init.c 40.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*  $Id: init.c,v 1.209 2002/02/09 19:49:31 davem Exp $
 *  arch/sparc64/mm/init.c
 *
 *  Copyright (C) 1996-1999 David S. Miller (davem@caip.rutgers.edu)
 *  Copyright (C) 1997-1999 Jakub Jelinek (jj@sunsite.mff.cuni.cz)
 */
 
#include <linux/config.h>
#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/string.h>
#include <linux/init.h>
#include <linux/bootmem.h>
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/slab.h>
#include <linux/initrd.h>
#include <linux/swap.h>
#include <linux/pagemap.h>
#include <linux/fs.h>
#include <linux/seq_file.h>
22
#include <linux/kprobes.h>
23
#include <linux/cache.h>
24
#include <linux/sort.h>
L
Linus Torvalds 已提交
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

#include <asm/head.h>
#include <asm/system.h>
#include <asm/page.h>
#include <asm/pgalloc.h>
#include <asm/pgtable.h>
#include <asm/oplib.h>
#include <asm/iommu.h>
#include <asm/io.h>
#include <asm/uaccess.h>
#include <asm/mmu_context.h>
#include <asm/tlbflush.h>
#include <asm/dma.h>
#include <asm/starfire.h>
#include <asm/tlb.h>
#include <asm/spitfire.h>
#include <asm/sections.h>

extern void device_scan(void);

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 79 80 81 82 83 84 85
#define MAX_BANKS	32

static struct linux_prom64_registers pavail[MAX_BANKS] __initdata;
static struct linux_prom64_registers pavail_rescan[MAX_BANKS] __initdata;
static int pavail_ents __initdata;
static int pavail_rescan_ents __initdata;

static int cmp_p64(const void *a, const void *b)
{
	const struct linux_prom64_registers *x = a, *y = b;

	if (x->phys_addr > y->phys_addr)
		return 1;
	if (x->phys_addr < y->phys_addr)
		return -1;
	return 0;
}

static void __init read_obp_memory(const char *property,
				   struct linux_prom64_registers *regs,
				   int *num_ents)
{
	int node = prom_finddevice("/memory");
	int prop_size = prom_getproplen(node, property);
	int ents, ret, i;

	ents = prop_size / sizeof(struct linux_prom64_registers);
	if (ents > MAX_BANKS) {
		prom_printf("The machine has more %s property entries than "
			    "this kernel can support (%d).\n",
			    property, MAX_BANKS);
		prom_halt();
	}

	ret = prom_getproperty(node, property, (char *) regs, prop_size);
	if (ret == -1) {
		prom_printf("Couldn't get %s property from /memory.\n");
		prom_halt();
	}

	*num_ents = ents;
86

87 88 89 90 91 92 93 94
	/* Sanitize what we got from the firmware, by page aligning
	 * everything.
	 */
	for (i = 0; i < ents; i++) {
		unsigned long base, size;

		base = regs[i].phys_addr;
		size = regs[i].reg_size;
95

96 97 98 99 100 101 102 103 104 105 106 107
		size &= PAGE_MASK;
		if (base & ~PAGE_MASK) {
			unsigned long new_base = PAGE_ALIGN(base);

			size -= new_base - base;
			if ((long) size < 0L)
				size = 0UL;
			base = new_base;
		}
		regs[i].phys_addr = base;
		regs[i].reg_size = size;
	}
108
	sort(regs, ents, sizeof(struct linux_prom64_registers),
109 110
	     cmp_p64, NULL);
}
L
Linus Torvalds 已提交
111

112
unsigned long *sparc64_valid_addr_bitmap __read_mostly;
L
Linus Torvalds 已提交
113 114

/* Ugly, but necessary... -DaveM */
115 116 117 118
unsigned long phys_base __read_mostly;
unsigned long kern_base __read_mostly;
unsigned long kern_size __read_mostly;
unsigned long pfn_base __read_mostly;
L
Linus Torvalds 已提交
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133

/* get_new_mmu_context() uses "cache + 1".  */
DEFINE_SPINLOCK(ctx_alloc_lock);
unsigned long tlb_context_cache = CTX_FIRST_VERSION - 1;
#define CTX_BMAP_SLOTS (1UL << (CTX_NR_BITS - 6))
unsigned long mmu_context_bmap[CTX_BMAP_SLOTS];

/* References to special section boundaries */
extern char  _start[], _end[];

/* Initial ramdisk setup */
extern unsigned long sparc_ramdisk_image64;
extern unsigned int sparc_ramdisk_image;
extern unsigned int sparc_ramdisk_size;

134
struct page *mem_map_zero __read_mostly;
L
Linus Torvalds 已提交
135

136 137 138 139 140 141
unsigned int sparc64_highest_unlocked_tlb_ent __read_mostly;

unsigned long sparc64_kern_pri_context __read_mostly;
unsigned long sparc64_kern_pri_nuc_bits __read_mostly;
unsigned long sparc64_kern_sec_context __read_mostly;

L
Linus Torvalds 已提交
142 143
int bigkernel = 0;

144
kmem_cache_t *pgtable_cache __read_mostly;
L
Linus Torvalds 已提交
145

146 147 148 149
static void zero_ctor(void *addr, kmem_cache_t *cache, unsigned long flags)
{
	clear_page(addr);
}
150

151
void pgtable_cache_init(void)
L
Linus Torvalds 已提交
152
{
153 154 155 156 157 158 159 160 161
	pgtable_cache = kmem_cache_create("pgtable_cache",
					  PAGE_SIZE, PAGE_SIZE,
					  SLAB_HWCACHE_ALIGN |
					  SLAB_MUST_HWCACHE_ALIGN,
					  zero_ctor,
					  NULL);
	if (!pgtable_cache) {
		prom_printf("pgtable_cache_init(): Could not create!\n");
		prom_halt();
L
Linus Torvalds 已提交
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
	}
}

#ifdef CONFIG_DEBUG_DCFLUSH
atomic_t dcpage_flushes = ATOMIC_INIT(0);
#ifdef CONFIG_SMP
atomic_t dcpage_flushes_xcall = ATOMIC_INIT(0);
#endif
#endif

__inline__ void flush_dcache_page_impl(struct page *page)
{
#ifdef CONFIG_DEBUG_DCFLUSH
	atomic_inc(&dcpage_flushes);
#endif

#ifdef DCACHE_ALIASING_POSSIBLE
	__flush_dcache_page(page_address(page),
			    ((tlb_type == spitfire) &&
			     page_mapping(page) != NULL));
#else
	if (page_mapping(page) != NULL &&
	    tlb_type == spitfire)
		__flush_icache_page(__pa(page_address(page)));
#endif
}

#define PG_dcache_dirty		PG_arch_1
190 191 192 193 194 195
#define PG_dcache_cpu_shift	24
#define PG_dcache_cpu_mask	(256 - 1)

#if NR_CPUS > 256
#error D-cache dirty tracking and thread_info->cpu need fixing for > 256 cpus
#endif
L
Linus Torvalds 已提交
196 197

#define dcache_dirty_cpu(page) \
198
	(((page)->flags >> PG_dcache_cpu_shift) & PG_dcache_cpu_mask)
L
Linus Torvalds 已提交
199 200 201 202

static __inline__ void set_dcache_dirty(struct page *page, int this_cpu)
{
	unsigned long mask = this_cpu;
203 204 205 206 207
	unsigned long non_cpu_bits;

	non_cpu_bits = ~(PG_dcache_cpu_mask << PG_dcache_cpu_shift);
	mask = (mask << PG_dcache_cpu_shift) | (1UL << PG_dcache_dirty);

