[net-next-2.6.git] / arch / powerpc / mm / mem.c

/*
 *  PowerPC version
 *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
 *
 *  Modifications by Paul Mackerras (PowerMac) (paulus@cs.anu.edu.au)
 *  and Cort Dougan (PReP) (cort@cs.nmt.edu)
 *    Copyright (C) 1996 Paul Mackerras
 *  PPC44x/36-bit changes by Matt Porter (mporter@mvista.com)
 *
 *  Derived from "arch/i386/mm/init.c"
 *    Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
 *
 *  This program is free software; you can redistribute it and/or
 *  modify it under the terms of the GNU General Public License
 *  as published by the Free Software Foundation; either version
 *  2 of the License, or (at your option) any later version.
 *
 */

#include <linux/module.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/gfp.h>
#include <linux/types.h>
#include <linux/mm.h>
#include <linux/stddef.h>
#include <linux/init.h>
#include <linux/bootmem.h>
#include <linux/highmem.h>
#include <linux/initrd.h>
#include <linux/pagemap.h>
#include <linux/suspend.h>
#include <linux/memblock.h>
#include <linux/hugetlb.h>

#include <asm/pgalloc.h>
#include <asm/prom.h>
#include <asm/io.h>
#include <asm/mmu_context.h>
#include <asm/pgtable.h>
#include <asm/mmu.h>
#include <asm/smp.h>
#include <asm/machdep.h>
#include <asm/btext.h>
#include <asm/tlb.h>
#include <asm/sections.h>
#include <asm/sparsemem.h>
#include <asm/vdso.h>
#include <asm/fixmap.h>
#include <asm/swiotlb.h>

#include "mmu_decl.h"

#ifndef CPU_FTR_COHERENT_ICACHE
#define CPU_FTR_COHERENT_ICACHE	0	/* XXX for now */
#define CPU_FTR_NOEXECUTE	0
#endif

int init_bootmem_done;
int mem_init_done;
phys_addr_t memory_limit;

#ifdef CONFIG_HIGHMEM
pte_t *kmap_pte;
pgprot_t kmap_prot;

EXPORT_SYMBOL(kmap_prot);
EXPORT_SYMBOL(kmap_pte);

static inline pte_t *virt_to_kpte(unsigned long vaddr)
{
	return pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr),
			vaddr), vaddr), vaddr);
}
#endif

int page_is_ram(unsigned long pfn)
{
#ifndef CONFIG_PPC64	/* XXX for now */
	return pfn < max_pfn;
#else
	unsigned long paddr = (pfn << PAGE_SHIFT);
	int i;
	for (i=0; i < memblock.memory.cnt; i++) {
		unsigned long base;

		base = memblock.memory.region[i].base;

		if ((paddr >= base) &&
			(paddr < (base + memblock.memory.region[i].size))) {
			return 1;
		}
	}

	return 0;
#endif
}

pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
			      unsigned long size, pgprot_t vma_prot)
{
	if (ppc_md.phys_mem_access_prot)
		return ppc_md.phys_mem_access_prot(file, pfn, size, vma_prot);

	if (!page_is_ram(pfn))
		vma_prot = pgprot_noncached(vma_prot);

	return vma_prot;
}
EXPORT_SYMBOL(phys_mem_access_prot);

#ifdef CONFIG_MEMORY_HOTPLUG

#ifdef CONFIG_NUMA
int memory_add_physaddr_to_nid(u64 start)
{
	return hot_add_scn_to_nid(start);
}
#endif

int arch_add_memory(int nid, u64 start, u64 size)
{
	struct pglist_data *pgdata;
	struct zone *zone;
	unsigned long start_pfn = start >> PAGE_SHIFT;
	unsigned long nr_pages = size >> PAGE_SHIFT;

	pgdata = NODE_DATA(nid);

	start = (unsigned long)__va(start);
	create_section_mapping(start, start + size);

	/* this should work for most non-highmem platforms */
	zone = pgdata->node_zones;

	return __add_pages(nid, zone, start_pfn, nr_pages);
}
#endif /* CONFIG_MEMORY_HOTPLUG */

/*
 * walk_memory_resource() needs to make sure there is no holes in a given
 * memory range.  PPC64 does not maintain the memory layout in /proc/iomem.
 * Instead it maintains it in memblock.memory structures.  Walk through the
 * memory regions, find holes and callback for contiguous regions.
 */
int
walk_system_ram_range(unsigned long start_pfn, unsigned long nr_pages,
		void *arg, int (*func)(unsigned long, unsigned long, void *))
{
	struct memblock_property res;
	unsigned long pfn, len;
	u64 end;
	int ret = -1;

	res.base = (u64) start_pfn << PAGE_SHIFT;
	res.size = (u64) nr_pages << PAGE_SHIFT;

	end = res.base + res.size - 1;
	while ((res.base < end) && (memblock_find(&res) >= 0)) {
		pfn = (unsigned long)(res.base >> PAGE_SHIFT);
		len = (unsigned long)(res.size >> PAGE_SHIFT);
		ret = (*func)(pfn, len, arg);
		if (ret)
			break;
		res.base += (res.size + 1);
		res.size = (end - res.base + 1);
	}
	return ret;
}
EXPORT_SYMBOL_GPL(walk_system_ram_range);

