dma-mapping.c 14.8 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1
/*
2
 *  linux/arch/arm/mm/dma-mapping.c
Linus Torvalds's avatar
Linus Torvalds committed
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
 *
 *  Copyright (C) 2000-2004 Russell King
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 *
 *  DMA uncached mapping support.
 */
#include <linux/module.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/errno.h>
#include <linux/list.h>
#include <linux/init.h>
#include <linux/device.h>
#include <linux/dma-mapping.h>

21
#include <asm/memory.h>
Linus Torvalds's avatar
Linus Torvalds committed
22 23
#include <asm/cacheflush.h>
#include <asm/tlbflush.h>
24 25 26 27 28 29
#include <asm/sizes.h>

/* Sanity check size */
#if (CONSISTENT_DMA_SIZE % SZ_2M)
#error "CONSISTENT_DMA_SIZE must be multiple of 2MiB"
#endif
Linus Torvalds's avatar
Linus Torvalds committed
30 31

#define CONSISTENT_END	(0xffe00000)
32 33
#define CONSISTENT_BASE	(CONSISTENT_END - CONSISTENT_DMA_SIZE)

Linus Torvalds's avatar
Linus Torvalds committed
34
#define CONSISTENT_OFFSET(x)	(((unsigned long)(x) - CONSISTENT_BASE) >> PAGE_SHIFT)
35 36 37
#define CONSISTENT_PTE_INDEX(x) (((unsigned long)(x) - CONSISTENT_BASE) >> PGDIR_SHIFT)
#define NUM_CONSISTENT_PTES (CONSISTENT_DMA_SIZE >> PGDIR_SHIFT)

Linus Torvalds's avatar
Linus Torvalds committed
38 39

/*
40
 * These are the page tables (2MB each) covering uncached, DMA consistent allocations
Linus Torvalds's avatar
Linus Torvalds committed
41
 */
42
static pte_t *consistent_pte[NUM_CONSISTENT_PTES];
Linus Torvalds's avatar
Linus Torvalds committed
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78
static DEFINE_SPINLOCK(consistent_lock);

/*
 * VM region handling support.
 *
 * This should become something generic, handling VM region allocations for
 * vmalloc and similar (ioremap, module space, etc).
 *
 * I envisage vmalloc()'s supporting vm_struct becoming:
 *
 *  struct vm_struct {
 *    struct vm_region	region;
 *    unsigned long	flags;
 *    struct page	**pages;
 *    unsigned int	nr_pages;
 *    unsigned long	phys_addr;
 *  };
 *
 * get_vm_area() would then call vm_region_alloc with an appropriate
 * struct vm_region head (eg):
 *
 *  struct vm_region vmalloc_head = {
 *	.vm_list	= LIST_HEAD_INIT(vmalloc_head.vm_list),
 *	.vm_start	= VMALLOC_START,
 *	.vm_end		= VMALLOC_END,
 *  };
 *
 * However, vmalloc_head.vm_start is variable (typically, it is dependent on
 * the amount of RAM found at boot time.)  I would imagine that get_vm_area()
 * would have to initialise this each time prior to calling vm_region_alloc().
 */
struct vm_region {
	struct list_head	vm_list;
	unsigned long		vm_start;
	unsigned long		vm_end;
	struct page		*vm_pages;
79
	int			vm_active;
Linus Torvalds's avatar
Linus Torvalds committed
80 81 82 83 84 85 86 87 88
};

static struct vm_region consistent_head = {
	.vm_list	= LIST_HEAD_INIT(consistent_head.vm_list),
	.vm_start	= CONSISTENT_BASE,
	.vm_end		= CONSISTENT_END,
};

static struct vm_region *
Al Viro's avatar
Al Viro committed
89
vm_region_alloc(struct vm_region *head, size_t size, gfp_t gfp)
Linus Torvalds's avatar
Linus Torvalds committed
90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
{
	unsigned long addr = head->vm_start, end = head->vm_end - size;
	unsigned long flags;
	struct vm_region *c, *new;

	new = kmalloc(sizeof(struct vm_region), gfp);
	if (!new)
		goto out;

	spin_lock_irqsave(&consistent_lock, flags);

	list_for_each_entry(c, &head->vm_list, vm_list) {
		if ((addr + size) < addr)
			goto nospc;
		if ((addr + size) <= c->vm_start)
			goto found;
		addr = c->vm_end;
		if (addr > end)
			goto nospc;
	}