L
Linus Torvalds 已提交
208 209 210 211 212 213
	__asm__ __volatile__("1:\n\t"
			     "ldx	[%2], %%g7\n\t"
			     "and	%%g7, %1, %%g1\n\t"
			     "or	%%g1, %0, %%g1\n\t"
			     "casx	[%2], %%g7, %%g1\n\t"
			     "cmp	%%g7, %%g1\n\t"
214
			     "membar	#StoreLoad | #StoreStore\n\t"
L
Linus Torvalds 已提交
215
			     "bne,pn	%%xcc, 1b\n\t"
216
			     " nop"
L
Linus Torvalds 已提交
217 218 219 220 221 222 223 224 225 226 227 228
			     : /* no outputs */
			     : "r" (mask), "r" (non_cpu_bits), "r" (&page->flags)
			     : "g1", "g7");
}

static __inline__ void clear_dcache_dirty_cpu(struct page *page, unsigned long cpu)
{
	unsigned long mask = (1UL << PG_dcache_dirty);

	__asm__ __volatile__("! test_and_clear_dcache_dirty\n"
			     "1:\n\t"
			     "ldx	[%2], %%g7\n\t"
229
			     "srlx	%%g7, %4, %%g1\n\t"
L
Linus Torvalds 已提交
230 231 232 233 234 235
			     "and	%%g1, %3, %%g1\n\t"
			     "cmp	%%g1, %0\n\t"
			     "bne,pn	%%icc, 2f\n\t"
			     " andn	%%g7, %1, %%g1\n\t"
			     "casx	[%2], %%g7, %%g1\n\t"
			     "cmp	%%g7, %%g1\n\t"
236
			     "membar	#StoreLoad | #StoreStore\n\t"
L
Linus Torvalds 已提交
237
			     "bne,pn	%%xcc, 1b\n\t"
238
			     " nop\n"
L
Linus Torvalds 已提交
239 240 241
			     "2:"
			     : /* no outputs */
			     : "r" (cpu), "r" (mask), "r" (&page->flags),
242 243
			       "i" (PG_dcache_cpu_mask),
			       "i" (PG_dcache_cpu_shift)
L
Linus Torvalds 已提交
244 245 246 247 248
			     : "g1", "g7");
}

void update_mmu_cache(struct vm_area_struct *vma, unsigned long address, pte_t pte)
{
249
	struct mm_struct *mm;
L
Linus Torvalds 已提交
250 251 252
	struct page *page;
	unsigned long pfn;
	unsigned long pg_flags;
253
	unsigned long mm_rss;
L
Linus Torvalds 已提交
254 255 256 257 258

	pfn = pte_pfn(pte);
	if (pfn_valid(pfn) &&
	    (page = pfn_to_page(pfn), page_mapping(page)) &&
	    ((pg_flags = page->flags) & (1UL << PG_dcache_dirty))) {
259 260
		int cpu = ((pg_flags >> PG_dcache_cpu_shift) &
			   PG_dcache_cpu_mask);
L
Linus Torvalds 已提交
261 262 263 264 265 266 267 268 269 270 271 272 273 274
		int this_cpu = get_cpu();

		/* This is just to optimize away some function calls
		 * in the SMP case.
		 */
		if (cpu == this_cpu)
			flush_dcache_page_impl(page);
		else
			smp_flush_dcache_page_impl(page, cpu);

		clear_dcache_dirty_cpu(page, cpu);

		put_cpu();
	}
275 276 277 278 279

	mm = vma->vm_mm;
	mm_rss = get_mm_rss(mm);
	if (mm_rss >= mm->context.tsb_rss_limit)
		tsb_grow(mm, mm_rss, GFP_ATOMIC);
L
Linus Torvalds 已提交
280 281 282 283
}

void flush_dcache_page(struct page *page)
{
284 285
	struct address_space *mapping;
	int this_cpu;
L
Linus Torvalds 已提交
286

287 288 289 290 291 292 293 294 295 296
	/* Do not bother with the expensive D-cache flush if it
	 * is merely the zero page.  The 'bigcore' testcase in GDB
	 * causes this case to run millions of times.
	 */
	if (page == ZERO_PAGE(0))
		return;

	this_cpu = get_cpu();

	mapping = page_mapping(page);
L
Linus Torvalds 已提交
297
	if (mapping && !mapping_mapped(mapping)) {
298
		int dirty = test_bit(PG_dcache_dirty, &page->flags);
L
Linus Torvalds 已提交
299
		if (dirty) {
300 301
			int dirty_cpu = dcache_dirty_cpu(page);

L
Linus Torvalds 已提交
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
			if (dirty_cpu == this_cpu)
				goto out;
			smp_flush_dcache_page_impl(page, dirty_cpu);
		}
		set_dcache_dirty(page, this_cpu);
	} else {
		/* We could delay the flush for the !page_mapping
		 * case too.  But that case is for exec env/arg
		 * pages and those are %99 certainly going to get
		 * faulted into the tlb (and thus flushed) anyways.
		 */
		flush_dcache_page_impl(page);
	}

out:
	put_cpu();
}

320
void __kprobes flush_icache_range(unsigned long start, unsigned long end)
L
Linus Torvalds 已提交
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
{
	/* Cheetah has coherent I-cache. */
	if (tlb_type == spitfire) {
		unsigned long kaddr;

		for (kaddr = start; kaddr < end; kaddr += PAGE_SIZE)
			__flush_icache_page(__get_phys(kaddr));
	}
}

unsigned long page_to_pfn(struct page *page)
{
	return (unsigned long) ((page - mem_map) + pfn_base);
}

struct page *pfn_to_page(unsigned long pfn)
{
	return (mem_map + (pfn - pfn_base));
}

void show_mem(void)
{
	printk("Mem-info:\n");
	show_free_areas();
	printk("Free swap:       %6ldkB\n",
	       nr_swap_pages << (PAGE_SHIFT-10));
	printk("%ld pages of RAM\n", num_physpages);
	printk("%d free pages\n", nr_free_pages());
}

void mmu_info(struct seq_file *m)
{
	if (tlb_type == cheetah)
		seq_printf(m, "MMU Type\t: Cheetah\n");
	else if (tlb_type == cheetah_plus)
		seq_printf(m, "MMU Type\t: Cheetah+\n");
	else if (tlb_type == spitfire)
		seq_printf(m, "MMU Type\t: Spitfire\n");
	else
		seq_printf(m, "MMU Type\t: ???\n");

#ifdef CONFIG_DEBUG_DCFLUSH
	seq_printf(m, "DCPageFlushes\t: %d\n",
		   atomic_read(&dcpage_flushes));
#ifdef CONFIG_SMP
	seq_printf(m, "DCPageFlushesXC\t: %d\n",
		   atomic_read(&dcpage_flushes_xcall));
#endif /* CONFIG_SMP */
#endif /* CONFIG_DEBUG_DCFLUSH */
}

struct linux_prom_translation {
	unsigned long virt;
	unsigned long size;
	unsigned long data;
};
377 378 379 380 381

/* Exported for kernel TLB miss handling in ktlb.S */
struct linux_prom_translation prom_trans[512] __read_mostly;
unsigned int prom_trans_ents __read_mostly;
unsigned int swapper_pgd_zero __read_mostly;
L
Linus Torvalds 已提交
382 383 384 385 386 387 388 389 390 391 392 393 394 395 396

extern unsigned long prom_boot_page;
extern void prom_remap(unsigned long physpage, unsigned long virtpage, int mmu_ihandle);
extern int prom_get_mmu_ihandle(void);
extern void register_prom_callbacks(void);