/*
 * Initialize the bootmem system and give it all the memory we
 * have available.  If we are using highmem, we only put the
 * lowmem into the bootmem system.
 */
#ifndef CONFIG_NEED_MULTIPLE_NODES
void __init do_init_bootmem(void)
{
	unsigned long i;
	unsigned long start, bootmap_pages;
	unsigned long total_pages;
	int boot_mapsize;

	max_low_pfn = max_pfn = memblock_end_of_DRAM() >> PAGE_SHIFT;
	total_pages = (memblock_end_of_DRAM() - memstart_addr) >> PAGE_SHIFT;
#ifdef CONFIG_HIGHMEM
	total_pages = total_lowmem >> PAGE_SHIFT;
	max_low_pfn = lowmem_end_addr >> PAGE_SHIFT;
#endif

	/*
	 * Find an area to use for the bootmem bitmap.  Calculate the size of
	 * bitmap required as (Total Memory) / PAGE_SIZE / BITS_PER_BYTE.
	 * Add 1 additional page in case the address isn't page-aligned.
	 */
	bootmap_pages = bootmem_bootmap_pages(total_pages);

	start = memblock_alloc(bootmap_pages << PAGE_SHIFT, PAGE_SIZE);

	min_low_pfn = MEMORY_START >> PAGE_SHIFT;
	boot_mapsize = init_bootmem_node(NODE_DATA(0), start >> PAGE_SHIFT, min_low_pfn, max_low_pfn);

	/* Add active regions with valid PFNs */
	for (i = 0; i < memblock.memory.cnt; i++) {
		unsigned long start_pfn, end_pfn;
		start_pfn = memblock.memory.region[i].base >> PAGE_SHIFT;
		end_pfn = start_pfn + memblock_size_pages(&memblock.memory, i);
		add_active_range(0, start_pfn, end_pfn);
	}

	/* Add all physical memory to the bootmem map, mark each area
	 * present.
	 */
#ifdef CONFIG_HIGHMEM
	free_bootmem_with_active_regions(0, lowmem_end_addr >> PAGE_SHIFT);

	/* reserve the sections we're already using */
	for (i = 0; i < memblock.reserved.cnt; i++) {
		unsigned long addr = memblock.reserved.region[i].base +
				     memblock_size_bytes(&memblock.reserved, i) - 1;
		if (addr < lowmem_end_addr)
			reserve_bootmem(memblock.reserved.region[i].base,
					memblock_size_bytes(&memblock.reserved, i),
					BOOTMEM_DEFAULT);
		else if (memblock.reserved.region[i].base < lowmem_end_addr) {
			unsigned long adjusted_size = lowmem_end_addr -
				      memblock.reserved.region[i].base;
			reserve_bootmem(memblock.reserved.region[i].base,
					adjusted_size, BOOTMEM_DEFAULT);
		}
	}
#else
	free_bootmem_with_active_regions(0, max_pfn);

	/* reserve the sections we're already using */
	for (i = 0; i < memblock.reserved.cnt; i++)
		reserve_bootmem(memblock.reserved.region[i].base,
				memblock_size_bytes(&memblock.reserved, i),
				BOOTMEM_DEFAULT);

#endif
	/* XXX need to clip this if using highmem? */
	sparse_memory_present_with_active_regions(0);

	init_bootmem_done = 1;
}

/* mark pages that don't exist as nosave */
static int __init mark_nonram_nosave(void)
{
	unsigned long memblock_next_region_start_pfn,
		      memblock_region_max_pfn;
	int i;

	for (i = 0; i < memblock.memory.cnt - 1; i++) {
		memblock_region_max_pfn =
			(memblock.memory.region[i].base >> PAGE_SHIFT) +
			(memblock.memory.region[i].size >> PAGE_SHIFT);
		memblock_next_region_start_pfn =
			memblock.memory.region[i+1].base >> PAGE_SHIFT;

		if (memblock_region_max_pfn < memblock_next_region_start_pfn)
			register_nosave_region(memblock_region_max_pfn,
					       memblock_next_region_start_pfn);
	}

	return 0;
}

/*
 * paging_init() sets up the page tables - in fact we've already done this.
 */
void __init paging_init(void)
{
	unsigned long total_ram = memblock_phys_mem_size();
	phys_addr_t top_of_ram = memblock_end_of_DRAM();
	unsigned long max_zone_pfns[MAX_NR_ZONES];

