mmu.c 22.5 KB
Newer Older
1 2 3 4 5 6 7 8 9
/*
 *  linux/arch/arm/mm/mmu.c
 *
 *  Copyright (C) 1995-2005 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.
 */
10
#include <linux/module.h>
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/init.h>
#include <linux/bootmem.h>
#include <linux/mman.h>
#include <linux/nodemask.h>

#include <asm/mach-types.h>
#include <asm/setup.h>
#include <asm/sizes.h>
#include <asm/tlb.h>

#include <asm/mach/arch.h>
#include <asm/mach/map.h>

#include "mm.h"

DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);

30
extern void _stext, _etext, __data_start, _end;
31 32 33 34 35 36 37
extern pgd_t swapper_pg_dir[PTRS_PER_PGD];

/*
 * empty_zero_page is a special page that is used for
 * zero-initialized data and COW.
 */
struct page *empty_zero_page;
38
EXPORT_SYMBOL(empty_zero_page);
39 40 41 42 43 44

/*
 * The pmd table for the upper-most set of pages.
 */
pmd_t *top_pmd;

45 46 47 48 49 50 51 52
#define CPOLICY_UNCACHED	0
#define CPOLICY_BUFFERED	1
#define CPOLICY_WRITETHROUGH	2
#define CPOLICY_WRITEBACK	3
#define CPOLICY_WRITEALLOC	4

static unsigned int cachepolicy __initdata = CPOLICY_WRITEBACK;
static unsigned int ecc_mask __initdata = 0;
53
pgprot_t pgprot_user;
54 55
pgprot_t pgprot_kernel;

56
EXPORT_SYMBOL(pgprot_user);
57 58 59 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 92 93 94 95
EXPORT_SYMBOL(pgprot_kernel);

struct cachepolicy {
	const char	policy[16];
	unsigned int	cr_mask;
	unsigned int	pmd;
	unsigned int	pte;
};

static struct cachepolicy cache_policies[] __initdata = {
	{
		.policy		= "uncached",
		.cr_mask	= CR_W|CR_C,
		.pmd		= PMD_SECT_UNCACHED,
		.pte		= 0,
	}, {
		.policy		= "buffered",
		.cr_mask	= CR_C,
		.pmd		= PMD_SECT_BUFFERED,
		.pte		= PTE_BUFFERABLE,
	}, {
		.policy		= "writethrough",
		.cr_mask	= 0,
		.pmd		= PMD_SECT_WT,
		.pte		= PTE_CACHEABLE,
	}, {
		.policy		= "writeback",
		.cr_mask	= 0,
		.pmd		= PMD_SECT_WB,
		.pte		= PTE_BUFFERABLE|PTE_CACHEABLE,
	}, {
		.policy		= "writealloc",
		.cr_mask	= 0,
		.pmd		= PMD_SECT_WBWA,
		.pte		= PTE_BUFFERABLE|PTE_CACHEABLE,
	}
};

/*
S
Simon Arlott 已提交
96
 * These are useful for identifying cache coherency
97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
 * problems by allowing the cache or the cache and
 * writebuffer to be turned off.  (Note: the write
 * buffer should not be on and the cache off).
 */
static void __init early_cachepolicy(char **p)
{
	int i;

	for (i = 0; i < ARRAY_SIZE(cache_policies); i++) {
		int len = strlen(cache_policies[i].policy);

		if (memcmp(*p, cache_policies[i].policy, len) == 0) {
			cachepolicy = i;
			cr_alignment &= ~cache_policies[i].cr_mask;
			cr_no_alignment &= ~cache_policies[i].cr_mask;
			*p += len;
			break;
		}
	}
	if (i == ARRAY_SIZE(cache_policies))
		printk(KERN_ERR "ERROR: unknown or unsupported cache policy\n");
118 119 120 121
	if (cpu_architecture() >= CPU_ARCH_ARMv6) {
		printk(KERN_WARNING "Only cachepolicy=writeback supported on ARMv6 and later\n");
		cachepolicy = CPOLICY_WRITEBACK;
	}
122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163
	flush_cache_all();
	set_cr(cr_alignment);
}
__early_param("cachepolicy=", early_cachepolicy);

