
|
1946 1947 1948 |
struct iova *piova; /* Make sure it's in range */ |
ba3959276
|
1949 |
end = min_t(u64, DOMAIN_MAX_ADDR(domain->gaw), end); |
f76aec76e
|
1950 |
if (!size || (IOVA_START_ADDR + size > end)) |
ba3959276
|
1951 1952 1953 |
return NULL; piova = alloc_iova(&domain->iovad, |
5b6985ce8
|
1954 |
size >> PAGE_SHIFT, IOVA_PFN(end), 1); |
ba3959276
|
1955 1956 |
return piova; } |
f76aec76e
|
1957 1958 |
static struct iova * __intel_alloc_iova(struct device *dev, struct dmar_domain *domain, |
bb9e6d650
|
1959 |
size_t size, u64 dma_mask) |
ba3959276
|
1960 |
{ |
ba3959276
|
1961 |
struct pci_dev *pdev = to_pci_dev(dev); |
ba3959276
|
1962 |
struct iova *iova = NULL; |
ba3959276
|
1963 |
|
bb9e6d650
|
1964 1965 1966 |
if (dma_mask <= DMA_32BIT_MASK || dmar_forcedac) iova = iommu_alloc_iova(domain, size, dma_mask); else { |
ba3959276
|
1967 1968 1969 |
/* * First try to allocate an io virtual address in * DMA_32BIT_MASK and if that fails then try allocating |
3609801ec
|
1970 |
* from higher range |
ba3959276
|
1971 |
*/ |
f76aec76e
|
1972 |
iova = iommu_alloc_iova(domain, size, DMA_32BIT_MASK); |
ba3959276
|
1973 |
if (!iova) |
bb9e6d650
|
1974 |
iova = iommu_alloc_iova(domain, size, dma_mask); |
ba3959276
|
1975 1976 1977 1978 |
} if (!iova) { printk(KERN_ERR"Allocating iova for %s failed", pci_name(pdev)); |
f76aec76e
|
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 |
return NULL; } return iova; } static struct dmar_domain * get_valid_domain_for_dev(struct pci_dev *pdev) { struct dmar_domain *domain; int ret; domain = get_domain_for_dev(pdev, DEFAULT_DOMAIN_ADDRESS_WIDTH); if (!domain) { printk(KERN_ERR "Allocating domain for %s failed", pci_name(pdev)); |
4fe05bbcd
|
1996 |
return NULL; |
ba3959276
|
1997 1998 1999 |
} /* make sure context mapping is ok */ |
5331fe6f5
|
2000 |
if (unlikely(!domain_context_mapped(pdev))) { |
ba3959276
|
2001 |
ret = domain_context_mapping(domain, pdev); |
f76aec76e
|
2002 2003 2004 2005 |
if (ret) { printk(KERN_ERR "Domain context map for %s failed", pci_name(pdev)); |
4fe05bbcd
|
2006 |
return NULL; |
f76aec76e
|
2007 |
} |
ba3959276
|
2008 |
} |
f76aec76e
|
2009 2010 |
return domain; } |
bb9e6d650
|
2011 2012 |
static dma_addr_t __intel_map_single(struct device *hwdev, phys_addr_t paddr, size_t size, int dir, u64 dma_mask) |
f76aec76e
|
2013 2014 |
{ struct pci_dev *pdev = to_pci_dev(hwdev); |
f76aec76e
|
2015 |
struct dmar_domain *domain; |
5b6985ce8
|
2016 |
phys_addr_t start_paddr; |
f76aec76e
|
2017 2018 |
struct iova *iova; int prot = 0; |
6865f0d19
|
2019 |
int ret; |
8c11e798e
|
2020 |
struct intel_iommu *iommu; |
f76aec76e
|
2021 2022 |
BUG_ON(dir == DMA_NONE); |
358dd8ac5
|
2023 |
if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO) |
6865f0d19
|
2024 |
return paddr; |
f76aec76e
|
2025 2026 2027 2028 |
domain = get_valid_domain_for_dev(pdev); if (!domain) return 0; |
8c11e798e
|
2029 |
iommu = domain_get_iommu(domain); |
6865f0d19
|
2030 |
size = aligned_size((u64)paddr, size); |
f76aec76e
|
2031 |
|
bb9e6d650
|
2032 |
iova = __intel_alloc_iova(hwdev, domain, size, pdev->dma_mask); |
f76aec76e
|
2033 2034 |
if (!iova) goto error; |
5b6985ce8
|
2035 |
start_paddr = (phys_addr_t)iova->pfn_lo << PAGE_SHIFT; |
f76aec76e
|
2036 |
|
ba3959276
|
2037 2038 2039 2040 2041 |
/* * Check if DMAR supports zero-length reads on write only * mappings.. */ if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL || \ |
8c11e798e
|
2042 |
!cap_zlr(iommu->cap)) |
ba3959276
|
2043 2044 2045 2046 |
prot |= DMA_PTE_READ; if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL) prot |= DMA_PTE_WRITE; /* |
6865f0d19
|
2047 |
* paddr - (paddr + size) might be partial page, we should map the whole |
ba3959276
|
2048 |
* page. Note: if two part of one page are separately mapped, we |
6865f0d19
|
2049 |
* might have two guest_addr mapping to the same host paddr, but this |
ba3959276
|
2050 2051 |
* is not a big problem */ |
6865f0d19
|
2052 |