#ifdef CONFIG_PPC32
	unsigned long v = __fix_to_virt(__end_of_fixed_addresses - 1);
	unsigned long end = __fix_to_virt(FIX_HOLE);

	for (; v < end; v += PAGE_SIZE)
		map_page(v, 0, 0); /* XXX gross */
#endif

#ifdef CONFIG_HIGHMEM
	map_page(PKMAP_BASE, 0, 0);	/* XXX gross */
	pkmap_page_table = virt_to_kpte(PKMAP_BASE);

	kmap_pte = virt_to_kpte(__fix_to_virt(FIX_KMAP_BEGIN));
	kmap_prot = PAGE_KERNEL;
#endif /* CONFIG_HIGHMEM */

	printk(KERN_DEBUG "Top of RAM: 0x%llx, Total RAM: 0x%lx\n",
	       (unsigned long long)top_of_ram, total_ram);
	printk(KERN_DEBUG "Memory hole size: %ldMB\n",
	       (long int)((top_of_ram - total_ram) >> 20));
	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
#ifdef CONFIG_HIGHMEM
	max_zone_pfns[ZONE_DMA] = lowmem_end_addr >> PAGE_SHIFT;
	max_zone_pfns[ZONE_HIGHMEM] = top_of_ram >> PAGE_SHIFT;
#else
	max_zone_pfns[ZONE_DMA] = top_of_ram >> PAGE_SHIFT;
#endif
	free_area_init_nodes(max_zone_pfns);

	mark_nonram_nosave();
}
#endif /* ! CONFIG_NEED_MULTIPLE_NODES */

void __init mem_init(void)
{
#ifdef CONFIG_NEED_MULTIPLE_NODES
	int nid;
#endif
	pg_data_t *pgdat;
	unsigned long i;
	struct page *page;
	unsigned long reservedpages = 0, codesize, initsize, datasize, bsssize;

#ifdef CONFIG_SWIOTLB
	if (ppc_swiotlb_enable)
		swiotlb_init(1);
#endif

	num_physpages = memblock.memory.size >> PAGE_SHIFT;
	high_memory = (void *) __va(max_low_pfn * PAGE_SIZE);

#ifdef CONFIG_NEED_MULTIPLE_NODES
        for_each_online_node(nid) {
		if (NODE_DATA(nid)->node_spanned_pages != 0) {
			printk("freeing bootmem node %d\n", nid);
			totalram_pages +=
				free_all_bootmem_node(NODE_DATA(nid));
		}
	}
#else
	max_mapnr = max_pfn;
	totalram_pages += free_all_bootmem();
#endif
	for_each_online_pgdat(pgdat) {
		for (i = 0; i < pgdat->node_spanned_pages; i++) {
			if (!pfn_valid(pgdat->node_start_pfn + i))
				continue;
			page = pgdat_page_nr(pgdat, i);
			if (PageReserved(page))
				reservedpages++;
		}
	}

	codesize = (unsigned long)&_sdata - (unsigned long)&_stext;
	datasize = (unsigned long)&_edata - (unsigned long)&_sdata;
	initsize = (unsigned long)&__init_end - (unsigned long)&__init_begin;
	bsssize = (unsigned long)&__bss_stop - (unsigned long)&__bss_start;

#ifdef CONFIG_HIGHMEM
	{
		unsigned long pfn, highmem_mapnr;

		highmem_mapnr = lowmem_end_addr >> PAGE_SHIFT;
		for (pfn = highmem_mapnr; pfn < max_mapnr; ++pfn) {
			struct page *page = pfn_to_page(pfn);
			if (memblock_is_reserved(pfn << PAGE_SHIFT))
				continue;
			ClearPageReserved(page);
			init_page_count(page);
			__free_page(page);
			totalhigh_pages++;
			reservedpages--;
		}
		totalram_pages += totalhigh_pages;
		printk(KERN_DEBUG "High memory: %luk\n",
		       totalhigh_pages << (PAGE_SHIFT-10));
	}
#endif /* CONFIG_HIGHMEM */

	printk(KERN_INFO "Memory: %luk/%luk available (%luk kernel code, "
	       "%luk reserved, %luk data, %luk bss, %luk init)\n",
		nr_free_pages() << (PAGE_SHIFT-10),
		num_physpages << (PAGE_SHIFT-10),
		codesize >> 10,
		reservedpages << (PAGE_SHIFT-10),
		datasize >> 10,
		bsssize >> 10,
		initsize >> 10);

#ifdef CONFIG_PPC32
	pr_info("Kernel virtual memory layout:\n");
	pr_info("  * 0x%08lx..0x%08lx  : fixmap\n", FIXADDR_START, FIXADDR_TOP);
#ifdef CONFIG_HIGHMEM
	pr_info("  * 0x%08lx..0x%08lx  : highmem PTEs\n",
		PKMAP_BASE, PKMAP_ADDR(LAST_PKMAP));
#endif /* CONFIG_HIGHMEM */
#ifdef CONFIG_NOT_COHERENT_CACHE
	pr_info("  * 0x%08lx..0x%08lx  : consistent mem\n",
		IOREMAP_TOP, IOREMAP_TOP + CONFIG_CONSISTENT_SIZE);
#endif /* CONFIG_NOT_COHERENT_CACHE */
	pr_info("  * 0x%08lx..0x%08lx  : early ioremap\n",
		ioremap_bot, IOREMAP_TOP);
	pr_info("  * 0x%08lx..0x%08lx  : vmalloc & ioremap\n",
		VMALLOC_START, VMALLOC_END);
#endif /* CONFIG_PPC32 */