static void __init early_nocache(char **__unused)
{
	char *p = "buffered";
	printk(KERN_WARNING "nocache is deprecated; use cachepolicy=%s\n", p);
	early_cachepolicy(&p);
}
__early_param("nocache", early_nocache);

static void __init early_nowrite(char **__unused)
{
	char *p = "uncached";
	printk(KERN_WARNING "nowb is deprecated; use cachepolicy=%s\n", p);
	early_cachepolicy(&p);
}
__early_param("nowb", early_nowrite);

static void __init early_ecc(char **p)
{
	if (memcmp(*p, "on", 2) == 0) {
		ecc_mask = PMD_PROTECTION;
		*p += 2;
	} else if (memcmp(*p, "off", 3) == 0) {
		ecc_mask = 0;
		*p += 3;
	}
}
__early_param("ecc=", early_ecc);

static int __init noalign_setup(char *__unused)
{
	cr_alignment &= ~CR_A;
	cr_no_alignment &= ~CR_A;
	set_cr(cr_alignment);
	return 1;
}
__setup("noalign", noalign_setup);

164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
#ifndef CONFIG_SMP
void adjust_cr(unsigned long mask, unsigned long set)
{
	unsigned long flags;

	mask &= ~CR_A;

	set &= mask;

	local_irq_save(flags);

	cr_no_alignment = (cr_no_alignment & ~mask) | set;
	cr_alignment = (cr_alignment & ~mask) | set;

	set_cr((get_cr() & ~mask) | set);

	local_irq_restore(flags);
}
#endif

184 185 186
#define PROT_PTE_DEVICE		L_PTE_PRESENT|L_PTE_YOUNG|L_PTE_DIRTY|L_PTE_WRITE
#define PROT_SECT_DEVICE	PMD_TYPE_SECT|PMD_SECT_XN|PMD_SECT_AP_WRITE

187
static struct mem_type mem_types[] = {
188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212
	[MT_DEVICE] = {		  /* Strongly ordered / ARMv6 shared device */
		.prot_pte	= PROT_PTE_DEVICE,
		.prot_l1	= PMD_TYPE_TABLE,
		.prot_sect	= PROT_SECT_DEVICE | PMD_SECT_UNCACHED,
		.domain		= DOMAIN_IO,
	},
	[MT_DEVICE_NONSHARED] = { /* ARMv6 non-shared device */
		.prot_pte	= PROT_PTE_DEVICE,
		.prot_pte_ext	= PTE_EXT_TEX(2),
		.prot_l1	= PMD_TYPE_TABLE,
		.prot_sect	= PROT_SECT_DEVICE | PMD_SECT_TEX(2),
		.domain		= DOMAIN_IO,
	},
	[MT_DEVICE_CACHED] = {	  /* ioremap_cached */
		.prot_pte	= PROT_PTE_DEVICE | L_PTE_CACHEABLE | L_PTE_BUFFERABLE,
		.prot_l1	= PMD_TYPE_TABLE,
		.prot_sect	= PROT_SECT_DEVICE | PMD_SECT_WB,
		.domain		= DOMAIN_IO,
	},	
	[MT_DEVICE_IXP2000] = {	  /* IXP2400 requires XCB=101 for on-chip I/O */
		.prot_pte	= PROT_PTE_DEVICE,
		.prot_l1	= PMD_TYPE_TABLE,
		.prot_sect	= PROT_SECT_DEVICE | PMD_SECT_BUFFERABLE |
				  PMD_SECT_TEX(1),
		.domain		= DOMAIN_IO,
213
	},
214 215 216 217 218 219
	[MT_DEVICE_WC] = {	/* ioremap_wc */
		.prot_pte	= PROT_PTE_DEVICE,
		.prot_l1	= PMD_TYPE_TABLE,
		.prot_sect	= PROT_SECT_DEVICE,
		.domain		= DOMAIN_IO,
	},
220
	[MT_CACHECLEAN] = {
221
		.prot_sect = PMD_TYPE_SECT | PMD_SECT_XN,
222 223 224
		.domain    = DOMAIN_KERNEL,
	},
	[MT_MINICLEAN] = {
225
		.prot_sect = PMD_TYPE_SECT | PMD_SECT_XN | PMD_SECT_MINICACHE,
226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
		.domain    = DOMAIN_KERNEL,
	},
	[MT_LOW_VECTORS] = {
		.prot_pte  = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY |
				L_PTE_EXEC,
		.prot_l1   = PMD_TYPE_TABLE,
		.domain    = DOMAIN_USER,
	},
	[MT_HIGH_VECTORS] = {
		.prot_pte  = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY |
				L_PTE_USER | L_PTE_EXEC,
		.prot_l1   = PMD_TYPE_TABLE,
		.domain    = DOMAIN_USER,
	},
	[MT_MEMORY] = {
241
		.prot_sect = PMD_TYPE_SECT | PMD_SECT_AP_WRITE,
242 243 244
		.domain    = DOMAIN_KERNEL,
	},
	[MT_ROM] = {
245
		.prot_sect = PMD_TYPE_SECT,
246 247 248 249
		.domain    = DOMAIN_KERNEL,
	},
};