 found:
	/*
	 * Insert this entry _before_ the one we found.
	 */
	list_add_tail(&new->vm_list, &c->vm_list);
	new->vm_start = addr;
	new->vm_end = addr + size;
118
	new->vm_active = 1;
Linus Torvalds's avatar
Linus Torvalds committed
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134

	spin_unlock_irqrestore(&consistent_lock, flags);
	return new;

 nospc:
	spin_unlock_irqrestore(&consistent_lock, flags);
	kfree(new);
 out:
	return NULL;
}

static struct vm_region *vm_region_find(struct vm_region *head, unsigned long addr)
{
	struct vm_region *c;
	
	list_for_each_entry(c, &head->vm_list, vm_list) {
135
		if (c->vm_active && c->vm_start == addr)
Linus Torvalds's avatar
Linus Torvalds committed
136 137 138 139 140 141 142 143 144 145 146 147
			goto out;
	}
	c = NULL;
 out:
	return c;
}

#ifdef CONFIG_HUGETLB_PAGE
#error ARM Coherent DMA allocator does not (yet) support huge TLB
#endif

static void *
Al Viro's avatar
Al Viro committed
148
__dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp,
Linus Torvalds's avatar
Linus Torvalds committed
149 150 151 152 153 154 155
	    pgprot_t prot)
{
	struct page *page;
	struct vm_region *c;
	unsigned long order;
	u64 mask = ISA_DMA_THRESHOLD, limit;

156
	if (!consistent_pte[0]) {
Linus Torvalds's avatar
Linus Torvalds committed
157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
		printk(KERN_ERR "%s: not initialised\n", __func__);
		dump_stack();
		return NULL;
	}

	if (dev) {
		mask = dev->coherent_dma_mask;

		/*
		 * Sanity check the DMA mask - it must be non-zero, and
		 * must be able to be satisfied by a DMA allocation.
		 */
		if (mask == 0) {
			dev_warn(dev, "coherent DMA mask is unset\n");
			goto no_page;
		}

		if ((~mask) & ISA_DMA_THRESHOLD) {
			dev_warn(dev, "coherent DMA mask %#llx is smaller "
				 "than system GFP_DMA mask %#llx\n",
				 mask, (unsigned long long)ISA_DMA_THRESHOLD);
			goto no_page;
		}
	}

	/*
	 * Sanity check the allocation size.
	 */
	size = PAGE_ALIGN(size);
	limit = (mask + 1) & ~mask;
	if ((limit && size >= limit) ||
	    size >= (CONSISTENT_END - CONSISTENT_BASE)) {
		printk(KERN_WARNING "coherent allocation too big "
		       "(requested %#x mask %#llx)\n", size, mask);
		goto no_page;
	}

	order = get_order(size);

	if (mask != 0xffffffff)
		gfp |= GFP_DMA;

	page = alloc_pages(gfp, order);
	if (!page)
		goto no_page;

	/*
	 * Invalidate any data that might be lurking in the
	 * kernel direct-mapped region for device DMA.
	 */
	{
208 209 210 211
		void *ptr = page_address(page);
		memset(ptr, 0, size);
		dmac_flush_range(ptr, ptr + size);
		outer_flush_range(__pa(ptr), __pa(ptr) + size);
Linus Torvalds's avatar
Linus Torvalds committed
212 213 214 215 216 217 218 219
	}

	/*
	 * Allocate a virtual address in the consistent mapping region.
	 */
	c = vm_region_alloc(&consistent_head, size,
			    gfp & ~(__GFP_DMA | __GFP_HIGHMEM));
	if (c) {
220
		pte_t *pte;
Linus Torvalds's avatar
Linus Torvalds committed
221
		struct page *end = page + (1 << order);
222 223
		int idx = CONSISTENT_PTE_INDEX(c->vm_start);
		u32 off = CONSISTENT_OFFSET(c->vm_start) & (PTRS_PER_PTE-1);
Linus Torvalds's avatar
Linus Torvalds committed
224

225
		pte = consistent_pte[idx] + off;
Linus Torvalds's avatar
Linus Torvalds committed
226 227
		c->vm_pages = page;

228 229
		split_page(page, order);

Linus Torvalds's avatar
Linus Torvalds committed
230 231 232 233 234 235 236 237 238 239 240 241
		/*
		 * Set the "dma handle"
		 */
		*handle = page_to_dma(dev, page);

		do {
			BUG_ON(!pte_none(*pte));