ret = domain_page_mapping(domain, start_paddr, |
5b6985ce8
|
2053 |
((u64)paddr) & PAGE_MASK, size, prot); |
ba3959276
|
2054 2055 |
if (ret) goto error; |
f76aec76e
|
2056 |
/* it's a non-present to present mapping */ |
8c11e798e
|
2057 |
ret = iommu_flush_iotlb_psi(iommu, domain->id, |
5b6985ce8
|
2058 |
start_paddr, size >> VTD_PAGE_SHIFT, 1); |
f76aec76e
|
2059 |
if (ret) |
8c11e798e
|
2060 |
iommu_flush_write_buffer(iommu); |
f76aec76e
|
2061 |
|
5b6985ce8
|
2062 |
return start_paddr + ((u64)paddr & (~PAGE_MASK)); |
ba3959276
|
2063 |
|
ba3959276
|
2064 |
error: |
f76aec76e
|
2065 2066 |
if (iova) __free_iova(&domain->iovad, iova); |
ba3959276
|
2067 2068 |
printk(KERN_ERR"Device %s request: %lx@%llx dir %d --- failed ", |
5b6985ce8
|
2069 |
pci_name(pdev), size, (unsigned long long)paddr, dir); |
ba3959276
|
2070 2071 |
return 0; } |
ffbbef5c0
|
2072 2073 2074 2075 |
static dma_addr_t intel_map_page(struct device *dev, struct page *page, unsigned long offset, size_t size, enum dma_data_direction dir, struct dma_attrs *attrs) |
bb9e6d650
|
2076 |
{ |
ffbbef5c0
|
2077 2078 |
return __intel_map_single(dev, page_to_phys(page) + offset, size, dir, to_pci_dev(dev)->dma_mask); |
bb9e6d650
|
2079 |
} |
5e0d2a6fc
|
2080 2081 |
static void flush_unmaps(void) { |
80b20dd85
|
2082 |
int i, j; |
5e0d2a6fc
|
2083 |
|
5e0d2a6fc
|
2084 2085 2086 2087 |
timer_on = 0; /* just flush them all */ for (i = 0; i < g_num_of_iommus; i++) { |
a2bb8459f
|
2088 2089 2090 |
struct intel_iommu *iommu = g_iommus[i]; if (!iommu) continue; |
c42d9f324
|
2091 |
|
a2bb8459f
|
2092 |
if (deferred_flush[i].next) { |
a77b67d40
|
2093 2094 |
iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH, 0); |
80b20dd85
|
2095 2096 2097 2098 2099 2100 |
for (j = 0; j < deferred_flush[i].next; j++) { __free_iova(&deferred_flush[i].domain[j]->iovad, deferred_flush[i].iova[j]); } deferred_flush[i].next = 0; } |
5e0d2a6fc
|
2101 |
} |
5e0d2a6fc
|
2102 |
list_size = 0; |
5e0d2a6fc
|
2103 2104 2105 2106 |
} static void flush_unmaps_timeout(unsigned long data) { |
80b20dd85
|
2107 2108 2109 |
unsigned long flags; spin_lock_irqsave(&async_umap_flush_lock, flags); |
5e0d2a6fc
|
2110 |
flush_unmaps(); |
80b20dd85
|
2111 |
spin_unlock_irqrestore(&async_umap_flush_lock, flags); |
5e0d2a6fc
|
2112 2113 2114 2115 2116 |
} static void add_unmap(struct dmar_domain *dom, struct iova *iova) { unsigned long flags; |
80b20dd85
|
2117 |
int next, iommu_id; |
8c11e798e
|
2118 |
struct intel_iommu *iommu; |
5e0d2a6fc
|
2119 2120 |
spin_lock_irqsave(&async_umap_flush_lock, flags); |
80b20dd85
|
2121 2122 |
if (list_size == HIGH_WATER_MARK) flush_unmaps(); |
8c11e798e
|
2123 2124 |
iommu = domain_get_iommu(dom); iommu_id = iommu->seq_id; |
c42d9f324
|
2125 |
|
80b20dd85
|
2126 2127 2128 2129 |
next = deferred_flush[iommu_id].next; deferred_flush[iommu_id].domain[next] = dom; deferred_flush[iommu_id].iova[next] = iova; deferred_flush[iommu_id].next++; |
5e0d2a6fc
|
2130 2131 2132 2133 2134 2135 2136 2137 |
if (!timer_on) { mod_timer(&unmap_timer, jiffies + msecs_to_jiffies(10)); timer_on = 1; } list_size++; spin_unlock_irqrestore(&async_umap_flush_lock, flags); } |
ffbbef5c0
|
2138 2139 2140 |
static void intel_unmap_page(struct device *dev, dma_addr_t dev_addr, size_t size, enum dma_data_direction dir, struct dma_attrs *attrs) |
ba3959276
|
2141 |
{ |
ba3959276
|
2142 |
struct pci_dev *pdev = to_pci_dev(dev); |
f76aec76e
|
2143 2144 |
struct dmar_domain *domain; unsigned long start_addr; |
ba3959276
|
2145 |
struct iova *iova; |
8c11e798e
|
2146 |
struct intel_iommu *iommu; |
ba3959276
|
2147 |
|
358dd8ac5
|
2148 |
if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO) |
f76aec76e
|
2149 |
return; |
ba3959276
|
2150 2151 |
domain = find_domain(pdev); BUG_ON(!domain); |
8c11e798e
|
2152 |
iommu = domain_get_iommu(domain); |
ba3959276
|
2153 |
iova = find_iova(&domain->iovad, IOVA_PFN(dev_addr)); |
f76aec76e
|
2154 |
if (!iova) |
ba3959276
|
2155 |
return; |