/* Exported for SMP bootup purposes. */
unsigned long kern_locked_tte_data;

/*
 * Translate PROM's mapping we capture at boot time into physical address.
 * The second parameter is only set from prom_callback() invocations.
 */
unsigned long prom_virt_to_phys(unsigned long promva, int *error)
{
397
	int i;
398

399 400
	for (i = 0; i < prom_trans_ents; i++) {
		struct linux_prom_translation *p = &prom_trans[i];
401

402 403 404
		if (promva >= p->virt &&
		    promva < (p->virt + p->size)) {
			unsigned long base = p->data & _PAGE_PADDR;
405

406 407 408
			if (error)
				*error = 0;
			return base + (promva & (8192 - 1));
409 410
		}
	}
411 412 413
	if (error)
		*error = 1;
	return 0UL;
414 415
}

416 417
/* The obp translations are saved based on 8k pagesize, since obp can
 * use a mixture of pagesizes. Misses to the LOW_OBP_ADDRESS ->
418
 * HI_OBP_ADDRESS range are handled in ktlb.S.
419
 */
420 421 422 423 424 425
static inline int in_obp_range(unsigned long vaddr)
{
	return (vaddr >= LOW_OBP_ADDRESS &&
		vaddr < HI_OBP_ADDRESS);
}

426
static int cmp_ptrans(const void *a, const void *b)
427
{
428
	const struct linux_prom_translation *x = a, *y = b;
429

430 431 432 433 434
	if (x->virt > y->virt)
		return 1;
	if (x->virt < y->virt)
		return -1;
	return 0;
435 436
}

437
/* Read OBP translations property into 'prom_trans[]'.  */
438
static void __init read_obp_translations(void)
439
{
440
	int n, node, ents, first, last, i;
L
Linus Torvalds 已提交
441 442 443

	node = prom_finddevice("/virtual-memory");
	n = prom_getproplen(node, "translations");
444
	if (unlikely(n == 0 || n == -1)) {
445
		prom_printf("prom_mappings: Couldn't get size.\n");
L
Linus Torvalds 已提交
446 447
		prom_halt();
	}
448 449
	if (unlikely(n > sizeof(prom_trans))) {
		prom_printf("prom_mappings: Size %Zd is too big.\n", n);
L
Linus Torvalds 已提交
450 451
		prom_halt();
	}
452

453
	if ((n = prom_getproperty(node, "translations",
454 455
				  (char *)&prom_trans[0],
				  sizeof(prom_trans))) == -1) {
456
		prom_printf("prom_mappings: Couldn't get property.\n");
L
Linus Torvalds 已提交
457 458
		prom_halt();
	}
459

460
	n = n / sizeof(struct linux_prom_translation);
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
	ents = n;

	sort(prom_trans, ents, sizeof(struct linux_prom_translation),
	     cmp_ptrans, NULL);

	/* Now kick out all the non-OBP entries.  */
	for (i = 0; i < ents; i++) {
		if (in_obp_range(prom_trans[i].virt))
			break;
	}
	first = i;
	for (; i < ents; i++) {
		if (!in_obp_range(prom_trans[i].virt))
			break;
	}
	last = i;

	for (i = 0; i < (last - first); i++) {
		struct linux_prom_translation *src = &prom_trans[i + first];
		struct linux_prom_translation *dest = &prom_trans[i];

		*dest = *src;
	}
	for (; i < ents; i++) {
		struct linux_prom_translation *dest = &prom_trans[i];
		dest->virt = dest->size = dest->data = 0x0UL;
	}

	prom_trans_ents = last - first;

	if (tlb_type == spitfire) {
		/* Clear diag TTE bits. */
		for (i = 0; i < prom_trans_ents; i++)
			prom_trans[i].data &= ~0x0003fe0000000000UL;
	}
497
}
L
Linus Torvalds 已提交
498

499
static void __init remap_kernel(void)
500 501 502 503
{
	unsigned long phys_page, tte_vaddr, tte_data;
	int tlb_ent = sparc64_highest_locked_tlbent();

L
Linus Torvalds 已提交
504
	tte_vaddr = (unsigned long) KERNBASE;
505 506 507 508
	phys_page = (prom_boot_mapping_phys_low >> 22UL) << 22UL;
	tte_data = (phys_page | (_PAGE_VALID | _PAGE_SZ4MB |
				 _PAGE_CP | _PAGE_CV | _PAGE_P |
				 _PAGE_L | _PAGE_W));
L
Linus Torvalds 已提交
509 510 511

	kern_locked_tte_data = tte_data;

512
	/* Now lock us into the TLBs via OBP. */
513 514
	prom_dtlb_load(tlb_ent, tte_data, tte_vaddr);
	prom_itlb_load(tlb_ent, tte_data, tte_vaddr);
L
Linus Torvalds 已提交
515
	if (bigkernel) {
516 517
		tlb_ent -= 1;
		prom_dtlb_load(tlb_ent,
518 519
			       tte_data + 0x400000, 
			       tte_vaddr + 0x400000);
520
		prom_itlb_load(tlb_ent,
521 522
			       tte_data + 0x400000, 
			       tte_vaddr + 0x400000);
L
Linus Torvalds 已提交
523
	}
524 525 526 527 528 529 530
	sparc64_highest_unlocked_tlb_ent = tlb_ent - 1;
	if (tlb_type == cheetah_plus) {
		sparc64_kern_pri_context = (CTX_CHEETAH_PLUS_CTX0 |
					    CTX_CHEETAH_PLUS_NUC);
		sparc64_kern_pri_nuc_bits = CTX_CHEETAH_PLUS_NUC;
		sparc64_kern_sec_context = CTX_CHEETAH_PLUS_CTX0;
	}
531
}
L
Linus Torvalds 已提交
532

533

534
static void __init inherit_prom_mappings(void)
535 536
{
	read_obp_translations();
537 538 539 540

	/* Now fixup OBP's idea about where we really are mapped. */
	prom_printf("Remapping the kernel... ");
	remap_kernel();
L
Linus Torvalds 已提交
541 542
	prom_printf("done.\n");

543
	prom_printf("Registering callbacks... ");
L
Linus Torvalds 已提交
544
	register_prom_callbacks();
545
	prom_printf("done.\n");
L
Linus Torvalds 已提交
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656
}

static int prom_ditlb_set;
struct prom_tlb_entry {
	int		tlb_ent;
	unsigned long	tlb_tag;
	unsigned long	tlb_data;
};
struct prom_tlb_entry prom_itlb[16], prom_dtlb[16];

void prom_world(int enter)
{
	unsigned long pstate;
	int i;

	if (!enter)
		set_fs((mm_segment_t) { get_thread_current_ds() });

	if (!prom_ditlb_set)
		return;