			/*
			 * x86 does not mark the pages reserved...
			 */
			SetPageReserved(page);
242
			set_pte_ext(pte, mk_pte(page, prot), 0);
Linus Torvalds's avatar
Linus Torvalds committed
243 244
			page++;
			pte++;
245 246 247 248 249
			off++;
			if (off >= PTRS_PER_PTE) {
				off = 0;
				pte = consistent_pte[++idx];
			}
Linus Torvalds's avatar
Linus Torvalds committed
250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274
		} while (size -= PAGE_SIZE);

		/*
		 * Free the otherwise unused pages.
		 */
		while (page < end) {
			__free_page(page);
			page++;
		}

		return (void *)c->vm_start;
	}

	if (page)
		__free_pages(page, order);
 no_page:
	*handle = ~0;
	return NULL;
}

/*
 * Allocate DMA-coherent memory space and return both the kernel remapped
 * virtual and bus address for that space.
 */
void *
Al Viro's avatar
Al Viro committed
275
dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp)
Linus Torvalds's avatar
Linus Torvalds committed
276
{
277 278 279 280 281
	void *memory;

	if (dma_alloc_from_coherent(dev, size, handle, &memory))
		return memory;

282 283 284 285 286 287 288 289 290 291 292
	if (arch_is_coherent()) {
		void *virt;

		virt = kmalloc(size, gfp);
		if (!virt)
			return NULL;
		*handle =  virt_to_dma(dev, virt);

		return virt;
	}

Linus Torvalds's avatar
Linus Torvalds committed
293 294 295 296 297 298 299 300 301 302
	return __dma_alloc(dev, size, handle, gfp,
			   pgprot_noncached(pgprot_kernel));
}
EXPORT_SYMBOL(dma_alloc_coherent);

/*
 * Allocate a writecombining region, in much the same way as
 * dma_alloc_coherent above.
 */
void *
Al Viro's avatar
Al Viro committed
303
dma_alloc_writecombine(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp)
Linus Torvalds's avatar
Linus Torvalds committed
304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
{
	return __dma_alloc(dev, size, handle, gfp,
			   pgprot_writecombine(pgprot_kernel));
}
EXPORT_SYMBOL(dma_alloc_writecombine);

static int dma_mmap(struct device *dev, struct vm_area_struct *vma,
		    void *cpu_addr, dma_addr_t dma_addr, size_t size)
{
	unsigned long flags, user_size, kern_size;
	struct vm_region *c;
	int ret = -ENXIO;

	user_size = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;

	spin_lock_irqsave(&consistent_lock, flags);
	c = vm_region_find(&consistent_head, (unsigned long)cpu_addr);
	spin_unlock_irqrestore(&consistent_lock, flags);

	if (c) {
		unsigned long off = vma->vm_pgoff;

		kern_size = (c->vm_end - c->vm_start) >> PAGE_SHIFT;

		if (off < kern_size &&
		    user_size <= (kern_size - off)) {
			ret = remap_pfn_range(vma, vma->vm_start,
					      page_to_pfn(c->vm_pages) + off,
					      user_size << PAGE_SHIFT,
					      vma->vm_page_prot);
		}
	}

	return ret;
}

int dma_mmap_coherent(struct device *dev, struct vm_area_struct *vma,
		      void *cpu_addr, dma_addr_t dma_addr, size_t size)
{
	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
	return dma_mmap(dev, vma, cpu_addr, dma_addr, size);
}
EXPORT_SYMBOL(dma_mmap_coherent);

int dma_mmap_writecombine(struct device *dev, struct vm_area_struct *vma,
			  void *cpu_addr, dma_addr_t dma_addr, size_t size)
{
	vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
	return dma_mmap(dev, vma, cpu_addr, dma_addr, size);
}
EXPORT_SYMBOL(dma_mmap_writecombine);

/*
 * free a page as defined by the above mapping.
358
 * Must not be called with IRQs disabled.
Linus Torvalds's avatar
Linus Torvalds committed
359 360 361 362 363 364
 */
void dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, dma_addr_t handle)
{
	struct vm_region *c;
	unsigned long flags, addr;
	pte_t *ptep;
365 366
	int idx;
	u32 off;
Linus Torvalds's avatar
Linus Torvalds committed
367

368 369
	WARN_ON(irqs_disabled());

370 371 372
	if (dma_release_from_coherent(dev, get_order(size), cpu_addr))
		return;

373 374 375 376 377
	if (arch_is_coherent()) {
		kfree(cpu_addr);
		return;
	}

Linus Torvalds's avatar
Linus Torvalds committed
378 379 380 381 382 383 384
	size = PAGE_ALIGN(size);

	spin_lock_irqsave(&consistent_lock, flags);
	c = vm_region_find(&consistent_head, (unsigned long)cpu_addr);
	if (!c)
		goto no_area;