250 251 252 253 254
const struct mem_type *get_mem_type(unsigned int type)
{
	return type < ARRAY_SIZE(mem_types) ? &mem_types[type] : NULL;
}

255 256 257 258 259 260 261 262 263 264 265
/*
 * Adjust the PMD section entries according to the CPU in use.
 */
static void __init build_mem_type_table(void)
{
	struct cachepolicy *cp;
	unsigned int cr = get_cr();
	unsigned int user_pgprot, kern_pgprot;
	int cpu_arch = cpu_architecture();
	int i;

266
	if (cpu_arch < CPU_ARCH_ARMv6) {
267
#if defined(CONFIG_CPU_DCACHE_DISABLE)
268 269
		if (cachepolicy > CPOLICY_BUFFERED)
			cachepolicy = CPOLICY_BUFFERED;
270
#elif defined(CONFIG_CPU_DCACHE_WRITETHROUGH)
271 272
		if (cachepolicy > CPOLICY_WRITETHROUGH)
			cachepolicy = CPOLICY_WRITETHROUGH;
273
#endif
274
	}
275 276 277 278 279 280
	if (cpu_arch < CPU_ARCH_ARMv5) {
		if (cachepolicy >= CPOLICY_WRITEALLOC)
			cachepolicy = CPOLICY_WRITEBACK;
		ecc_mask = 0;
	}

281 282 283 284 285 286 287 288 289 290 291 292 293 294
	/*
	 * On non-Xscale3 ARMv5-and-older systems, use CB=01
	 * (Uncached/Buffered) for ioremap_wc() mappings.  On XScale3
	 * and ARMv6+, use TEXCB=00100 mappings (Inner/Outer Uncacheable
	 * in xsc3 parlance, Uncached Normal in ARMv6 parlance).
	 */
	if (cpu_is_xsc3() || cpu_arch >= CPU_ARCH_ARMv6) {
		mem_types[MT_DEVICE_WC].prot_pte_ext |= PTE_EXT_TEX(1);
		mem_types[MT_DEVICE_WC].prot_sect |= PMD_SECT_TEX(1);
	} else {
		mem_types[MT_DEVICE_WC].prot_pte |= L_PTE_BUFFERABLE;
		mem_types[MT_DEVICE_WC].prot_sect |= PMD_SECT_BUFFERABLE;
	}

295
	/*
296 297 298
	 * ARMv5 and lower, bit 4 must be set for page tables.
	 * (was: cache "update-able on write" bit on ARM610)
	 * However, Xscale cores require this bit to be cleared.
299
	 */
300 301
	if (cpu_is_xscale()) {
		for (i = 0; i < ARRAY_SIZE(mem_types); i++) {
302
			mem_types[i].prot_sect &= ~PMD_BIT4;
303 304 305 306
			mem_types[i].prot_l1 &= ~PMD_BIT4;
		}
	} else if (cpu_arch < CPU_ARCH_ARMv6) {
		for (i = 0; i < ARRAY_SIZE(mem_types); i++) {
307 308
			if (mem_types[i].prot_l1)
				mem_types[i].prot_l1 |= PMD_BIT4;
309 310 311 312
			if (mem_types[i].prot_sect)
				mem_types[i].prot_sect |= PMD_BIT4;
		}
	}
313 314 315 316 317 318 319 320 321 322 323

	cp = &cache_policies[cachepolicy];
	kern_pgprot = user_pgprot = cp->pte;