	/* Make sure the following runs atomically. */
	__asm__ __volatile__("flushw\n\t"
			     "rdpr	%%pstate, %0\n\t"
			     "wrpr	%0, %1, %%pstate"
			     : "=r" (pstate)
			     : "i" (PSTATE_IE));

	if (enter) {
		/* Install PROM world. */
		for (i = 0; i < 16; i++) {
			if (prom_dtlb[i].tlb_ent != -1) {
				__asm__ __volatile__("stxa %0, [%1] %2\n\t"
						     "membar #Sync"
					: : "r" (prom_dtlb[i].tlb_tag), "r" (TLB_TAG_ACCESS),
					"i" (ASI_DMMU));
				if (tlb_type == spitfire)
					spitfire_put_dtlb_data(prom_dtlb[i].tlb_ent,
							       prom_dtlb[i].tlb_data);
				else if (tlb_type == cheetah || tlb_type == cheetah_plus)
					cheetah_put_ldtlb_data(prom_dtlb[i].tlb_ent,
							       prom_dtlb[i].tlb_data);
			}
			if (prom_itlb[i].tlb_ent != -1) {
				__asm__ __volatile__("stxa %0, [%1] %2\n\t"
						     "membar #Sync"
						     : : "r" (prom_itlb[i].tlb_tag),
						     "r" (TLB_TAG_ACCESS),
						     "i" (ASI_IMMU));
				if (tlb_type == spitfire)
					spitfire_put_itlb_data(prom_itlb[i].tlb_ent,
							       prom_itlb[i].tlb_data);
				else if (tlb_type == cheetah || tlb_type == cheetah_plus)
					cheetah_put_litlb_data(prom_itlb[i].tlb_ent,
							       prom_itlb[i].tlb_data);
			}
		}
	} else {
		for (i = 0; i < 16; i++) {
			if (prom_dtlb[i].tlb_ent != -1) {
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
					: : "r" (TLB_TAG_ACCESS), "i" (ASI_DMMU));
				if (tlb_type == spitfire)
					spitfire_put_dtlb_data(prom_dtlb[i].tlb_ent, 0x0UL);
				else
					cheetah_put_ldtlb_data(prom_dtlb[i].tlb_ent, 0x0UL);
			}
			if (prom_itlb[i].tlb_ent != -1) {
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : : "r" (TLB_TAG_ACCESS),
						     "i" (ASI_IMMU));
				if (tlb_type == spitfire)
					spitfire_put_itlb_data(prom_itlb[i].tlb_ent, 0x0UL);
				else
					cheetah_put_litlb_data(prom_itlb[i].tlb_ent, 0x0UL);
			}
		}
	}
	__asm__ __volatile__("wrpr	%0, 0, %%pstate"
			     : : "r" (pstate));
}

void inherit_locked_prom_mappings(int save_p)
{
	int i;
	int dtlb_seen = 0;
	int itlb_seen = 0;

	/* Fucking losing PROM has more mappings in the TLB, but
	 * it (conveniently) fails to mention any of these in the
	 * translations property.  The only ones that matter are
	 * the locked PROM tlb entries, so we impose the following
	 * irrecovable rule on the PROM, it is allowed 8 locked
	 * entries in the ITLB and 8 in the DTLB.
	 *
	 * Supposedly the upper 16GB of the address space is
	 * reserved for OBP, BUT I WISH THIS WAS DOCUMENTED
	 * SOMEWHERE!!!!!!!!!!!!!!!!!  Furthermore the entire interface
	 * used between the client program and the firmware on sun5
	 * systems to coordinate mmu mappings is also COMPLETELY
	 * UNDOCUMENTED!!!!!! Thanks S(t)un!
	 */
	if (save_p) {
		for (i = 0; i < 16; i++) {
			prom_itlb[i].tlb_ent = -1;
			prom_dtlb[i].tlb_ent = -1;
		}
	}
	if (tlb_type == spitfire) {
657 658
		int high = sparc64_highest_unlocked_tlb_ent;
		for (i = 0; i <= high; i++) {
L
Linus Torvalds 已提交
659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745
			unsigned long data;

			/* Spitfire Errata #32 workaround */
			/* NOTE: Always runs on spitfire, so no cheetah+
			 *       page size encodings.
			 */
			__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
					     "flush	%%g6"
					     : /* No outputs */
					     : "r" (0),
					     "r" (PRIMARY_CONTEXT), "i" (ASI_DMMU));

			data = spitfire_get_dtlb_data(i);
			if ((data & (_PAGE_L|_PAGE_VALID)) == (_PAGE_L|_PAGE_VALID)) {
				unsigned long tag;

				/* Spitfire Errata #32 workaround */
				/* NOTE: Always runs on spitfire, so no
				 *       cheetah+ page size encodings.
				 */
				__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
						     "flush	%%g6"
						     : /* No outputs */
						     : "r" (0),
						     "r" (PRIMARY_CONTEXT), "i" (ASI_DMMU));

				tag = spitfire_get_dtlb_tag(i);
				if (save_p) {
					prom_dtlb[dtlb_seen].tlb_ent = i;
					prom_dtlb[dtlb_seen].tlb_tag = tag;
					prom_dtlb[dtlb_seen].tlb_data = data;
				}
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : : "r" (TLB_TAG_ACCESS), "i" (ASI_DMMU));
				spitfire_put_dtlb_data(i, 0x0UL);

				dtlb_seen++;
				if (dtlb_seen > 15)
					break;
			}
		}

		for (i = 0; i < high; i++) {
			unsigned long data;

			/* Spitfire Errata #32 workaround */
			/* NOTE: Always runs on spitfire, so no
			 *       cheetah+ page size encodings.
			 */
			__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
					     "flush	%%g6"
					     : /* No outputs */
					     : "r" (0),
					     "r" (PRIMARY_CONTEXT), "i" (ASI_DMMU));

			data = spitfire_get_itlb_data(i);
			if ((data & (_PAGE_L|_PAGE_VALID)) == (_PAGE_L|_PAGE_VALID)) {
				unsigned long tag;

				/* Spitfire Errata #32 workaround */
				/* NOTE: Always runs on spitfire, so no
				 *       cheetah+ page size encodings.
				 */
				__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
						     "flush	%%g6"
						     : /* No outputs */
						     : "r" (0),
						     "r" (PRIMARY_CONTEXT), "i" (ASI_DMMU));

				tag = spitfire_get_itlb_tag(i);
				if (save_p) {
					prom_itlb[itlb_seen].tlb_ent = i;
					prom_itlb[itlb_seen].tlb_tag = tag;
					prom_itlb[itlb_seen].tlb_data = data;
				}
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : : "r" (TLB_TAG_ACCESS), "i" (ASI_IMMU));
				spitfire_put_itlb_data(i, 0x0UL);

				itlb_seen++;
				if (itlb_seen > 15)
					break;
			}
		}
	} else if (tlb_type == cheetah || tlb_type == cheetah_plus) {
746
		int high = sparc64_highest_unlocked_tlb_ent;
L
Linus Torvalds 已提交
747

748
		for (i = 0; i <= high; i++) {
L
Linus Torvalds 已提交
749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068
			unsigned long data;

			data = cheetah_get_ldtlb_data(i);
			if ((data & (_PAGE_L|_PAGE_VALID)) == (_PAGE_L|_PAGE_VALID)) {
				unsigned long tag;

				tag = cheetah_get_ldtlb_tag(i);
				if (save_p) {
					prom_dtlb[dtlb_seen].tlb_ent = i;
					prom_dtlb[dtlb_seen].tlb_tag = tag;
					prom_dtlb[dtlb_seen].tlb_data = data;
				}
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : : "r" (TLB_TAG_ACCESS), "i" (ASI_DMMU));
				cheetah_put_ldtlb_data(i, 0x0UL);

				dtlb_seen++;
				if (dtlb_seen > 15)
					break;
			}
		}

		for (i = 0; i < high; i++) {
			unsigned long data;

			data = cheetah_get_litlb_data(i);
			if ((data & (_PAGE_L|_PAGE_VALID)) == (_PAGE_L|_PAGE_VALID)) {
				unsigned long tag;

				tag = cheetah_get_litlb_tag(i);
				if (save_p) {
					prom_itlb[itlb_seen].tlb_ent = i;
					prom_itlb[itlb_seen].tlb_tag = tag;
					prom_itlb[itlb_seen].tlb_data = data;
				}
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : : "r" (TLB_TAG_ACCESS), "i" (ASI_IMMU));
				cheetah_put_litlb_data(i, 0x0UL);

				itlb_seen++;
				if (itlb_seen > 15)
					break;
			}
		}
	} else {
		/* Implement me :-) */
		BUG();
	}
	if (save_p)
		prom_ditlb_set = 1;
}