ba3959276
|
2156 |
|
5b6985ce8
|
2157 |
start_addr = iova->pfn_lo << PAGE_SHIFT; |
f76aec76e
|
2158 |
size = aligned_size((u64)dev_addr, size); |
ba3959276
|
2159 |
|
f76aec76e
|
2160 2161 |
pr_debug("Device %s unmapping: %lx@%llx ", |
5b6985ce8
|
2162 |
pci_name(pdev), size, (unsigned long long)start_addr); |
ba3959276
|
2163 |
|
f76aec76e
|
2164 2165 2166 2167 |
/* clear the whole page */ dma_pte_clear_range(domain, start_addr, start_addr + size); /* free page tables */ dma_pte_free_pagetable(domain, start_addr, start_addr + size); |
5e0d2a6fc
|
2168 |
if (intel_iommu_strict) { |
8c11e798e
|
2169 |
if (iommu_flush_iotlb_psi(iommu, |
5b6985ce8
|
2170 |
domain->id, start_addr, size >> VTD_PAGE_SHIFT, 0)) |
8c11e798e
|
2171 |
iommu_flush_write_buffer(iommu); |
5e0d2a6fc
|
2172 2173 2174 2175 2176 2177 2178 2179 |
/* free iova */ __free_iova(&domain->iovad, iova); } else { add_unmap(domain, iova); /* * queue up the release of the unmap to save the 1/6th of the * cpu used up by the iotlb flush operation... */ |
5e0d2a6fc
|
2180 |
} |
ba3959276
|
2181 |
} |
d7ab5c46a
|
2182 2183 |
static void intel_unmap_single(struct device *dev, dma_addr_t dev_addr, size_t size, int dir) |
ffbbef5c0
|
2184 2185 2186 |
{ intel_unmap_page(dev, dev_addr, size, dir, NULL); } |
d7ab5c46a
|
2187 2188 |
static void *intel_alloc_coherent(struct device *hwdev, size_t size, dma_addr_t *dma_handle, gfp_t flags) |
ba3959276
|
2189 2190 2191 |
{ void *vaddr; int order; |
5b6985ce8
|
2192 |
size = PAGE_ALIGN(size); |
ba3959276
|
2193 2194 2195 2196 2197 2198 2199 |
order = get_order(size); flags &= ~(GFP_DMA | GFP_DMA32); vaddr = (void *)__get_free_pages(flags, order); if (!vaddr) return NULL; memset(vaddr, 0, size); |
bb9e6d650
|
2200 2201 2202 |
*dma_handle = __intel_map_single(hwdev, virt_to_bus(vaddr), size, DMA_BIDIRECTIONAL, hwdev->coherent_dma_mask); |
ba3959276
|
2203 2204 2205 2206 2207 |
if (*dma_handle) return vaddr; free_pages((unsigned long)vaddr, order); return NULL; } |
d7ab5c46a
|
2208 2209 |
static void intel_free_coherent(struct device *hwdev, size_t size, void *vaddr, dma_addr_t dma_handle) |
ba3959276
|
2210 2211 |
{ int order; |
5b6985ce8
|
2212 |
size = PAGE_ALIGN(size); |
ba3959276
|
2213 2214 2215 2216 2217 |
order = get_order(size); intel_unmap_single(hwdev, dma_handle, size, DMA_BIDIRECTIONAL); free_pages((unsigned long)vaddr, order); } |
12d4d40e6
|
2218 |
#define SG_ENT_VIRT_ADDRESS(sg) (sg_virt((sg))) |
5b6985ce8
|
2219 |
|
d7ab5c46a
|
2220 2221 2222 |
static void intel_unmap_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, enum dma_data_direction dir, struct dma_attrs *attrs) |
ba3959276
|
2223 2224 2225 2226 |
{ int i; struct pci_dev *pdev = to_pci_dev(hwdev); struct dmar_domain *domain; |
f76aec76e
|
2227 2228 2229 2230 |
unsigned long start_addr; struct iova *iova; size_t size = 0; void *addr; |
c03ab37cb
|
2231 |
struct scatterlist *sg; |
8c11e798e
|
2232 |
struct intel_iommu *iommu; |
ba3959276
|
2233 |
|
358dd8ac5
|
2234 |
if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO) |
ba3959276
|
2235 2236 2237 |
return; domain = find_domain(pdev); |
8c11e798e
|
2238 2239 2240 |
BUG_ON(!domain); iommu = domain_get_iommu(domain); |
ba3959276
|
2241 |
|
c03ab37cb
|
2242 |
iova = find_iova(&domain->iovad, IOVA_PFN(sglist[0].dma_address)); |
f76aec76e
|
2243 2244 |
if (!iova) return; |
c03ab37cb
|
2245 |
for_each_sg(sglist, sg, nelems, i) { |
f76aec76e
|
2246 2247 2248 |
addr = SG_ENT_VIRT_ADDRESS(sg); size += aligned_size((u64)addr, sg->length); } |
5b6985ce8
|
2249 |
start_addr = iova->pfn_lo << PAGE_SHIFT; |
f76aec76e
|
2250 2251 2252 2253 2254 |
/* clear the whole page */ dma_pte_clear_range(domain, start_addr, start_addr + size); /* free page tables */ dma_pte_free_pagetable(domain, start_addr, start_addr + size); |
8c11e798e
|
2255 |
if (iommu_flush_iotlb_psi(iommu, domain->id, start_addr, |
5b6985ce8
|
2256 |
size >> VTD_PAGE_SHIFT, 0)) |
8c11e798e
|
2257 |