	/*
	 * Enable CPU-specific coherency if supported.
	 * (Only available on XSC3 at the moment.)
	 */
	if (arch_is_coherent()) {
		if (cpu_is_xsc3()) {
			mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S;
324
			mem_types[MT_MEMORY].prot_pte |= L_PTE_SHARED;
325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
		}
	}

	/*
	 * ARMv6 and above have extended page tables.
	 */
	if (cpu_arch >= CPU_ARCH_ARMv6 && (cr & CR_XP)) {
		/*
		 * Mark cache clean areas and XIP ROM read only
		 * from SVC mode and no access from userspace.
		 */
		mem_types[MT_ROM].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE;
		mem_types[MT_MINICLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE;
		mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE;

		/*
		 * Mark the device area as "shared device"
		 */
		mem_types[MT_DEVICE].prot_pte |= L_PTE_BUFFERABLE;
		mem_types[MT_DEVICE].prot_sect |= PMD_SECT_BUFFERED;

#ifdef CONFIG_SMP
		/*
		 * Mark memory with the "shared" attribute for SMP systems
		 */
		user_pgprot |= L_PTE_SHARED;
		kern_pgprot |= L_PTE_SHARED;
		mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S;
#endif
	}

	for (i = 0; i < 16; i++) {
		unsigned long v = pgprot_val(protection_map[i]);
		v = (v & ~(L_PTE_BUFFERABLE|L_PTE_CACHEABLE)) | user_pgprot;
		protection_map[i] = __pgprot(v);
	}

	mem_types[MT_LOW_VECTORS].prot_pte |= kern_pgprot;
	mem_types[MT_HIGH_VECTORS].prot_pte |= kern_pgprot;

	if (cpu_arch >= CPU_ARCH_ARMv5) {
#ifndef CONFIG_SMP
		/*
		 * Only use write-through for non-SMP systems
		 */
		mem_types[MT_LOW_VECTORS].prot_pte &= ~L_PTE_BUFFERABLE;
		mem_types[MT_HIGH_VECTORS].prot_pte &= ~L_PTE_BUFFERABLE;
#endif
	} else {
		mem_types[MT_MINICLEAN].prot_sect &= ~PMD_SECT_TEX(1);
	}

377
	pgprot_user   = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | user_pgprot);
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397
	pgprot_kernel = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG |
				 L_PTE_DIRTY | L_PTE_WRITE |
				 L_PTE_EXEC | kern_pgprot);

	mem_types[MT_LOW_VECTORS].prot_l1 |= ecc_mask;
	mem_types[MT_HIGH_VECTORS].prot_l1 |= ecc_mask;
	mem_types[MT_MEMORY].prot_sect |= ecc_mask | cp->pmd;
	mem_types[MT_ROM].prot_sect |= cp->pmd;

	switch (cp->pmd) {
	case PMD_SECT_WT:
		mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_WT;
		break;
	case PMD_SECT_WB:
	case PMD_SECT_WBWA:
		mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_WB;
		break;
	}
	printk("Memory policy: ECC %sabled, Data cache %s\n",
		ecc_mask ? "en" : "dis", cp->policy);
398 399 400 401 402 403 404 405

	for (i = 0; i < ARRAY_SIZE(mem_types); i++) {
		struct mem_type *t = &mem_types[i];
		if (t->prot_l1)
			t->prot_l1 |= PMD_DOMAIN(t->domain);
		if (t->prot_sect)
			t->prot_sect |= PMD_DOMAIN(t->domain);
	}
406 407 408 409
}

#define vectors_base()	(vectors_high() ? 0xffff0000 : 0)

410 411 412
static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr,
				  unsigned long end, unsigned long pfn,
				  const struct mem_type *type)
413
{
414
	pte_t *pte;
415

416 417 418 419
	if (pmd_none(*pmd)) {
		pte = alloc_bootmem_low_pages(2 * PTRS_PER_PTE * sizeof(pte_t));
		__pmd_populate(pmd, __pa(pte) | type->prot_l1);
	}
420