	mem_init_done = 1;
}

/*
 * This is called when a page has been modified by the kernel.
 * It just marks the page as not i-cache clean.  We do the i-cache
 * flush later when the page is given to a user process, if necessary.
 */
void flush_dcache_page(struct page *page)
{
	if (cpu_has_feature(CPU_FTR_COHERENT_ICACHE))
		return;
	/* avoid an atomic op if possible */
	if (test_bit(PG_arch_1, &page->flags))
		clear_bit(PG_arch_1, &page->flags);
}
EXPORT_SYMBOL(flush_dcache_page);

void flush_dcache_icache_page(struct page *page)
{
#ifdef CONFIG_HUGETLB_PAGE
	if (PageCompound(page)) {
		flush_dcache_icache_hugepage(page);
		return;
	}
#endif
#ifdef CONFIG_BOOKE
	{
		void *start = kmap_atomic(page, KM_PPC_SYNC_ICACHE);
		__flush_dcache_icache(start);
		kunmap_atomic(start, KM_PPC_SYNC_ICACHE);
	}
#elif defined(CONFIG_8xx) || defined(CONFIG_PPC64)
	/* On 8xx there is no need to kmap since highmem is not supported */
	__flush_dcache_icache(page_address(page)); 
#else
	__flush_dcache_icache_phys(page_to_pfn(page) << PAGE_SHIFT);
#endif
}

void clear_user_page(void *page, unsigned long vaddr, struct page *pg)
{
	clear_page(page);

	/*
	 * We shouldnt have to do this, but some versions of glibc
	 * require it (ld.so assumes zero filled pages are icache clean)
	 * - Anton
	 */
	flush_dcache_page(pg);
}
EXPORT_SYMBOL(clear_user_page);

void copy_user_page(void *vto, void *vfrom, unsigned long vaddr,
		    struct page *pg)
{
	copy_page(vto, vfrom);

	/*
	 * We should be able to use the following optimisation, however
	 * there are two problems.
	 * Firstly a bug in some versions of binutils meant PLT sections
	 * were not marked executable.
	 * Secondly the first word in the GOT section is blrl, used
	 * to establish the GOT address. Until recently the GOT was
	 * not marked executable.
	 * - Anton
	 */
#if 0
	if (!vma->vm_file && ((vma->vm_flags & VM_EXEC) == 0))
		return;
#endif

	flush_dcache_page(pg);
}

void flush_icache_user_range(struct vm_area_struct *vma, struct page *page,
			     unsigned long addr, int len)
{
	unsigned long maddr;

	maddr = (unsigned long) kmap(page) + (addr & ~PAGE_MASK);
	flush_icache_range(maddr, maddr + len);
	kunmap(page);
}
EXPORT_SYMBOL(flush_icache_user_range);

/*
 * This is called at the end of handling a user page fault, when the
 * fault has been handled by updating a PTE in the linux page tables.
 * We use it to preload an HPTE into the hash table corresponding to
 * the updated linux PTE.
 * 
 * This must always be called with the pte lock held.
 */
void update_mmu_cache(struct vm_area_struct *vma, unsigned long address,
		      pte_t *ptep)
{
#ifdef CONFIG_PPC_STD_MMU
	unsigned long access = 0, trap;

	/* We only want HPTEs for linux PTEs that have _PAGE_ACCESSED set */
	if (!pte_young(*ptep) || address >= TASK_SIZE)
		return;