iommu_flush_write_buffer(iommu); |
f76aec76e
|
2258 2259 2260 |
/* free iova */ __free_iova(&domain->iovad, iova); |
ba3959276
|
2261 |
} |
ba3959276
|
2262 |
static int intel_nontranslate_map_sg(struct device *hddev, |
c03ab37cb
|
2263 |
struct scatterlist *sglist, int nelems, int dir) |
ba3959276
|
2264 2265 |
{ int i; |
c03ab37cb
|
2266 |
struct scatterlist *sg; |
ba3959276
|
2267 |
|
c03ab37cb
|
2268 |
for_each_sg(sglist, sg, nelems, i) { |
12d4d40e6
|
2269 |
BUG_ON(!sg_page(sg)); |
c03ab37cb
|
2270 2271 |
sg->dma_address = virt_to_bus(SG_ENT_VIRT_ADDRESS(sg)); sg->dma_length = sg->length; |
ba3959276
|
2272 2273 2274 |
} return nelems; } |
d7ab5c46a
|
2275 2276 |
static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, enum dma_data_direction dir, struct dma_attrs *attrs) |
ba3959276
|
2277 2278 2279 |
{ void *addr; int i; |
ba3959276
|
2280 2281 |
struct pci_dev *pdev = to_pci_dev(hwdev); struct dmar_domain *domain; |
f76aec76e
|
2282 2283 2284 2285 2286 |
size_t size = 0; int prot = 0; size_t offset = 0; struct iova *iova = NULL; int ret; |
c03ab37cb
|
2287 |
struct scatterlist *sg; |
f76aec76e
|
2288 |
unsigned long start_addr; |
8c11e798e
|
2289 |
struct intel_iommu *iommu; |
ba3959276
|
2290 2291 |
BUG_ON(dir == DMA_NONE); |
358dd8ac5
|
2292 |
if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO) |
c03ab37cb
|
2293 |
return intel_nontranslate_map_sg(hwdev, sglist, nelems, dir); |
ba3959276
|
2294 |
|
f76aec76e
|
2295 2296 2297 |
domain = get_valid_domain_for_dev(pdev); if (!domain) return 0; |
8c11e798e
|
2298 |
iommu = domain_get_iommu(domain); |
c03ab37cb
|
2299 |
for_each_sg(sglist, sg, nelems, i) { |
ba3959276
|
2300 |
addr = SG_ENT_VIRT_ADDRESS(sg); |
f76aec76e
|
2301 2302 2303 |
addr = (void *)virt_to_phys(addr); size += aligned_size((u64)addr, sg->length); } |
bb9e6d650
|
2304 |
iova = __intel_alloc_iova(hwdev, domain, size, pdev->dma_mask); |
f76aec76e
|
2305 |
if (!iova) { |
c03ab37cb
|
2306 |
sglist->dma_length = 0; |
f76aec76e
|
2307 2308 2309 2310 2311 2312 2313 2314 |
return 0; } /* * Check if DMAR supports zero-length reads on write only * mappings.. */ if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL || \ |
8c11e798e
|
2315 |
!cap_zlr(iommu->cap)) |
f76aec76e
|
2316 2317 2318 |
prot |= DMA_PTE_READ; if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL) prot |= DMA_PTE_WRITE; |
5b6985ce8
|
2319 |
start_addr = iova->pfn_lo << PAGE_SHIFT; |
f76aec76e
|
2320 |
offset = 0; |
c03ab37cb
|
2321 |
for_each_sg(sglist, sg, nelems, i) { |
f76aec76e
|
2322 2323 2324 2325 |
addr = SG_ENT_VIRT_ADDRESS(sg); addr = (void *)virt_to_phys(addr); size = aligned_size((u64)addr, sg->length); ret = domain_page_mapping(domain, start_addr + offset, |
5b6985ce8
|
2326 |
((u64)addr) & PAGE_MASK, |
f76aec76e
|
2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 |
size, prot); if (ret) { /* clear the page */ dma_pte_clear_range(domain, start_addr, start_addr + offset); /* free page tables */ dma_pte_free_pagetable(domain, start_addr, start_addr + offset); /* free iova */ __free_iova(&domain->iovad, iova); |
ba3959276
|
2337 2338 |
return 0; } |
f76aec76e
|
2339 |
sg->dma_address = start_addr + offset + |
5b6985ce8
|
2340 |
((u64)addr & (~PAGE_MASK)); |
ba3959276
|
2341 |
sg->dma_length = sg->length; |
f76aec76e
|
2342 |
offset += size; |
ba3959276
|
2343 |
} |
ba3959276
|
2344 |
/* it's a non-present to present mapping */ |
8c11e798e
|
2345 |
if (iommu_flush_iotlb_psi(iommu, domain->id, |
5b6985ce8
|
2346 |
start_addr, offset >> VTD_PAGE_SHIFT, 1)) |
8c11e798e
|
2347 |
iommu_flush_write_buffer(iommu); |
ba3959276
|
2348 2349 |
return nelems; } |
dfb805e83
|
2350 2351 2352 2353 |
static int intel_mapping_error(struct device *dev, dma_addr_t dma_addr) { return !dma_addr; } |
160c1d8e4
|
2354 |
struct dma_map_ops intel_dma_ops = { |
ba3959276
|
2355 2356 |
.alloc_coherent = intel_alloc_coherent, .free_coherent = intel_free_coherent, |
ba3959276
|
2357 2358 |
.map_sg = intel_map_sg, .unmap_sg = intel_unmap_sg, |