421 422
	pte = pte_offset_kernel(pmd, addr);
	do {
423 424
		set_pte_ext(pte, pfn_pte(pfn, __pgprot(type->prot_pte)),
			    type->prot_pte_ext);
425 426
		pfn++;
	} while (pte++, addr += PAGE_SIZE, addr != end);
427 428
}

429 430 431
static void __init alloc_init_section(pgd_t *pgd, unsigned long addr,
				      unsigned long end, unsigned long phys,
				      const struct mem_type *type)
432
{
433
	pmd_t *pmd = pmd_offset(pgd, addr);
434

435 436 437 438 439 440 441 442
	/*
	 * Try a section mapping - end, addr and phys must all be aligned
	 * to a section boundary.  Note that PMDs refer to the individual
	 * L1 entries, whereas PGDs refer to a group of L1 entries making
	 * up one logical pointer to an L2 table.
	 */
	if (((addr | end | phys) & ~SECTION_MASK) == 0) {
		pmd_t *p = pmd;
443

444 445 446 447 448 449 450
		if (addr & SECTION_SIZE)
			pmd++;

		do {
			*pmd = __pmd(phys | type->prot_sect);
			phys += SECTION_SIZE;
		} while (pmd++, addr += SECTION_SIZE, addr != end);
451

452 453 454 455 456 457 458 459
		flush_pmd_entry(p);
	} else {
		/*
		 * No need to loop; pte's aren't interested in the
		 * individual L1 entries.
		 */
		alloc_init_pte(pmd, addr, end, __phys_to_pfn(phys), type);
	}
460 461
}

462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519
static void __init create_36bit_mapping(struct map_desc *md,
					const struct mem_type *type)
{
	unsigned long phys, addr, length, end;
	pgd_t *pgd;

	addr = md->virtual;
	phys = (unsigned long)__pfn_to_phys(md->pfn);
	length = PAGE_ALIGN(md->length);

	if (!(cpu_architecture() >= CPU_ARCH_ARMv6 || cpu_is_xsc3())) {
		printk(KERN_ERR "MM: CPU does not support supersection "
		       "mapping for 0x%08llx at 0x%08lx\n",
		       __pfn_to_phys((u64)md->pfn), addr);
		return;
	}

	/* N.B.	ARMv6 supersections are only defined to work with domain 0.
	 *	Since domain assignments can in fact be arbitrary, the
	 *	'domain == 0' check below is required to insure that ARMv6
	 *	supersections are only allocated for domain 0 regardless
	 *	of the actual domain assignments in use.
	 */
	if (type->domain) {
		printk(KERN_ERR "MM: invalid domain in supersection "
		       "mapping for 0x%08llx at 0x%08lx\n",
		       __pfn_to_phys((u64)md->pfn), addr);
		return;
	}

	if ((addr | length | __pfn_to_phys(md->pfn)) & ~SUPERSECTION_MASK) {
		printk(KERN_ERR "MM: cannot create mapping for "
		       "0x%08llx at 0x%08lx invalid alignment\n",
		       __pfn_to_phys((u64)md->pfn), addr);
		return;
	}

	/*
	 * Shift bits [35:32] of address into bits [23:20] of PMD
	 * (See ARMv6 spec).
	 */
	phys |= (((md->pfn >> (32 - PAGE_SHIFT)) & 0xF) << 20);

	pgd = pgd_offset_k(addr);
	end = addr + length;
	do {
		pmd_t *pmd = pmd_offset(pgd, addr);
		int i;

		for (i = 0; i < 16; i++)
			*pmd++ = __pmd(phys | type->prot_sect | PMD_SECT_SUPER);

		addr += SUPERSECTION_SIZE;
		phys += SUPERSECTION_SIZE;
		pgd += SUPERSECTION_SIZE >> PGDIR_SHIFT;
	} while (addr != end);
}