/* Give PROM back his world, done during reboots... */
void prom_reload_locked(void)
{
	int i;

	for (i = 0; i < 16; i++) {
		if (prom_dtlb[i].tlb_ent != -1) {
			__asm__ __volatile__("stxa %0, [%1] %2\n\t"
					     "membar #Sync"
				: : "r" (prom_dtlb[i].tlb_tag), "r" (TLB_TAG_ACCESS),
				"i" (ASI_DMMU));
			if (tlb_type == spitfire)
				spitfire_put_dtlb_data(prom_dtlb[i].tlb_ent,
						       prom_dtlb[i].tlb_data);
			else if (tlb_type == cheetah || tlb_type == cheetah_plus)
				cheetah_put_ldtlb_data(prom_dtlb[i].tlb_ent,
						      prom_dtlb[i].tlb_data);
		}

		if (prom_itlb[i].tlb_ent != -1) {
			__asm__ __volatile__("stxa %0, [%1] %2\n\t"
					     "membar #Sync"
					     : : "r" (prom_itlb[i].tlb_tag),
					     "r" (TLB_TAG_ACCESS),
					     "i" (ASI_IMMU));
			if (tlb_type == spitfire)
				spitfire_put_itlb_data(prom_itlb[i].tlb_ent,
						       prom_itlb[i].tlb_data);
			else
				cheetah_put_litlb_data(prom_itlb[i].tlb_ent,
						       prom_itlb[i].tlb_data);
		}
	}
}

#ifdef DCACHE_ALIASING_POSSIBLE
void __flush_dcache_range(unsigned long start, unsigned long end)
{
	unsigned long va;

	if (tlb_type == spitfire) {
		int n = 0;

		for (va = start; va < end; va += 32) {
			spitfire_put_dcache_tag(va & 0x3fe0, 0x0);
			if (++n >= 512)
				break;
		}
	} else {
		start = __pa(start);
		end = __pa(end);
		for (va = start; va < end; va += 32)
			__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
					     "membar #Sync"
					     : /* no outputs */
					     : "r" (va),
					       "i" (ASI_DCACHE_INVALIDATE));
	}
}
#endif /* DCACHE_ALIASING_POSSIBLE */

/* If not locked, zap it. */
void __flush_tlb_all(void)
{
	unsigned long pstate;
	int i;

	__asm__ __volatile__("flushw\n\t"
			     "rdpr	%%pstate, %0\n\t"
			     "wrpr	%0, %1, %%pstate"
			     : "=r" (pstate)
			     : "i" (PSTATE_IE));
	if (tlb_type == spitfire) {
		for (i = 0; i < 64; i++) {
			/* Spitfire Errata #32 workaround */
			/* NOTE: Always runs on spitfire, so no
			 *       cheetah+ page size encodings.
			 */
			__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
					     "flush	%%g6"
					     : /* No outputs */
					     : "r" (0),
					     "r" (PRIMARY_CONTEXT), "i" (ASI_DMMU));

			if (!(spitfire_get_dtlb_data(i) & _PAGE_L)) {
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : /* no outputs */
						     : "r" (TLB_TAG_ACCESS), "i" (ASI_DMMU));
				spitfire_put_dtlb_data(i, 0x0UL);
			}

			/* Spitfire Errata #32 workaround */
			/* NOTE: Always runs on spitfire, so no
			 *       cheetah+ page size encodings.
			 */
			__asm__ __volatile__("stxa	%0, [%1] %2\n\t"
					     "flush	%%g6"
					     : /* No outputs */
					     : "r" (0),
					     "r" (PRIMARY_CONTEXT), "i" (ASI_DMMU));

			if (!(spitfire_get_itlb_data(i) & _PAGE_L)) {
				__asm__ __volatile__("stxa %%g0, [%0] %1\n\t"
						     "membar #Sync"
						     : /* no outputs */
						     : "r" (TLB_TAG_ACCESS), "i" (ASI_IMMU));
				spitfire_put_itlb_data(i, 0x0UL);
			}
		}
	} else if (tlb_type == cheetah || tlb_type == cheetah_plus) {
		cheetah_flush_dtlb_all();
		cheetah_flush_itlb_all();
	}
	__asm__ __volatile__("wrpr	%0, 0, %%pstate"
			     : : "r" (pstate));
}

/* Caller does TLB context flushing on local CPU if necessary.
 * The caller also ensures that CTX_VALID(mm->context) is false.
 *
 * We must be careful about boundary cases so that we never
 * let the user have CTX 0 (nucleus) or we ever use a CTX
 * version of zero (and thus NO_CONTEXT would not be caught
 * by version mis-match tests in mmu_context.h).
 */
void get_new_mmu_context(struct mm_struct *mm)
{
	unsigned long ctx, new_ctx;
	unsigned long orig_pgsz_bits;
	

	spin_lock(&ctx_alloc_lock);
	orig_pgsz_bits = (mm->context.sparc64_ctx_val & CTX_PGSZ_MASK);
	ctx = (tlb_context_cache + 1) & CTX_NR_MASK;
	new_ctx = find_next_zero_bit(mmu_context_bmap, 1 << CTX_NR_BITS, ctx);
	if (new_ctx >= (1 << CTX_NR_BITS)) {
		new_ctx = find_next_zero_bit(mmu_context_bmap, ctx, 1);
		if (new_ctx >= ctx) {
			int i;
			new_ctx = (tlb_context_cache & CTX_VERSION_MASK) +
				CTX_FIRST_VERSION;
			if (new_ctx == 1)
				new_ctx = CTX_FIRST_VERSION;

			/* Don't call memset, for 16 entries that's just
			 * plain silly...
			 */
			mmu_context_bmap[0] = 3;
			mmu_context_bmap[1] = 0;
			mmu_context_bmap[2] = 0;
			mmu_context_bmap[3] = 0;
			for (i = 4; i < CTX_BMAP_SLOTS; i += 4) {
				mmu_context_bmap[i + 0] = 0;
				mmu_context_bmap[i + 1] = 0;
				mmu_context_bmap[i + 2] = 0;
				mmu_context_bmap[i + 3] = 0;
			}
			goto out;
		}
	}
	mmu_context_bmap[new_ctx>>6] |= (1UL << (new_ctx & 63));
	new_ctx |= (tlb_context_cache & CTX_VERSION_MASK);
out:
	tlb_context_cache = new_ctx;
	mm->context.sparc64_ctx_val = new_ctx | orig_pgsz_bits;
	spin_unlock(&ctx_alloc_lock);
}

void sparc_ultra_dump_itlb(void)
{
        int slot;