ffbbef5c0
|
2359 2360 |
.map_page = intel_map_page, .unmap_page = intel_unmap_page, |
dfb805e83
|
2361 |
.mapping_error = intel_mapping_error, |
ba3959276
|
2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 |
}; static inline int iommu_domain_cache_init(void) { int ret = 0; iommu_domain_cache = kmem_cache_create("iommu_domain", sizeof(struct dmar_domain), 0, SLAB_HWCACHE_ALIGN, NULL); if (!iommu_domain_cache) { printk(KERN_ERR "Couldn't create iommu_domain cache "); ret = -ENOMEM; } return ret; } static inline int iommu_devinfo_cache_init(void) { int ret = 0; iommu_devinfo_cache = kmem_cache_create("iommu_devinfo", sizeof(struct device_domain_info), 0, SLAB_HWCACHE_ALIGN, |
ba3959276
|
2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 |
NULL); if (!iommu_devinfo_cache) { printk(KERN_ERR "Couldn't create devinfo cache "); ret = -ENOMEM; } return ret; } static inline int iommu_iova_cache_init(void) { int ret = 0; iommu_iova_cache = kmem_cache_create("iommu_iova", sizeof(struct iova), 0, SLAB_HWCACHE_ALIGN, |
ba3959276
|
2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 |
NULL); if (!iommu_iova_cache) { printk(KERN_ERR "Couldn't create iova cache "); ret = -ENOMEM; } return ret; } static int __init iommu_init_mempool(void) { int ret; ret = iommu_iova_cache_init(); if (ret) return ret; ret = iommu_domain_cache_init(); if (ret) goto domain_error; ret = iommu_devinfo_cache_init(); if (!ret) return ret; kmem_cache_destroy(iommu_domain_cache); domain_error: kmem_cache_destroy(iommu_iova_cache); return -ENOMEM; } static void __init iommu_exit_mempool(void) { kmem_cache_destroy(iommu_devinfo_cache); kmem_cache_destroy(iommu_domain_cache); kmem_cache_destroy(iommu_iova_cache); } |
ba3959276
|
2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 |
static void __init init_no_remapping_devices(void) { struct dmar_drhd_unit *drhd; for_each_drhd_unit(drhd) { if (!drhd->include_all) { int i; for (i = 0; i < drhd->devices_cnt; i++) if (drhd->devices[i] != NULL) break; /* ignore DMAR unit if no pci devices exist */ if (i == drhd->devices_cnt) drhd->ignored = 1; } } if (dmar_map_gfx) return; for_each_drhd_unit(drhd) { int i; if (drhd->ignored || drhd->include_all) continue; for (i = 0; i < drhd->devices_cnt; i++) if (drhd->devices[i] && !IS_GFX_DEVICE(drhd->devices[i])) break; if (i < drhd->devices_cnt) continue; /* bypass IOMMU if it is just for gfx devices */ drhd->ignored = 1; for (i = 0; i < drhd->devices_cnt; i++) { if (!drhd->devices[i]) continue; |
358dd8ac5
|
2485 |
drhd->devices[i]->dev.archdata.iommu = DUMMY_DEVICE_DOMAIN_INFO; |
ba3959276
|
2486 2487 2488 2489 2490 2491 2492 |
} } } int __init intel_iommu_init(void) { int ret = 0; |
ba3959276
|
2493 2494 |
if (dmar_table_init()) return -ENODEV; |
1886e8a90
|
2495 2496 |
if (dmar_dev_scope_init()) return -ENODEV; |
2ae210106
|
2497 2498 2499 2500 2501 2502 |
/* * Check the need for DMA-remapping initialization now. * Above initialization will also be used by Interrupt-remapping. */ if (no_iommu || swiotlb || dmar_disabled) return -ENODEV; |
ba3959276
|
2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 |
iommu_init_mempool(); dmar_init_reserved_ranges(); init_no_remapping_devices(); ret = init_dmars(); if (ret) { printk(KERN_ERR "IOMMU: dmar init failed "); put_iova_domain(&reserved_iova_list); iommu_exit_mempool(); return ret; } printk(KERN_INFO "PCI-DMA: Intel(R) Virtualization Technology for Directed I/O "); |
5e0d2a6fc
|
2519 |
init_timer(&unmap_timer); |
ba3959276
|
2520 2521 |
force_iommu = 1; dma_ops = &intel_dma_ops; |
a8bcbb0de
|
2522 2523 |
register_iommu(&intel_iommu_ops); |
ba3959276
|
2524 2525 |
return 0; } |
e820482cd
|
2526 |
|
c7151a8df
|
2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 |