520 521 522 523 524 525 526 527 528
/*
 * Create the page directory entries and any necessary
 * page tables for the mapping specified by `md'.  We
 * are able to cope here with varying sizes and address
 * offsets, and we take full advantage of sections and
 * supersections.
 */
void __init create_mapping(struct map_desc *md)
{
529
	unsigned long phys, addr, length, end;
530
	const struct mem_type *type;
531
	pgd_t *pgd;
532 533 534 535 536 537 538 539 540 541 542 543 544 545 546

	if (md->virtual != vectors_base() && md->virtual < TASK_SIZE) {
		printk(KERN_WARNING "BUG: not creating mapping for "
		       "0x%08llx at 0x%08lx in user region\n",
		       __pfn_to_phys((u64)md->pfn), md->virtual);
		return;
	}

	if ((md->type == MT_DEVICE || md->type == MT_ROM) &&
	    md->virtual >= PAGE_OFFSET && md->virtual < VMALLOC_END) {
		printk(KERN_WARNING "BUG: mapping for 0x%08llx at 0x%08lx "
		       "overlaps vmalloc space\n",
		       __pfn_to_phys((u64)md->pfn), md->virtual);
	}

547
	type = &mem_types[md->type];
548 549 550 551

	/*
	 * Catch 36-bit addresses
	 */
552 553 554
	if (md->pfn >= 0x100000) {
		create_36bit_mapping(md, type);
		return;
555 556
	}

557
	addr = md->virtual & PAGE_MASK;
558
	phys = (unsigned long)__pfn_to_phys(md->pfn);
559
	length = PAGE_ALIGN(md->length + (md->virtual & ~PAGE_MASK));
560

561
	if (type->prot_l1 == 0 && ((addr | phys | length) & ~SECTION_MASK)) {
562 563
		printk(KERN_WARNING "BUG: map for 0x%08lx at 0x%08lx can not "
		       "be mapped using pages, ignoring.\n",
564
		       __pfn_to_phys(md->pfn), addr);
565 566 567
		return;
	}

568 569 570 571
	pgd = pgd_offset_k(addr);
	end = addr + length;
	do {
		unsigned long next = pgd_addr_end(addr, end);
572

573
		alloc_init_section(pgd, addr, next, phys, type);
574

575 576 577
		phys += next - addr;
		addr = next;
	} while (pgd++, addr != end);
578 579 580 581 582 583 584 585 586 587 588 589 590
}

/*
 * Create the architecture specific mappings
 */
void __init iotable_init(struct map_desc *io_desc, int nr)
{
	int i;

	for (i = 0; i < nr; i++)
		create_mapping(io_desc + i);
}

591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639
static int __init check_membank_valid(struct membank *mb)
{
	/*
	 * Check whether this memory region has non-zero size.
	 */
	if (mb->size == 0)
		return 0;

	/*
	 * Check whether this memory region would entirely overlap
	 * the vmalloc area.
	 */
	if (phys_to_virt(mb->start) >= VMALLOC_MIN) {
		printk(KERN_NOTICE "Ignoring RAM at %.8lx-%.8lx "
			"(vmalloc region overlap).\n",
			mb->start, mb->start + mb->size - 1);
		return 0;
	}

	/*
	 * Check whether this memory region would partially overlap
	 * the vmalloc area.
	 */
	if (phys_to_virt(mb->start + mb->size) < phys_to_virt(mb->start) ||
	    phys_to_virt(mb->start + mb->size) > VMALLOC_MIN) {
		unsigned long newsize = VMALLOC_MIN - phys_to_virt(mb->start);

		printk(KERN_NOTICE "Truncating RAM at %.8lx-%.8lx "
			"to -%.8lx (vmalloc region overlap).\n",
			mb->start, mb->start + mb->size - 1,
			mb->start + newsize - 1);
		mb->size = newsize;
	}

	return 1;
}

static void __init sanity_check_meminfo(struct meminfo *mi)
{
	int i;
	int j;

	for (i = 0, j = 0; i < mi->nr_banks; i++) {
		if (check_membank_valid(&mi->bank[i]))
			mi->bank[j++] = mi->bank[i];
	}
	mi->nr_banks = j;
}

640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677
static inline void prepare_page_table(struct meminfo *mi)
{
	unsigned long addr;

	/*
	 * Clear out all the mappings below the kernel image.
	 */
	for (addr = 0; addr < MODULE_START; addr += PGDIR_SIZE)
		pmd_clear(pmd_off_k(addr));