	/* We try to figure out if we are coming from an instruction
	 * access fault and pass that down to __hash_page so we avoid
	 * double-faulting on execution of fresh text. We have to test
	 * for regs NULL since init will get here first thing at boot
	 *
	 * We also avoid filling the hash if not coming from a fault
	 */
	if (current->thread.regs == NULL)
		return;
	trap = TRAP(current->thread.regs);
	if (trap == 0x400)
		access |= _PAGE_EXEC;
	else if (trap != 0x300)
		return;
	hash_preload(vma->vm_mm, address, access, trap);
#endif /* CONFIG_PPC_STD_MMU */
}
Commit	Line	Data
14cf11af PM	1	/*
	2	* PowerPC version
	3	* Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
	4	*
	5	* Modifications by Paul Mackerras (PowerMac) (paulus@cs.anu.edu.au)
	6	* and Cort Dougan (PReP) (cort@cs.nmt.edu)
	7	* Copyright (C) 1996 Paul Mackerras
14cf11af PM	8	* PPC44x/36-bit changes by Matt Porter (mporter@mvista.com)
	9	*
	10	* Derived from "arch/i386/mm/init.c"
	11	* Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds
	12	*
	13	* This program is free software; you can redistribute it and/or
	14	* modify it under the terms of the GNU General Public License
	15	* as published by the Free Software Foundation; either version
	16	* 2 of the License, or (at your option) any later version.
	17	*
	18	*/
	19
14cf11af PM	20	#include <linux/module.h>
	21	#include <linux/sched.h>
	22	#include <linux/kernel.h>
	23	#include <linux/errno.h>
	24	#include <linux/string.h>
5a0e3ad6	25	#include <linux/gfp.h>
14cf11af PM	26	#include <linux/types.h>
	27	#include <linux/mm.h>
	28	#include <linux/stddef.h>
	29	#include <linux/init.h>
	30	#include <linux/bootmem.h>
	31	#include <linux/highmem.h>
	32	#include <linux/initrd.h>
	33	#include <linux/pagemap.h>
4e8ad3e8	34	#include <linux/suspend.h>
95f72d1e	35	#include <linux/memblock.h>
0895ecda	36	#include <linux/hugetlb.h>
14cf11af PM	37
	38	#include <asm/pgalloc.h>
	39	#include <asm/prom.h>
	40	#include <asm/io.h>
	41	#include <asm/mmu_context.h>
	42	#include <asm/pgtable.h>
	43	#include <asm/mmu.h>
	44	#include <asm/smp.h>
	45	#include <asm/machdep.h>
	46	#include <asm/btext.h>
	47	#include <asm/tlb.h>
7c8c6b97	48	#include <asm/sections.h>
db7f37de	49	#include <asm/sparsemem.h>
ab1f9dac	50	#include <asm/vdso.h>
2c419bde	51	#include <asm/fixmap.h>
a9327296	52	#include <asm/swiotlb.h>
14cf11af	53
14cf11af PM	54	#include "mmu_decl.h"
	55
	56	#ifndef CPU_FTR_COHERENT_ICACHE
	57	#define CPU_FTR_COHERENT_ICACHE 0 /* XXX for now */
	58	#define CPU_FTR_NOEXECUTE 0
	59	#endif
	60
7c8c6b97 PM	61	int init_bootmem_done;
7c8c6b97 PM	62	int mem_init_done;
49a84965	63	phys_addr_t memory_limit;
7c8c6b97	64
2c419bde KG	65	#ifdef CONFIG_HIGHMEM
	66	pte_t *kmap_pte;
	67	pgprot_t kmap_prot;
	68
	69	EXPORT_SYMBOL(kmap_prot);
	70	EXPORT_SYMBOL(kmap_pte);
	71
	72	static inline pte_t *virt_to_kpte(unsigned long vaddr)
	73	{
	74	return pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr),
	75	vaddr), vaddr), vaddr);
	76	}
	77	#endif
	78
14cf11af PM	79	int page_is_ram(unsigned long pfn)
14cf11af PM	80	{
14cf11af	81	#ifndef CONFIG_PPC64 /* XXX for now */
a880e762	82	return pfn < max_pfn;
14cf11af	83	#else
a880e762	84	unsigned long paddr = (pfn << PAGE_SHIFT);
14cf11af	85	int i;
95f72d1e	86	for (i=0; i < memblock.memory.cnt; i++) {
14cf11af PM	87	unsigned long base;
14cf11af PM	88
95f72d1e	89	base = memblock.memory.region[i].base;
14cf11af PM	90
14cf11af PM	91	if ((paddr >= base) &&
95f72d1e	92	(paddr < (base + memblock.memory.region[i].size))) {
14cf11af PM	93	return 1;
	94	}
	95	}
	96
	97	return 0;
	98	#endif
	99	}
14cf11af	100
8b150478	101	pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
14cf11af PM	102	unsigned long size, pgprot_t vma_prot)
	103	{
	104	if (ppc_md.phys_mem_access_prot)
8b150478	105	return ppc_md.phys_mem_access_prot(file, pfn, size, vma_prot);
14cf11af	106
8b150478	107	if (!page_is_ram(pfn))
64b3d0e8 BH	108	vma_prot = pgprot_noncached(vma_prot);
64b3d0e8 BH	109
14cf11af PM	110	return vma_prot;