static int vm_domain_add_dev_info(struct dmar_domain *domain, struct pci_dev *pdev) { struct device_domain_info *info; unsigned long flags; info = alloc_devinfo_mem(); if (!info) return -ENOMEM; info->bus = pdev->bus->number; info->devfn = pdev->devfn; info->dev = pdev; info->domain = domain; spin_lock_irqsave(&device_domain_lock, flags); list_add(&info->link, &domain->devices); list_add(&info->global, &device_domain_list); pdev->dev.archdata.iommu = info; spin_unlock_irqrestore(&device_domain_lock, flags); return 0; } static void vm_domain_remove_one_dev_info(struct dmar_domain *domain, struct pci_dev *pdev) { struct device_domain_info *info; struct intel_iommu *iommu; unsigned long flags; int found = 0; struct list_head *entry, *tmp; iommu = device_to_iommu(pdev->bus->number, pdev->devfn); if (!iommu) return; spin_lock_irqsave(&device_domain_lock, flags); list_for_each_safe(entry, tmp, &domain->devices) { info = list_entry(entry, struct device_domain_info, link); if (info->bus == pdev->bus->number && info->devfn == pdev->devfn) { list_del(&info->link); list_del(&info->global); if (info->dev) info->dev->dev.archdata.iommu = NULL; spin_unlock_irqrestore(&device_domain_lock, flags); iommu_detach_dev(iommu, info->bus, info->devfn); free_devinfo_mem(info); spin_lock_irqsave(&device_domain_lock, flags); if (found) break; else continue; } /* if there is no other devices under the same iommu * owned by this domain, clear this iommu in iommu_bmp * update iommu count and coherency */ if (device_to_iommu(info->bus, info->devfn) == iommu) found = 1; } if (found == 0) { unsigned long tmp_flags; spin_lock_irqsave(&domain->iommu_lock, tmp_flags); clear_bit(iommu->seq_id, &domain->iommu_bmp); domain->iommu_count--; domain_update_iommu_coherency(domain); spin_unlock_irqrestore(&domain->iommu_lock, tmp_flags); } spin_unlock_irqrestore(&device_domain_lock, flags); } static void vm_domain_remove_all_dev_info(struct dmar_domain *domain) { struct device_domain_info *info; struct intel_iommu *iommu; unsigned long flags1, flags2; spin_lock_irqsave(&device_domain_lock, flags1); while (!list_empty(&domain->devices)) { info = list_entry(domain->devices.next, struct device_domain_info, link); list_del(&info->link); list_del(&info->global); if (info->dev) info->dev->dev.archdata.iommu = NULL; spin_unlock_irqrestore(&device_domain_lock, flags1); iommu = device_to_iommu(info->bus, info->devfn); iommu_detach_dev(iommu, info->bus, info->devfn); /* clear this iommu in iommu_bmp, update iommu count * and coherency */ spin_lock_irqsave(&domain->iommu_lock, flags2); if (test_and_clear_bit(iommu->seq_id, &domain->iommu_bmp)) { domain->iommu_count--; domain_update_iommu_coherency(domain); } spin_unlock_irqrestore(&domain->iommu_lock, flags2); free_devinfo_mem(info); spin_lock_irqsave(&device_domain_lock, flags1); } spin_unlock_irqrestore(&device_domain_lock, flags1); } |
5e98c4b1d
|
2642 2643 |
/* domain id for virtual machine, it won't be set in context */ static unsigned long vm_domid; |
fe40f1e02
|
2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 |
static int vm_domain_min_agaw(struct dmar_domain *domain) { int i; int min_agaw = domain->agaw; i = find_first_bit(&domain->iommu_bmp, g_num_of_iommus); for (; i < g_num_of_iommus; ) { if (min_agaw > g_iommus[i]->agaw) min_agaw = g_iommus[i]->agaw; i = find_next_bit(&domain->iommu_bmp, g_num_of_iommus, i+1); } return min_agaw; } |
5e98c4b1d
|
2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 |
static struct dmar_domain *iommu_alloc_vm_domain(void) { struct dmar_domain *domain; domain = alloc_domain_mem(); if (!domain) return NULL; domain->id = vm_domid++; memset(&domain->iommu_bmp, 0, sizeof(unsigned long)); domain->flags = DOMAIN_FLAG_VIRTUAL_MACHINE; return domain; } static int vm_domain_init(struct dmar_domain *domain, int guest_width) { int adjust_width; init_iova_domain(&domain->iovad, DMA_32BIT_PFN); spin_lock_init(&domain->mapping_lock); spin_lock_init(&domain->iommu_lock); domain_reserve_special_ranges(domain); /* calculate AGAW */ domain->gaw = guest_width; adjust_width = guestwidth_to_adjustwidth(guest_width); domain->agaw = width_to_agaw(adjust_width); INIT_LIST_HEAD(&domain->devices); domain->iommu_count = 0; domain->iommu_coherency = 0; |