#ifdef CONFIG_XIP_KERNEL
	/* The XIP kernel is mapped in the module area -- skip over it */
	addr = ((unsigned long)&_etext + PGDIR_SIZE - 1) & PGDIR_MASK;
#endif
	for ( ; addr < PAGE_OFFSET; addr += PGDIR_SIZE)
		pmd_clear(pmd_off_k(addr));

	/*
	 * Clear out all the kernel space mappings, except for the first
	 * memory bank, up to the end of the vmalloc region.
	 */
	for (addr = __phys_to_virt(mi->bank[0].start + mi->bank[0].size);
	     addr < VMALLOC_END; addr += PGDIR_SIZE)
		pmd_clear(pmd_off_k(addr));
}

/*
 * Reserve the various regions of node 0
 */
void __init reserve_node_zero(pg_data_t *pgdat)
{
	unsigned long res_size = 0;

	/*
	 * Register the kernel text and data with bootmem.
	 * Note that this can only be in node 0.
	 */
#ifdef CONFIG_XIP_KERNEL
678 679
	reserve_bootmem_node(pgdat, __pa(&__data_start), &_end - &__data_start,
			BOOTMEM_DEFAULT);
680
#else
681 682
	reserve_bootmem_node(pgdat, __pa(&_stext), &_end - &_stext,
			BOOTMEM_DEFAULT);
683 684 685 686 687 688 689
#endif

	/*
	 * Reserve the page tables.  These are already in use,
	 * and can only be in node 0.
	 */
	reserve_bootmem_node(pgdat, __pa(swapper_pg_dir),
690
			     PTRS_PER_PGD * sizeof(pgd_t), BOOTMEM_DEFAULT);
691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709

	/*
	 * Hmm... This should go elsewhere, but we really really need to
	 * stop things allocating the low memory; ideally we need a better
	 * implementation of GFP_DMA which does not assume that DMA-able
	 * memory starts at zero.
	 */
	if (machine_is_integrator() || machine_is_cintegrator())
		res_size = __pa(swapper_pg_dir) - PHYS_OFFSET;

	/*
	 * These should likewise go elsewhere.  They pre-reserve the
	 * screen memory region at the start of main system memory.
	 */
	if (machine_is_edb7211())
		res_size = 0x00020000;
	if (machine_is_p720t())
		res_size = 0x00014000;

710 711 712
	/* H1940 and RX3715 need to reserve this for suspend */

	if (machine_is_h1940() || machine_is_rx3715()) {
713 714 715 716
		reserve_bootmem_node(pgdat, 0x30003000, 0x1000,
				BOOTMEM_DEFAULT);
		reserve_bootmem_node(pgdat, 0x30081000, 0x1000,
				BOOTMEM_DEFAULT);
717 718
	}

719 720 721 722 723 724 725 726
#ifdef CONFIG_SA1111
	/*
	 * Because of the SA1111 DMA bug, we want to preserve our
	 * precious DMA-able memory...
	 */
	res_size = __pa(swapper_pg_dir) - PHYS_OFFSET;
#endif
	if (res_size)
727 728
		reserve_bootmem_node(pgdat, PHYS_OFFSET, res_size,
				BOOTMEM_DEFAULT);
729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824
}

/*
 * Set up device the mappings.  Since we clear out the page tables for all
 * mappings above VMALLOC_END, we will remove any debug device mappings.
 * This means you have to be careful how you debug this function, or any
 * called function.  This means you can't use any function or debugging
 * method which may touch any device, otherwise the kernel _will_ crash.
 */
static void __init devicemaps_init(struct machine_desc *mdesc)
{
	struct map_desc map;
	unsigned long addr;
	void *vectors;

	/*
	 * Allocate the vector page early.
	 */
	vectors = alloc_bootmem_low_pages(PAGE_SIZE);
	BUG_ON(!vectors);

	for (addr = VMALLOC_END; addr; addr += PGDIR_SIZE)
		pmd_clear(pmd_off_k(addr));