	111	}
	112	EXPORT_SYMBOL(phys_mem_access_prot);
	113
23fd0775 PM	114	#ifdef CONFIG_MEMORY_HOTPLUG
23fd0775 PM	115
bc02af93 YG	116	#ifdef CONFIG_NUMA
	117	int memory_add_physaddr_to_nid(u64 start)
	118	{
	119	return hot_add_scn_to_nid(start);
	120	}
	121	#endif
	122
fa90f70a	123	int arch_add_memory(int nid, u64 start, u64 size)
23fd0775	124	{
237a0989	125	struct pglist_data *pgdata;
23fd0775 PM	126	struct zone *zone;
	127	unsigned long start_pfn = start >> PAGE_SHIFT;
	128	unsigned long nr_pages = size >> PAGE_SHIFT;
	129
237a0989 MK	130	pgdata = NODE_DATA(nid);
237a0989 MK	131
2d0eee14	132	start = (unsigned long)__va(start);
54b79248 MK	133	create_section_mapping(start, start + size);
54b79248 MK	134
23fd0775 PM	135	/* this should work for most non-highmem platforms */
	136	zone = pgdata->node_zones;
	137
c04fc586	138	return __add_pages(nid, zone, start_pfn, nr_pages);
23fd0775	139	}
0d579944	140	#endif /* CONFIG_MEMORY_HOTPLUG */
a99824f3 BP	141
	142	/*
	143	* walk_memory_resource() needs to make sure there is no holes in a given
9d88a2eb	144	* memory range. PPC64 does not maintain the memory layout in /proc/iomem.
95f72d1e	145	* Instead it maintains it in memblock.memory structures. Walk through the
9d88a2eb	146	* memory regions, find holes and callback for contiguous regions.
a99824f3 BP	147	*/
a99824f3 BP	148	int
908eedc6 KH	149	walk_system_ram_range(unsigned long start_pfn, unsigned long nr_pages,
908eedc6 KH	150	void arg, int (func)(unsigned long, unsigned long, void *))
a99824f3	151	{
95f72d1e	152	struct memblock_property res;
9d88a2eb BP	153	unsigned long pfn, len;
	154	u64 end;
	155	int ret = -1;
	156
	157	res.base = (u64) start_pfn << PAGE_SHIFT;
	158	res.size = (u64) nr_pages << PAGE_SHIFT;
	159
	160	end = res.base + res.size - 1;
95f72d1e	161	while ((res.base < end) && (memblock_find(&res) >= 0)) {
9d88a2eb BP	162	pfn = (unsigned long)(res.base >> PAGE_SHIFT);
	163	len = (unsigned long)(res.size >> PAGE_SHIFT);
	164	ret = (*func)(pfn, len, arg);
	165	if (ret)
	166	break;
	167	res.base += (res.size + 1);
	168	res.size = (end - res.base + 1);
	169	}
	170	return ret;
a99824f3	171	}
908eedc6	172	EXPORT_SYMBOL_GPL(walk_system_ram_range);
a99824f3	173
7c8c6b97 PM	174	/*
	175	* Initialize the bootmem system and give it all the memory we
	176	* have available. If we are using highmem, we only put the
	177	* lowmem into the bootmem system.
	178	*/
	179	#ifndef CONFIG_NEED_MULTIPLE_NODES
	180	void __init do_init_bootmem(void)
	181	{
	182	unsigned long i;
	183	unsigned long start, bootmap_pages;
	184	unsigned long total_pages;
	185	int boot_mapsize;
	186
95f72d1e YL	187	max_low_pfn = max_pfn = memblock_end_of_DRAM() >> PAGE_SHIFT;
95f72d1e YL	188	total_pages = (memblock_end_of_DRAM() - memstart_addr) >> PAGE_SHIFT;
7c8c6b97 PM	189	#ifdef CONFIG_HIGHMEM
7c8c6b97 PM	190	total_pages = total_lowmem >> PAGE_SHIFT;
d7917ba7	191	max_low_pfn = lowmem_end_addr >> PAGE_SHIFT;
7c8c6b97 PM	192	#endif
	193
	194	/*
	195	* Find an area to use for the bootmem bitmap. Calculate the size of
	196	* bitmap required as (Total Memory) / PAGE_SIZE / BITS_PER_BYTE.
	197	* Add 1 additional page in case the address isn't page-aligned.
	198	*/
	199	bootmap_pages = bootmem_bootmap_pages(total_pages);
	200
95f72d1e	201	start = memblock_alloc(bootmap_pages << PAGE_SHIFT, PAGE_SIZE);
7c8c6b97	202
37dd2bad KG	203	min_low_pfn = MEMORY_START >> PAGE_SHIFT;
37dd2bad KG	204	boot_mapsize = init_bootmem_node(NODE_DATA(0), start >> PAGE_SHIFT, min_low_pfn, max_low_pfn);
7c8c6b97	205
c67c3cb4	206	/* Add active regions with valid PFNs */
95f72d1e	207	for (i = 0; i < memblock.memory.cnt; i++) {
c67c3cb4	208	unsigned long start_pfn, end_pfn;
95f72d1e YL	209	start_pfn = memblock.memory.region[i].base >> PAGE_SHIFT;
95f72d1e YL	210	end_pfn = start_pfn + memblock_size_pages(&memblock.memory, i);
c67c3cb4 MG	211	add_active_range(0, start_pfn, end_pfn);
	212	}
	213
7c8c6b97 PM	214	/* Add all physical memory to the bootmem map, mark each area