fe40f1e02
|
2693 |
domain->max_addr = 0; |
5e98c4b1d
|
2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 |
/* always allocate the top pgd */ domain->pgd = (struct dma_pte *)alloc_pgtable_page(); if (!domain->pgd) return -ENOMEM; domain_flush_cache(domain, domain->pgd, PAGE_SIZE); return 0; } static void iommu_free_vm_domain(struct dmar_domain *domain) { unsigned long flags; struct dmar_drhd_unit *drhd; struct intel_iommu *iommu; unsigned long i; unsigned long ndomains; for_each_drhd_unit(drhd) { if (drhd->ignored) continue; iommu = drhd->iommu; ndomains = cap_ndoms(iommu->cap); i = find_first_bit(iommu->domain_ids, ndomains); for (; i < ndomains; ) { if (iommu->domains[i] == domain) { spin_lock_irqsave(&iommu->lock, flags); clear_bit(i, iommu->domain_ids); iommu->domains[i] = NULL; spin_unlock_irqrestore(&iommu->lock, flags); break; } i = find_next_bit(iommu->domain_ids, ndomains, i+1); } } } static void vm_domain_exit(struct dmar_domain *domain) { u64 end; /* Domain 0 is reserved, so dont process it */ if (!domain) return; vm_domain_remove_all_dev_info(domain); /* destroy iovas */ put_iova_domain(&domain->iovad); end = DOMAIN_MAX_ADDR(domain->gaw); end = end & (~VTD_PAGE_MASK); /* clear ptes */ dma_pte_clear_range(domain, 0, end); /* free page tables */ dma_pte_free_pagetable(domain, 0, end); iommu_free_vm_domain(domain); free_domain_mem(domain); } |
5d450806e
|
2754 |
static int intel_iommu_domain_init(struct iommu_domain *domain) |
387179464
|
2755 |
{ |
5d450806e
|
2756 |
struct dmar_domain *dmar_domain; |
387179464
|
2757 |
|
5d450806e
|
2758 2759 |
dmar_domain = iommu_alloc_vm_domain(); if (!dmar_domain) { |
387179464
|
2760 |
printk(KERN_ERR |
5d450806e
|
2761 2762 2763 |
"intel_iommu_domain_init: dmar_domain == NULL "); return -ENOMEM; |
387179464
|
2764 |
} |
5d450806e
|
2765 |
if (vm_domain_init(dmar_domain, DEFAULT_DOMAIN_ADDRESS_WIDTH)) { |
387179464
|
2766 |
printk(KERN_ERR |
5d450806e
|
2767 2768 2769 2770 |
"intel_iommu_domain_init() failed "); vm_domain_exit(dmar_domain); return -ENOMEM; |
387179464
|
2771 |
} |
5d450806e
|
2772 |
domain->priv = dmar_domain; |
faa3d6f5f
|
2773 |
|
5d450806e
|
2774 |
return 0; |
387179464
|
2775 |
} |
387179464
|
2776 |
|
5d450806e
|
2777 |
static void intel_iommu_domain_destroy(struct iommu_domain *domain) |
387179464
|
2778 |
{ |
5d450806e
|
2779 2780 2781 2782 |
struct dmar_domain *dmar_domain = domain->priv; domain->priv = NULL; vm_domain_exit(dmar_domain); |
387179464
|
2783 |
} |
387179464
|
2784 |
|
4c5478c94
|
2785 2786 |
static int intel_iommu_attach_device(struct iommu_domain *domain, struct device *dev) |
387179464
|
2787 |
{ |
4c5478c94
|
2788 2789 |
struct dmar_domain *dmar_domain = domain->priv; struct pci_dev *pdev = to_pci_dev(dev); |
fe40f1e02
|
2790 2791 2792 |
struct intel_iommu *iommu; int addr_width; u64 end; |
faa3d6f5f
|
2793 2794 2795 2796 2797 2798 2799 2800 |
int ret; /* normally pdev is not mapped */ if (unlikely(domain_context_mapped(pdev))) { struct dmar_domain *old_domain; old_domain = find_domain(pdev); if (old_domain) { |
4c5478c94
|
2801 |
if (dmar_domain->flags & DOMAIN_FLAG_VIRTUAL_MACHINE) |
faa3d6f5f
|
2802 2803 2804 2805 2806 |
vm_domain_remove_one_dev_info(old_domain, pdev); else domain_remove_dev_info(old_domain); } } |
fe40f1e02
|
2807 2808 2809 2810 2811 2812 2813 2814 |
iommu = device_to_iommu(pdev->bus->number, pdev->devfn); if (!iommu) return -ENODEV; /* check if this iommu agaw is sufficient for max mapped address */ addr_width = agaw_to_width(iommu->agaw); end = DOMAIN_MAX_ADDR(addr_width); end = end & VTD_PAGE_MASK; |
4c5478c94
|
2815 |
if (end < dmar_domain->max_addr) { |
fe40f1e02
|
2816 2817 2818 |
printk(KERN_ERR "%s: iommu agaw (%d) is not " "sufficient for the mapped address (%llx) ", |
4c5478c94
|
2819 |
__func__, iommu->agaw, dmar_domain->max_addr); |
fe40f1e02
|
2820 2821 |
return -EFAULT; } |