385 386 387
	c->vm_active = 0;
	spin_unlock_irqrestore(&consistent_lock, flags);

Linus Torvalds's avatar
Linus Torvalds committed
388 389 390 391 392 393 394
	if ((c->vm_end - c->vm_start) != size) {
		printk(KERN_ERR "%s: freeing wrong coherent size (%ld != %d)\n",
		       __func__, c->vm_end - c->vm_start, size);
		dump_stack();
		size = c->vm_end - c->vm_start;
	}

395 396 397
	idx = CONSISTENT_PTE_INDEX(c->vm_start);
	off = CONSISTENT_OFFSET(c->vm_start) & (PTRS_PER_PTE-1);
	ptep = consistent_pte[idx] + off;
Linus Torvalds's avatar
Linus Torvalds committed
398 399 400 401 402 403 404
	addr = c->vm_start;
	do {
		pte_t pte = ptep_get_and_clear(&init_mm, addr, ptep);
		unsigned long pfn;

		ptep++;
		addr += PAGE_SIZE;
405 406 407 408 409
		off++;
		if (off >= PTRS_PER_PTE) {
			off = 0;
			ptep = consistent_pte[++idx];
		}
Linus Torvalds's avatar
Linus Torvalds committed
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432

		if (!pte_none(pte) && pte_present(pte)) {
			pfn = pte_pfn(pte);

			if (pfn_valid(pfn)) {
				struct page *page = pfn_to_page(pfn);

				/*
				 * x86 does not mark the pages reserved...
				 */
				ClearPageReserved(page);

				__free_page(page);
				continue;
			}
		}

		printk(KERN_CRIT "%s: bad page in kernel page table\n",
		       __func__);
	} while (size -= PAGE_SIZE);

	flush_tlb_kernel_range(c->vm_start, c->vm_end);

433
	spin_lock_irqsave(&consistent_lock, flags);
Linus Torvalds's avatar
Linus Torvalds committed
434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455
	list_del(&c->vm_list);
	spin_unlock_irqrestore(&consistent_lock, flags);

	kfree(c);
	return;

 no_area:
	spin_unlock_irqrestore(&consistent_lock, flags);
	printk(KERN_ERR "%s: trying to free invalid coherent area: %p\n",
	       __func__, cpu_addr);
	dump_stack();
}
EXPORT_SYMBOL(dma_free_coherent);

/*
 * Initialise the consistent memory allocation.
 */
static int __init consistent_init(void)
{
	pgd_t *pgd;
	pmd_t *pmd;
	pte_t *pte;
456 457
	int ret = 0, i = 0;
	u32 base = CONSISTENT_BASE;
Linus Torvalds's avatar
Linus Torvalds committed
458 459

	do {
460 461
		pgd = pgd_offset(&init_mm, base);
		pmd = pmd_alloc(&init_mm, pgd, base);
Linus Torvalds's avatar
Linus Torvalds committed
462 463 464 465 466 467 468
		if (!pmd) {
			printk(KERN_ERR "%s: no pmd tables\n", __func__);
			ret = -ENOMEM;
			break;
		}
		WARN_ON(!pmd_none(*pmd));

469
		pte = pte_alloc_kernel(pmd, base);
Linus Torvalds's avatar
Linus Torvalds committed
470 471 472 473 474 475
		if (!pte) {
			printk(KERN_ERR "%s: no pte tables\n", __func__);
			ret = -ENOMEM;
			break;
		}

476 477 478
		consistent_pte[i++] = pte;
		base += (1 << PGDIR_SHIFT);
	} while (base < CONSISTENT_END);
Linus Torvalds's avatar
Linus Torvalds committed
479 480 481 482 483 484 485 486

	return ret;
}

core_initcall(consistent_init);

/*
 * Make an area consistent for devices.
487 488 489
 * Note: Drivers should NOT use this function directly, as it will break
 * platforms with CONFIG_DMABOUNCE.
 * Use the driver DMA support - see dma-mapping.h (dma_sync_*)
Linus Torvalds's avatar
Linus Torvalds committed
490
 */
491
void dma_cache_maint(const void *start, size_t size, int direction)
Linus Torvalds's avatar
Linus Torvalds committed
492
{
493
	const void *end = start + size;
Linus Torvalds's avatar
Linus Torvalds committed
494