	if (tlb_type == spitfire) {
		printk ("Contents of itlb: ");
		for (slot = 0; slot < 14; slot++) printk ("    ");
		printk ("%2x:%016lx,%016lx\n",
			0,
			spitfire_get_itlb_tag(0), spitfire_get_itlb_data(0));
		for (slot = 1; slot < 64; slot+=3) {
			printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx %2x:%016lx,%016lx\n", 
				slot,
				spitfire_get_itlb_tag(slot), spitfire_get_itlb_data(slot),
				slot+1,
				spitfire_get_itlb_tag(slot+1), spitfire_get_itlb_data(slot+1),
				slot+2,
				spitfire_get_itlb_tag(slot+2), spitfire_get_itlb_data(slot+2));
		}
	} else if (tlb_type == cheetah || tlb_type == cheetah_plus) {
		printk ("Contents of itlb0:\n");
		for (slot = 0; slot < 16; slot+=2) {
			printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx\n",
				slot,
				cheetah_get_litlb_tag(slot), cheetah_get_litlb_data(slot),
				slot+1,
				cheetah_get_litlb_tag(slot+1), cheetah_get_litlb_data(slot+1));
		}
		printk ("Contents of itlb2:\n");
		for (slot = 0; slot < 128; slot+=2) {
			printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx\n",
				slot,
				cheetah_get_itlb_tag(slot), cheetah_get_itlb_data(slot),
				slot+1,
				cheetah_get_itlb_tag(slot+1), cheetah_get_itlb_data(slot+1));
		}
	}
}

void sparc_ultra_dump_dtlb(void)
{
        int slot;

	if (tlb_type == spitfire) {
		printk ("Contents of dtlb: ");
		for (slot = 0; slot < 14; slot++) printk ("    ");
		printk ("%2x:%016lx,%016lx\n", 0,
			spitfire_get_dtlb_tag(0), spitfire_get_dtlb_data(0));
		for (slot = 1; slot < 64; slot+=3) {
			printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx %2x:%016lx,%016lx\n", 
				slot,
				spitfire_get_dtlb_tag(slot), spitfire_get_dtlb_data(slot),
				slot+1,
				spitfire_get_dtlb_tag(slot+1), spitfire_get_dtlb_data(slot+1),
				slot+2,
				spitfire_get_dtlb_tag(slot+2), spitfire_get_dtlb_data(slot+2));
		}
	} else if (tlb_type == cheetah || tlb_type == cheetah_plus) {
		printk ("Contents of dtlb0:\n");
		for (slot = 0; slot < 16; slot+=2) {
			printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx\n",
				slot,
				cheetah_get_ldtlb_tag(slot), cheetah_get_ldtlb_data(slot),
				slot+1,
				cheetah_get_ldtlb_tag(slot+1), cheetah_get_ldtlb_data(slot+1));
		}
		printk ("Contents of dtlb2:\n");
		for (slot = 0; slot < 512; slot+=2) {
			printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx\n",
				slot,
				cheetah_get_dtlb_tag(slot, 2), cheetah_get_dtlb_data(slot, 2),
				slot+1,
				cheetah_get_dtlb_tag(slot+1, 2), cheetah_get_dtlb_data(slot+1, 2));
		}
		if (tlb_type == cheetah_plus) {
			printk ("Contents of dtlb3:\n");
			for (slot = 0; slot < 512; slot+=2) {
				printk ("%2x:%016lx,%016lx %2x:%016lx,%016lx\n",
					slot,
					cheetah_get_dtlb_tag(slot, 3), cheetah_get_dtlb_data(slot, 3),
					slot+1,
					cheetah_get_dtlb_tag(slot+1, 3), cheetah_get_dtlb_data(slot+1, 3));
			}
		}
	}
}

extern unsigned long cmdline_memory_size;

unsigned long __init bootmem_init(unsigned long *pages_avail)
{
	unsigned long bootmap_size, start_pfn, end_pfn;
	unsigned long end_of_phys_memory = 0UL;
	unsigned long bootmap_pfn, bytes_avail, size;
	int i;

#ifdef CONFIG_DEBUG_BOOTMEM
1069
	prom_printf("bootmem_init: Scan pavail, ");
L
Linus Torvalds 已提交
1070 1071 1072
#endif

	bytes_avail = 0UL;
1073 1074 1075 1076
	for (i = 0; i < pavail_ents; i++) {
		end_of_phys_memory = pavail[i].phys_addr +
			pavail[i].reg_size;
		bytes_avail += pavail[i].reg_size;
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081 1082 1083
		if (cmdline_memory_size) {
			if (bytes_avail > cmdline_memory_size) {
				unsigned long slack = bytes_avail - cmdline_memory_size;

				bytes_avail -= slack;
				end_of_phys_memory -= slack;

1084 1085 1086 1087 1088
				pavail[i].reg_size -= slack;
				if ((long)pavail[i].reg_size <= 0L) {
					pavail[i].phys_addr = 0xdeadbeefUL;
					pavail[i].reg_size = 0UL;
					pavail_ents = i;
L
Linus Torvalds 已提交
1089
				} else {
1090 1091 1092
					pavail[i+1].reg_size = 0Ul;
					pavail[i+1].phys_addr = 0xdeadbeefUL;
					pavail_ents = i + 1;
L
Linus Torvalds 已提交
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145
				}
				break;
			}
		}
	}

	*pages_avail = bytes_avail >> PAGE_SHIFT;

	/* Start with page aligned address of last symbol in kernel
	 * image.  The kernel is hard mapped below PAGE_OFFSET in a
	 * 4MB locked TLB translation.
	 */
	start_pfn = PAGE_ALIGN(kern_base + kern_size) >> PAGE_SHIFT;

	bootmap_pfn = start_pfn;

	end_pfn = end_of_phys_memory >> PAGE_SHIFT;

#ifdef CONFIG_BLK_DEV_INITRD
	/* Now have to check initial ramdisk, so that bootmap does not overwrite it */
	if (sparc_ramdisk_image || sparc_ramdisk_image64) {
		unsigned long ramdisk_image = sparc_ramdisk_image ?
			sparc_ramdisk_image : sparc_ramdisk_image64;
		if (ramdisk_image >= (unsigned long)_end - 2 * PAGE_SIZE)
			ramdisk_image -= KERNBASE;
		initrd_start = ramdisk_image + phys_base;
		initrd_end = initrd_start + sparc_ramdisk_size;
		if (initrd_end > end_of_phys_memory) {
			printk(KERN_CRIT "initrd extends beyond end of memory "
		                 	 "(0x%016lx > 0x%016lx)\ndisabling initrd\n",
			       initrd_end, end_of_phys_memory);
			initrd_start = 0;
		}
		if (initrd_start) {
			if (initrd_start >= (start_pfn << PAGE_SHIFT) &&
			    initrd_start < (start_pfn << PAGE_SHIFT) + 2 * PAGE_SIZE)
				bootmap_pfn = PAGE_ALIGN (initrd_end) >> PAGE_SHIFT;
		}
	}
#endif	
	/* Initialize the boot-time allocator. */
	max_pfn = max_low_pfn = end_pfn;
	min_low_pfn = pfn_base;

#ifdef CONFIG_DEBUG_BOOTMEM
	prom_printf("init_bootmem(min[%lx], bootmap[%lx], max[%lx])\n",
		    min_low_pfn, bootmap_pfn, max_low_pfn);
#endif
	bootmap_size = init_bootmem_node(NODE_DATA(0), bootmap_pfn, pfn_base, end_pfn);

	/* Now register the available physical memory with the
	 * allocator.
	 */
1146
	for (i = 0; i < pavail_ents; i++) {
L
Linus Torvalds 已提交
1147
#ifdef CONFIG_DEBUG_BOOTMEM
1148 1149
		prom_printf("free_bootmem(pavail:%d): base[%lx] size[%lx]\n",
			    i, pavail[i].phys_addr, pavail[i].reg_size);
L
Linus Torvalds 已提交
1150
#endif
1151
		free_bootmem(pavail[i].phys_addr, pavail[i].reg_size);
L
Linus Torvalds 已提交
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 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191
	}

#ifdef CONFIG_BLK_DEV_INITRD
	if (initrd_start) {
		size = initrd_end - initrd_start;