4c5478c94
|
2822 |
ret = domain_context_mapping(dmar_domain, pdev); |
faa3d6f5f
|
2823 2824 |
if (ret) return ret; |
4c5478c94
|
2825 |
ret = vm_domain_add_dev_info(dmar_domain, pdev); |
faa3d6f5f
|
2826 |
return ret; |
387179464
|
2827 |
} |
387179464
|
2828 |
|
4c5478c94
|
2829 2830 |
static void intel_iommu_detach_device(struct iommu_domain *domain, struct device *dev) |
387179464
|
2831 |
{ |
4c5478c94
|
2832 2833 2834 2835 |
struct dmar_domain *dmar_domain = domain->priv; struct pci_dev *pdev = to_pci_dev(dev); vm_domain_remove_one_dev_info(dmar_domain, pdev); |
faa3d6f5f
|
2836 |
} |
c7151a8df
|
2837 |
|
dde57a210
|
2838 2839 2840 |
static int intel_iommu_map_range(struct iommu_domain *domain, unsigned long iova, phys_addr_t hpa, size_t size, int iommu_prot) |
faa3d6f5f
|
2841 |
{ |
dde57a210
|
2842 |
struct dmar_domain *dmar_domain = domain->priv; |
fe40f1e02
|
2843 2844 |
u64 max_addr; int addr_width; |
dde57a210
|
2845 |
int prot = 0; |
faa3d6f5f
|
2846 |
int ret; |
fe40f1e02
|
2847 |
|
dde57a210
|
2848 2849 2850 2851 |
if (iommu_prot & IOMMU_READ) prot |= DMA_PTE_READ; if (iommu_prot & IOMMU_WRITE) prot |= DMA_PTE_WRITE; |
fe40f1e02
|
2852 |
max_addr = (iova & VTD_PAGE_MASK) + VTD_PAGE_ALIGN(size); |
dde57a210
|
2853 |
if (dmar_domain->max_addr < max_addr) { |
fe40f1e02
|
2854 2855 2856 2857 |
int min_agaw; u64 end; /* check if minimum agaw is sufficient for mapped address */ |
dde57a210
|
2858 |
min_agaw = vm_domain_min_agaw(dmar_domain); |
fe40f1e02
|
2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 |
addr_width = agaw_to_width(min_agaw); end = DOMAIN_MAX_ADDR(addr_width); end = end & VTD_PAGE_MASK; if (end < max_addr) { printk(KERN_ERR "%s: iommu agaw (%d) is not " "sufficient for the mapped address (%llx) ", __func__, min_agaw, max_addr); return -EFAULT; } |
dde57a210
|
2869 |
dmar_domain->max_addr = max_addr; |
fe40f1e02
|
2870 |
} |
dde57a210
|
2871 |
ret = domain_page_mapping(dmar_domain, iova, hpa, size, prot); |
faa3d6f5f
|
2872 |
return ret; |
387179464
|
2873 |
} |
387179464
|
2874 |
|
dde57a210
|
2875 2876 |
static void intel_iommu_unmap_range(struct iommu_domain *domain, unsigned long iova, size_t size) |
387179464
|
2877 |
{ |
dde57a210
|
2878 |
struct dmar_domain *dmar_domain = domain->priv; |
faa3d6f5f
|
2879 2880 2881 2882 2883 |
dma_addr_t base; /* The address might not be aligned */ base = iova & VTD_PAGE_MASK; size = VTD_PAGE_ALIGN(size); |
dde57a210
|
2884 |
dma_pte_clear_range(dmar_domain, base, base + size); |
fe40f1e02
|
2885 |
|
dde57a210
|
2886 2887 |
if (dmar_domain->max_addr == base + size) dmar_domain->max_addr = base; |
387179464
|
2888 |
} |
387179464
|
2889 |
|
d14d65777
|
2890 2891 |
static phys_addr_t intel_iommu_iova_to_phys(struct iommu_domain *domain, unsigned long iova) |
387179464
|
2892 |
{ |
d14d65777
|
2893 |
struct dmar_domain *dmar_domain = domain->priv; |
387179464
|
2894 |
struct dma_pte *pte; |
faa3d6f5f
|
2895 |
u64 phys = 0; |
387179464
|
2896 |
|
d14d65777
|
2897 |
pte = addr_to_dma_pte(dmar_domain, iova); |
387179464
|
2898 |
if (pte) |
faa3d6f5f
|
2899 |
phys = dma_pte_addr(pte); |
387179464
|
2900 |
|
faa3d6f5f
|
2901 |
return phys; |
387179464
|
2902 |
} |
a8bcbb0de
|
2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 |
static struct iommu_ops intel_iommu_ops = { .domain_init = intel_iommu_domain_init, .domain_destroy = intel_iommu_domain_destroy, .attach_dev = intel_iommu_attach_device, .detach_dev = intel_iommu_detach_device, .map = intel_iommu_map_range, .unmap = intel_iommu_unmap_range, .iova_to_phys = intel_iommu_iova_to_phys, }; |
9af88143b
|
2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 |
static void __devinit quirk_iommu_rwbf(struct pci_dev *dev) { /* * Mobile 4 Series Chipset neglects to set RWBF capability, * but needs it: */ printk(KERN_INFO "DMAR: Forcing write-buffer flush capability "); rwbf_quirk = 1; } DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2a40, quirk_iommu_rwbf); |