495
	BUG_ON(!virt_addr_valid(start) || !virt_addr_valid(end - 1));
496

Linus Torvalds's avatar
Linus Torvalds committed
497 498 499
	switch (direction) {
	case DMA_FROM_DEVICE:		/* invalidate only */
		dmac_inv_range(start, end);
500
		outer_inv_range(__pa(start), __pa(end));
Linus Torvalds's avatar
Linus Torvalds committed
501 502 503
		break;
	case DMA_TO_DEVICE:		/* writeback only */
		dmac_clean_range(start, end);
504
		outer_clean_range(__pa(start), __pa(end));
Linus Torvalds's avatar
Linus Torvalds committed
505 506 507
		break;
	case DMA_BIDIRECTIONAL:		/* writeback and invalidate */
		dmac_flush_range(start, end);
508
		outer_flush_range(__pa(start), __pa(end));
Linus Torvalds's avatar
Linus Torvalds committed
509 510 511 512 513
		break;
	default:
		BUG();
	}
}
514
EXPORT_SYMBOL(dma_cache_maint);
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535

/**
 * dma_map_sg - map a set of SG buffers for streaming mode DMA
 * @dev: valid struct device pointer, or NULL for ISA and EISA-like devices
 * @sg: list of buffers
 * @nents: number of buffers to map
 * @dir: DMA transfer direction
 *
 * Map a set of buffers described by scatterlist in streaming mode for DMA.
 * This is the scatter-gather version of the dma_map_single interface.
 * Here the scatter gather list elements are each tagged with the
 * appropriate dma address and length.  They are obtained via
 * sg_dma_{address,length}.
 *
 * Device ownership issues as mentioned for dma_map_single are the same
 * here.
 */
int dma_map_sg(struct device *dev, struct scatterlist *sg, int nents,
		enum dma_data_direction dir)
{
	struct scatterlist *s;
536
	int i, j;
537 538

	for_each_sg(sg, s, nents, i) {
539 540 541 542
		s->dma_address = dma_map_page(dev, sg_page(s), s->offset,
						s->length, dir);
		if (dma_mapping_error(dev, s->dma_address))
			goto bad_mapping;
543 544
	}
	return nents;
545 546 547 548 549

 bad_mapping:
	for_each_sg(sg, s, i, j)
		dma_unmap_page(dev, sg_dma_address(s), sg_dma_len(s), dir);
	return 0;
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565
}
EXPORT_SYMBOL(dma_map_sg);

/**
 * dma_unmap_sg - unmap a set of SG buffers mapped by dma_map_sg
 * @dev: valid struct device pointer, or NULL for ISA and EISA-like devices
 * @sg: list of buffers
 * @nents: number of buffers to unmap (returned from dma_map_sg)
 * @dir: DMA transfer direction (same as was passed to dma_map_sg)
 *
 * Unmap a set of streaming mode DMA translations.  Again, CPU access
 * rules concerning calls here are the same as for dma_unmap_single().
 */
void dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nents,
		enum dma_data_direction dir)
{
566 567 568 569 570
	struct scatterlist *s;
	int i;

	for_each_sg(sg, s, nents, i)
		dma_unmap_page(dev, sg_dma_address(s), sg_dma_len(s), dir);
571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587
}
EXPORT_SYMBOL(dma_unmap_sg);

/**
 * dma_sync_sg_for_cpu
 * @dev: valid struct device pointer, or NULL for ISA and EISA-like devices
 * @sg: list of buffers
 * @nents: number of buffers to map (returned from dma_map_sg)
 * @dir: DMA transfer direction (same as was passed to dma_map_sg)
 */
void dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg,
			int nents, enum dma_data_direction dir)
{
	struct scatterlist *s;
	int i;

	for_each_sg(sg, s, nents, i) {
588 589
		dmabounce_sync_for_cpu(dev, sg_dma_address(s), 0,
					sg_dma_len(s), dir);
590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607
	}
}
EXPORT_SYMBOL(dma_sync_sg_for_cpu);

/**
 * dma_sync_sg_for_device
 * @dev: valid struct device pointer, or NULL for ISA and EISA-like devices
 * @sg: list of buffers
 * @nents: number of buffers to map (returned from dma_map_sg)
 * @dir: DMA transfer direction (same as was passed to dma_map_sg)
 */
void dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg,
			int nents, enum dma_data_direction dir)
{
	struct scatterlist *s;
	int i;

	for_each_sg(sg, s, nents, i) {
608 609 610 611
		if (!dmabounce_sync_for_device(dev, sg_dma_address(s), 0,
					sg_dma_len(s), dir))
			continue;

612 613 614 615 616
		if (!arch_is_coherent())
			dma_cache_maint(sg_virt(s), s->length, dir);
	}
}
EXPORT_SYMBOL(dma_sync_sg_for_device);