	215	* present.
	216	*/
7c8c6b97	217	#ifdef CONFIG_HIGHMEM
d7917ba7	218	free_bootmem_with_active_regions(0, lowmem_end_addr >> PAGE_SHIFT);
f98eeb4e KG	219
f98eeb4e KG	220	/* reserve the sections we're already using */
95f72d1e YL	221	for (i = 0; i < memblock.reserved.cnt; i++) {
	222	unsigned long addr = memblock.reserved.region[i].base +
	223	memblock_size_bytes(&memblock.reserved, i) - 1;
d7917ba7	224	if (addr < lowmem_end_addr)
95f72d1e YL	225	reserve_bootmem(memblock.reserved.region[i].base,
95f72d1e YL	226	memblock_size_bytes(&memblock.reserved, i),
72a7fe39	227	BOOTMEM_DEFAULT);
95f72d1e	228	else if (memblock.reserved.region[i].base < lowmem_end_addr) {
d7917ba7	229	unsigned long adjusted_size = lowmem_end_addr -
95f72d1e YL	230	memblock.reserved.region[i].base;
95f72d1e YL	231	reserve_bootmem(memblock.reserved.region[i].base,
72a7fe39	232	adjusted_size, BOOTMEM_DEFAULT);
f98eeb4e KG	233	}
f98eeb4e KG	234	}
c67c3cb4 MG	235	#else
c67c3cb4 MG	236	free_bootmem_with_active_regions(0, max_pfn);
7c8c6b97 PM	237
7c8c6b97 PM	238	/* reserve the sections we're already using */
95f72d1e YL	239	for (i = 0; i < memblock.reserved.cnt; i++)
	240	reserve_bootmem(memblock.reserved.region[i].base,
	241	memblock_size_bytes(&memblock.reserved, i),
72a7fe39	242	BOOTMEM_DEFAULT);
7c8c6b97	243
f98eeb4e	244	#endif
7c8c6b97	245	/* XXX need to clip this if using highmem? */
c67c3cb4 MG	246	sparse_memory_present_with_active_regions(0);
c67c3cb4 MG	247
7c8c6b97 PM	248	init_bootmem_done = 1;
	249	}
	250
4e8ad3e8 JB	251	/* mark pages that don't exist as nosave */
	252	static int __init mark_nonram_nosave(void)
	253	{
95f72d1e YL	254	unsigned long memblock_next_region_start_pfn,
95f72d1e YL	255	memblock_region_max_pfn;
4e8ad3e8 JB	256	int i;
4e8ad3e8 JB	257
95f72d1e YL	258	for (i = 0; i < memblock.memory.cnt - 1; i++) {
	259	memblock_region_max_pfn =
	260	(memblock.memory.region[i].base >> PAGE_SHIFT) +
	261	(memblock.memory.region[i].size >> PAGE_SHIFT);
	262	memblock_next_region_start_pfn =
	263	memblock.memory.region[i+1].base >> PAGE_SHIFT;
4e8ad3e8	264
95f72d1e YL	265	if (memblock_region_max_pfn < memblock_next_region_start_pfn)
	266	register_nosave_region(memblock_region_max_pfn,
	267	memblock_next_region_start_pfn);
4e8ad3e8 JB	268	}
	269
	270	return 0;
	271	}
	272
7c8c6b97 PM	273	/*
	274	* paging_init() sets up the page tables - in fact we've already done this.
	275	*/
	276	void __init paging_init(void)
	277	{
95f72d1e YL	278	unsigned long total_ram = memblock_phys_mem_size();
95f72d1e YL	279	phys_addr_t top_of_ram = memblock_end_of_DRAM();
c67c3cb4	280	unsigned long max_zone_pfns[MAX_NR_ZONES];
7c8c6b97	281
2c419bde KG	282	#ifdef CONFIG_PPC32
	283	unsigned long v = __fix_to_virt(__end_of_fixed_addresses - 1);
	284	unsigned long end = __fix_to_virt(FIX_HOLE);
	285
	286	for (; v < end; v += PAGE_SIZE)
	287	map_page(v, 0, 0); /* XXX gross */
	288	#endif
	289
7c8c6b97 PM	290	#ifdef CONFIG_HIGHMEM
7c8c6b97 PM	291	map_page(PKMAP_BASE, 0, 0); /* XXX gross */
2c419bde KG	292	pkmap_page_table = virt_to_kpte(PKMAP_BASE);
	293
	294	kmap_pte = virt_to_kpte(__fix_to_virt(FIX_KMAP_BEGIN));
7c8c6b97 PM	295	kmap_prot = PAGE_KERNEL;
	296	#endif /* CONFIG_HIGHMEM */
	297
2bf3016f	298	printk(KERN_DEBUG "Top of RAM: 0x%llx, Total RAM: 0x%lx\n",
fb610635	299	(unsigned long long)top_of_ram, total_ram);
e110b281	300	printk(KERN_DEBUG "Memory hole size: %ldMB\n",
2bf3016f	301	(long int)((top_of_ram - total_ram) >> 20));
6391af17	302	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
7c8c6b97	303	#ifdef CONFIG_HIGHMEM
d7917ba7	304	max_zone_pfns[ZONE_DMA] = lowmem_end_addr >> PAGE_SHIFT;
6391af17	305	max_zone_pfns[ZONE_HIGHMEM] = top_of_ram >> PAGE_SHIFT;
7c8c6b97	306	#else
6391af17	307	max_zone_pfns[ZONE_DMA] = top_of_ram >> PAGE_SHIFT;
c67c3cb4 MG	308	#endif
c67c3cb4 MG	309	free_area_init_nodes(max_zone_pfns);
4e8ad3e8 JB	310
4e8ad3e8 JB	311	mark_nonram_nosave();
7c8c6b97 PM	312	}
	313	#endif /* ! CONFIG_NEED_MULTIPLE_NODES */