		/* Resert the initrd image area. */
#ifdef CONFIG_DEBUG_BOOTMEM
		prom_printf("reserve_bootmem(initrd): base[%llx] size[%lx]\n",
			initrd_start, initrd_end);
#endif
		reserve_bootmem(initrd_start, size);
		*pages_avail -= PAGE_ALIGN(size) >> PAGE_SHIFT;

		initrd_start += PAGE_OFFSET;
		initrd_end += PAGE_OFFSET;
	}
#endif
	/* Reserve the kernel text/data/bss. */
#ifdef CONFIG_DEBUG_BOOTMEM
	prom_printf("reserve_bootmem(kernel): base[%lx] size[%lx]\n", kern_base, kern_size);
#endif
	reserve_bootmem(kern_base, kern_size);
	*pages_avail -= PAGE_ALIGN(kern_size) >> PAGE_SHIFT;

	/* Reserve the bootmem map.   We do not account for it
	 * in pages_avail because we will release that memory
	 * in free_all_bootmem.
	 */
	size = bootmap_size;
#ifdef CONFIG_DEBUG_BOOTMEM
	prom_printf("reserve_bootmem(bootmap): base[%lx] size[%lx]\n",
		    (bootmap_pfn << PAGE_SHIFT), size);
#endif
	reserve_bootmem((bootmap_pfn << PAGE_SHIFT), size);
	*pages_avail -= PAGE_ALIGN(size) >> PAGE_SHIFT;

	return end_pfn;
}

1192 1193 1194 1195 1196 1197 1198 1199
#ifdef CONFIG_DEBUG_PAGEALLOC
static unsigned long kernel_map_range(unsigned long pstart, unsigned long pend, pgprot_t prot)
{
	unsigned long vstart = PAGE_OFFSET + pstart;
	unsigned long vend = PAGE_OFFSET + pend;
	unsigned long alloc_bytes = 0UL;

	if ((vstart & ~PAGE_MASK) || (vend & ~PAGE_MASK)) {
1200
		prom_printf("kernel_map: Unaligned physmem[%lx:%lx]\n",
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246
			    vstart, vend);
		prom_halt();
	}

	while (vstart < vend) {
		unsigned long this_end, paddr = __pa(vstart);
		pgd_t *pgd = pgd_offset_k(vstart);
		pud_t *pud;
		pmd_t *pmd;
		pte_t *pte;

		pud = pud_offset(pgd, vstart);
		if (pud_none(*pud)) {
			pmd_t *new;

			new = __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, PAGE_SIZE);
			alloc_bytes += PAGE_SIZE;
			pud_populate(&init_mm, pud, new);
		}

		pmd = pmd_offset(pud, vstart);
		if (!pmd_present(*pmd)) {
			pte_t *new;

			new = __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, PAGE_SIZE);
			alloc_bytes += PAGE_SIZE;
			pmd_populate_kernel(&init_mm, pmd, new);
		}

		pte = pte_offset_kernel(pmd, vstart);
		this_end = (vstart + PMD_SIZE) & PMD_MASK;
		if (this_end > vend)
			this_end = vend;

		while (vstart < this_end) {
			pte_val(*pte) = (paddr | pgprot_val(prot));

			vstart += PAGE_SIZE;
			paddr += PAGE_SIZE;
			pte++;
		}
	}

	return alloc_bytes;
}

1247 1248 1249
static struct linux_prom64_registers pall[MAX_BANKS] __initdata;
static int pall_ents __initdata;

1250 1251 1252 1253
extern unsigned int kvmap_linear_patch[1];

static void __init kernel_physical_mapping_init(void)
{
1254
	unsigned long i, mem_alloced = 0UL;
1255

1256 1257 1258
	read_obp_memory("reg", &pall[0], &pall_ents);

	for (i = 0; i < pall_ents; i++) {
1259 1260
		unsigned long phys_start, phys_end;

1261 1262
		phys_start = pall[i].phys_addr;
		phys_end = phys_start + pall[i].reg_size;
1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283
		mem_alloced += kernel_map_range(phys_start, phys_end,
						PAGE_KERNEL);
	}

	printk("Allocated %ld bytes for kernel page tables.\n",
	       mem_alloced);

	kvmap_linear_patch[0] = 0x01000000; /* nop */
	flushi(&kvmap_linear_patch[0]);

	__flush_tlb_all();
}

void kernel_map_pages(struct page *page, int numpages, int enable)
{
	unsigned long phys_start = page_to_pfn(page) << PAGE_SHIFT;
	unsigned long phys_end = phys_start + (numpages * PAGE_SIZE);

	kernel_map_range(phys_start, phys_end,
			 (enable ? PAGE_KERNEL : __pgprot(0)));

1284 1285 1286
	flush_tsb_kernel_range(PAGE_OFFSET + phys_start,
			       PAGE_OFFSET + phys_end);

1287 1288 1289 1290 1291 1292 1293 1294
	/* we should perform an IPI and flush all tlbs,
	 * but that can deadlock->flush only current cpu.
	 */
	__flush_tlb_kernel_range(PAGE_OFFSET + phys_start,
				 PAGE_OFFSET + phys_end);
}
#endif

1295 1296
unsigned long __init find_ecache_flush_span(unsigned long size)
{
1297 1298
	int i;

1299 1300 1301
	for (i = 0; i < pavail_ents; i++) {
		if (pavail[i].reg_size >= size)
			return pavail[i].phys_addr;
1302 1303
	}

1304
	return ~0UL;
1305 1306
}

L
Linus Torvalds 已提交
1307 1308 1309 1310 1311
/* paging_init() sets up the page tables */

extern void cheetah_ecache_flush_init(void);

static unsigned long last_valid_pfn;
1312
pgd_t swapper_pg_dir[2048];
L
Linus Torvalds 已提交
1313 1314 1315

void __init paging_init(void)
{
1316
	unsigned long end_pfn, pages_avail, shift;
1317 1318
	unsigned long real_end, i;

1319 1320
	/* Find available physical memory... */
	read_obp_memory("available", &pavail[0], &pavail_ents);
1321 1322

	phys_base = 0xffffffffffffffffUL;
1323 1324
	for (i = 0; i < pavail_ents; i++)
		phys_base = min(phys_base, pavail[i].phys_addr);
1325 1326 1327 1328 1329

	pfn_base = phys_base >> PAGE_SHIFT;

	kern_base = (prom_boot_mapping_phys_low >> 22UL) << 22UL;
	kern_size = (unsigned long)&_end - (unsigned long)KERNBASE;
L
Linus Torvalds 已提交
1330 1331 1332

	set_bit(0, mmu_context_bmap);

1333 1334
	shift = kern_base + PAGE_OFFSET - ((unsigned long)KERNBASE);

L
Linus Torvalds 已提交
1335 1336 1337
	real_end = (unsigned long)_end;
	if ((real_end > ((unsigned long)KERNBASE + 0x400000)))
		bigkernel = 1;
1338 1339 1340
	if ((real_end > ((unsigned long)KERNBASE + 0x800000))) {
		prom_printf("paging_init: Kernel > 8MB, too large.\n");
		prom_halt();
L
Linus Torvalds 已提交
1341
	}
1342 1343

	/* Set kernel pgd to upper alias so physical page computations
L
Linus Torvalds 已提交
1344 1345 1346 1347
	 * work.
	 */
	init_mm.pgd += ((shift) / (sizeof(pgd_t)));
	
1348
	memset(swapper_low_pmd_dir, 0, sizeof(swapper_low_pmd_dir));
L
Linus Torvalds 已提交
1349 1350 1351