	/*
	 * Map the kernel if it is XIP.
	 * It is always first in the modulearea.
	 */
#ifdef CONFIG_XIP_KERNEL
	map.pfn = __phys_to_pfn(CONFIG_XIP_PHYS_ADDR & SECTION_MASK);
	map.virtual = MODULE_START;
	map.length = ((unsigned long)&_etext - map.virtual + ~SECTION_MASK) & SECTION_MASK;
	map.type = MT_ROM;
	create_mapping(&map);
#endif

	/*
	 * Map the cache flushing regions.
	 */
#ifdef FLUSH_BASE
	map.pfn = __phys_to_pfn(FLUSH_BASE_PHYS);
	map.virtual = FLUSH_BASE;
	map.length = SZ_1M;
	map.type = MT_CACHECLEAN;
	create_mapping(&map);
#endif
#ifdef FLUSH_BASE_MINICACHE
	map.pfn = __phys_to_pfn(FLUSH_BASE_PHYS + SZ_1M);
	map.virtual = FLUSH_BASE_MINICACHE;
	map.length = SZ_1M;
	map.type = MT_MINICLEAN;
	create_mapping(&map);
#endif

	/*
	 * Create a mapping for the machine vectors at the high-vectors
	 * location (0xffff0000).  If we aren't using high-vectors, also
	 * create a mapping at the low-vectors virtual address.
	 */
	map.pfn = __phys_to_pfn(virt_to_phys(vectors));
	map.virtual = 0xffff0000;
	map.length = PAGE_SIZE;
	map.type = MT_HIGH_VECTORS;
	create_mapping(&map);

	if (!vectors_high()) {
		map.virtual = 0;
		map.type = MT_LOW_VECTORS;
		create_mapping(&map);
	}

	/*
	 * Ask the machine support to map in the statically mapped devices.
	 */
	if (mdesc->map_io)
		mdesc->map_io();

	/*
	 * Finally flush the caches and tlb to ensure that we're in a
	 * consistent state wrt the writebuffer.  This also ensures that
	 * any write-allocated cache lines in the vector page are written
	 * back.  After this point, we can start to touch devices again.
	 */
	local_flush_tlb_all();
	flush_cache_all();
}

/*
 * paging_init() sets up the page tables, initialises the zone memory
 * maps, and sets up the zero page, bad page and bad page tables.
 */
void __init paging_init(struct meminfo *mi, struct machine_desc *mdesc)
{
	void *zero_page;

	build_mem_type_table();
825
	sanity_check_meminfo(mi);
826 827 828 829 830 831 832 833 834 835 836 837 838 839
	prepare_page_table(mi);
	bootmem_init(mi);
	devicemaps_init(mdesc);

	top_pmd = pmd_off_k(0xffff0000);

	/*
	 * allocate the zero page.  Note that we count on this going ok.
	 */
	zero_page = alloc_bootmem_low_pages(PAGE_SIZE);
	memzero(zero_page, PAGE_SIZE);
	empty_zero_page = virt_to_page(zero_page);
	flush_dcache_page(empty_zero_page);
}
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

/*
 * In order to soft-boot, we need to insert a 1:1 mapping in place of
 * the user-mode pages.  This will then ensure that we have predictable
 * results when turning the mmu off
 */
void setup_mm_for_reboot(char mode)
{
	unsigned long base_pmdval;
	pgd_t *pgd;
	int i;

	if (current->mm && current->mm->pgd)
		pgd = current->mm->pgd;
	else
		pgd = init_mm.pgd;

	base_pmdval = PMD_SECT_AP_WRITE | PMD_SECT_AP_READ | PMD_TYPE_SECT;
	if (cpu_architecture() <= CPU_ARCH_ARMv5TEJ && !cpu_is_xscale())
		base_pmdval |= PMD_BIT4;

	for (i = 0; i < FIRST_USER_PGD_NR + USER_PTRS_PER_PGD; i++, pgd++) {
		unsigned long pmdval = (i << PGDIR_SHIFT) | base_pmdval;
		pmd_t *pmd;

		pmd = pmd_off(pgd, i << PGDIR_SHIFT);
		pmd[0] = __pmd(pmdval);
		pmd[1] = __pmd(pmdval + (1 << (PGDIR_SHIFT - 1)));
		flush_pmd_entry(pmd);
	}
}