	/* Now can init the kernel/bad page tables. */
	pud_set(pud_offset(&swapper_pg_dir[0], 0),
1352
		swapper_low_pmd_dir + (shift / sizeof(pgd_t)));
L
Linus Torvalds 已提交
1353
	
1354
	swapper_pgd_zero = pgd_val(swapper_pg_dir[0]);
L
Linus Torvalds 已提交
1355
	
1356
	inherit_prom_mappings();
1357
	
L
Linus Torvalds 已提交
1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
	/* Ok, we can use our TLB miss and window trap handlers safely.
	 * We need to do a quick peek here to see if we are on StarFire
	 * or not, so setup_tba can setup the IRQ globals correctly (it
	 * needs to get the hard smp processor id correctly).
	 */
	{
		extern void setup_tba(int);
		setup_tba(this_is_starfire);
	}

1368 1369 1370
	inherit_locked_prom_mappings(1);

	__flush_tlb_all();
1371

1372 1373 1374 1375
	/* Setup bootmem... */
	pages_avail = 0;
	last_valid_pfn = end_pfn = bootmem_init(&pages_avail);

1376 1377 1378 1379
#ifdef CONFIG_DEBUG_PAGEALLOC
	kernel_physical_mapping_init();
#endif

L
Linus Torvalds 已提交
1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403
	{
		unsigned long zones_size[MAX_NR_ZONES];
		unsigned long zholes_size[MAX_NR_ZONES];
		unsigned long npages;
		int znum;

		for (znum = 0; znum < MAX_NR_ZONES; znum++)
			zones_size[znum] = zholes_size[znum] = 0;

		npages = end_pfn - pfn_base;
		zones_size[ZONE_DMA] = npages;
		zholes_size[ZONE_DMA] = npages - pages_avail;

		free_area_init_node(0, &contig_page_data, zones_size,
				    phys_base >> PAGE_SHIFT, zholes_size);
	}

	device_scan();
}

static void __init taint_real_pages(void)
{
	int i;

1404
	read_obp_memory("available", &pavail_rescan[0], &pavail_rescan_ents);
L
Linus Torvalds 已提交
1405

1406
	/* Find changes discovered in the physmem available rescan and
L
Linus Torvalds 已提交
1407 1408
	 * reserve the lost portions in the bootmem maps.
	 */
1409
	for (i = 0; i < pavail_ents; i++) {
L
Linus Torvalds 已提交
1410 1411
		unsigned long old_start, old_end;

1412
		old_start = pavail[i].phys_addr;
L
Linus Torvalds 已提交
1413
		old_end = old_start +
1414
			pavail[i].reg_size;
L
Linus Torvalds 已提交
1415 1416 1417
		while (old_start < old_end) {
			int n;

1418
			for (n = 0; pavail_rescan_ents; n++) {
L
Linus Torvalds 已提交
1419 1420
				unsigned long new_start, new_end;

1421 1422 1423
				new_start = pavail_rescan[n].phys_addr;
				new_end = new_start +
					pavail_rescan[n].reg_size;
L
Linus Torvalds 已提交
1424 1425 1426

				if (new_start <= old_start &&
				    new_end >= (old_start + PAGE_SIZE)) {
1427 1428
					set_bit(old_start >> 22,
						sparc64_valid_addr_bitmap);
L
Linus Torvalds 已提交
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447
					goto do_next_page;
				}
			}
			reserve_bootmem(old_start, PAGE_SIZE);

		do_next_page:
			old_start += PAGE_SIZE;
		}
	}
}

void __init mem_init(void)
{
	unsigned long codepages, datapages, initpages;
	unsigned long addr, last;
	int i;

	i = last_valid_pfn >> ((22 - PAGE_SHIFT) + 6);
	i += 1;
1448
	sparc64_valid_addr_bitmap = (unsigned long *) alloc_bootmem(i << 3);
L
Linus Torvalds 已提交
1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500
	if (sparc64_valid_addr_bitmap == NULL) {
		prom_printf("mem_init: Cannot alloc valid_addr_bitmap.\n");
		prom_halt();
	}
	memset(sparc64_valid_addr_bitmap, 0, i << 3);

	addr = PAGE_OFFSET + kern_base;
	last = PAGE_ALIGN(kern_size) + addr;
	while (addr < last) {
		set_bit(__pa(addr) >> 22, sparc64_valid_addr_bitmap);
		addr += PAGE_SIZE;
	}

	taint_real_pages();

	max_mapnr = last_valid_pfn - pfn_base;
	high_memory = __va(last_valid_pfn << PAGE_SHIFT);

#ifdef CONFIG_DEBUG_BOOTMEM
	prom_printf("mem_init: Calling free_all_bootmem().\n");
#endif
	totalram_pages = num_physpages = free_all_bootmem() - 1;

	/*
	 * Set up the zero page, mark it reserved, so that page count
	 * is not manipulated when freeing the page from user ptes.
	 */
	mem_map_zero = alloc_pages(GFP_KERNEL|__GFP_ZERO, 0);
	if (mem_map_zero == NULL) {
		prom_printf("paging_init: Cannot alloc zero page.\n");
		prom_halt();
	}
	SetPageReserved(mem_map_zero);

	codepages = (((unsigned long) _etext) - ((unsigned long) _start));
	codepages = PAGE_ALIGN(codepages) >> PAGE_SHIFT;
	datapages = (((unsigned long) _edata) - ((unsigned long) _etext));
	datapages = PAGE_ALIGN(datapages) >> PAGE_SHIFT;
	initpages = (((unsigned long) __init_end) - ((unsigned long) __init_begin));
	initpages = PAGE_ALIGN(initpages) >> PAGE_SHIFT;

	printk("Memory: %uk available (%ldk kernel code, %ldk data, %ldk init) [%016lx,%016lx]\n",
	       nr_free_pages() << (PAGE_SHIFT-10),
	       codepages << (PAGE_SHIFT-10),
	       datapages << (PAGE_SHIFT-10), 
	       initpages << (PAGE_SHIFT-10), 
	       PAGE_OFFSET, (last_valid_pfn << PAGE_SHIFT));

	if (tlb_type == cheetah || tlb_type == cheetah_plus)
		cheetah_ecache_flush_init();
}

1501
void free_initmem(void)
L
Linus Torvalds 已提交
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
{
	unsigned long addr, initend;

	/*
	 * The init section is aligned to 8k in vmlinux.lds. Page align for >8k pagesizes.
	 */
	addr = PAGE_ALIGN((unsigned long)(__init_begin));
	initend = (unsigned long)(__init_end) & PAGE_MASK;
	for (; addr < initend; addr += PAGE_SIZE) {
		unsigned long page;
		struct page *p;

		page = (addr +
			((unsigned long) __va(kern_base)) -
			((unsigned long) KERNBASE));
		memset((void *)addr, 0xcc, PAGE_SIZE);
		p = virt_to_page(page);

		ClearPageReserved(p);
		set_page_count(p, 1);
		__free_page(p);
		num_physpages++;
		totalram_pages++;
	}
}

#ifdef CONFIG_BLK_DEV_INITRD
void free_initrd_mem(unsigned long start, unsigned long end)
{
	if (start < end)
		printk ("Freeing initrd memory: %ldk freed\n", (end - start) >> 10);
	for (; start < end; start += PAGE_SIZE) {
		struct page *p = virt_to_page(start);

		ClearPageReserved(p);
		set_page_count(p, 1);
		__free_page(p);
		num_physpages++;
		totalram_pages++;
	}
}
#endif