	314
	315	void __init mem_init(void)
	316	{
	317	#ifdef CONFIG_NEED_MULTIPLE_NODES
	318	int nid;
	319	#endif
	320	pg_data_t *pgdat;
	321	unsigned long i;
	322	struct page *page;
	323	unsigned long reservedpages = 0, codesize, initsize, datasize, bsssize;
	324
a9327296 FT	325	#ifdef CONFIG_SWIOTLB
	326	if (ppc_swiotlb_enable)
	327	swiotlb_init(1);
	328	#endif
	329
95f72d1e	330	num_physpages = memblock.memory.size >> PAGE_SHIFT;
7c8c6b97 PM	331	high_memory = (void ) __va(max_low_pfn PAGE_SIZE);
	332
	333	#ifdef CONFIG_NEED_MULTIPLE_NODES
	334	for_each_online_node(nid) {
	335	if (NODE_DATA(nid)->node_spanned_pages != 0) {
c258dd40	336	printk("freeing bootmem node %d\n", nid);
7c8c6b97 PM	337	totalram_pages +=
	338	free_all_bootmem_node(NODE_DATA(nid));
	339	}
	340	}
	341	#else
fb6d73d3	342	max_mapnr = max_pfn;
7c8c6b97 PM	343	totalram_pages += free_all_bootmem();
7c8c6b97 PM	344	#endif
ec936fc5	345	for_each_online_pgdat(pgdat) {
7c8c6b97	346	for (i = 0; i < pgdat->node_spanned_pages; i++) {
fb6d73d3 PM	347	if (!pfn_valid(pgdat->node_start_pfn + i))
fb6d73d3 PM	348	continue;
7c8c6b97 PM	349	page = pgdat_page_nr(pgdat, i);
	350	if (PageReserved(page))
	351	reservedpages++;
	352	}
	353	}
	354
	355	codesize = (unsigned long)&_sdata - (unsigned long)&_stext;
bcb35576	356	datasize = (unsigned long)&_edata - (unsigned long)&_sdata;
7c8c6b97 PM	357	initsize = (unsigned long)&__init_end - (unsigned long)&__init_begin;
	358	bsssize = (unsigned long)&__bss_stop - (unsigned long)&__bss_start;
	359
	360	#ifdef CONFIG_HIGHMEM
	361	{
	362	unsigned long pfn, highmem_mapnr;
	363
d7917ba7	364	highmem_mapnr = lowmem_end_addr >> PAGE_SHIFT;
7c8c6b97 PM	365	for (pfn = highmem_mapnr; pfn < max_mapnr; ++pfn) {
7c8c6b97 PM	366	struct page *page = pfn_to_page(pfn);
95f72d1e	367	if (memblock_is_reserved(pfn << PAGE_SHIFT))
f98eeb4e	368	continue;
7c8c6b97	369	ClearPageReserved(page);
7835e98b	370	init_page_count(page);
7c8c6b97 PM	371	__free_page(page);
7c8c6b97 PM	372	totalhigh_pages++;
f98eeb4e	373	reservedpages--;
7c8c6b97 PM	374	}
7c8c6b97 PM	375	totalram_pages += totalhigh_pages;
e110b281	376	printk(KERN_DEBUG "High memory: %luk\n",
7c8c6b97 PM	377	totalhigh_pages << (PAGE_SHIFT-10));
	378	}
	379	#endif /* CONFIG_HIGHMEM */
	380
	381	printk(KERN_INFO "Memory: %luk/%luk available (%luk kernel code, "
	382	"%luk reserved, %luk data, %luk bss, %luk init)\n",
cc013a88	383	nr_free_pages() << (PAGE_SHIFT-10),
7c8c6b97 PM	384	num_physpages << (PAGE_SHIFT-10),
	385	codesize >> 10,
	386	reservedpages << (PAGE_SHIFT-10),
	387	datasize >> 10,
	388	bsssize >> 10,
	389	initsize >> 10);
	390
f637a49e BH	391	#ifdef CONFIG_PPC32
	392	pr_info("Kernel virtual memory layout:\n");
	393	pr_info(" * 0x%08lx..0x%08lx : fixmap\n", FIXADDR_START, FIXADDR_TOP);
	394	#ifdef CONFIG_HIGHMEM
	395	pr_info(" * 0x%08lx..0x%08lx : highmem PTEs\n",
	396	PKMAP_BASE, PKMAP_ADDR(LAST_PKMAP));
	397	#endif /* CONFIG_HIGHMEM */
8b31e49d BH	398	#ifdef CONFIG_NOT_COHERENT_CACHE
	399	pr_info(" * 0x%08lx..0x%08lx : consistent mem\n",
	400	IOREMAP_TOP, IOREMAP_TOP + CONFIG_CONSISTENT_SIZE);
	401	#endif /* CONFIG_NOT_COHERENT_CACHE */
f637a49e BH	402	pr_info(" * 0x%08lx..0x%08lx : early ioremap\n",
	403	ioremap_bot, IOREMAP_TOP);
	404	pr_info(" * 0x%08lx..0x%08lx : vmalloc & ioremap\n",
	405	VMALLOC_START, VMALLOC_END);
	406	#endif /* CONFIG_PPC32 */
	407
7c8c6b97	408	mem_init_done = 1;
7c8c6b97 PM	409	}
7c8c6b97 PM	410
14cf11af PM	411	/*
	412	* This is called when a page has been modified by the kernel.
	413	* It just marks the page as not i-cache clean. We do the i-cache
	414	* flush later when the page is given to a user process, if necessary.
	415	*/
	416	void flush_dcache_page(struct page *page)
	417	{
	418	if (cpu_has_feature(CPU_FTR_COHERENT_ICACHE))
	419	return;
	420	/* avoid an atomic op if possible */
	421	if (test_bit(PG_arch_1, &page->flags))
	422	clear_bit(PG_arch_1, &page->flags);
	423	}
	424	EXPORT_SYMBOL(flush_dcache_page);
	425
	426	void flush_dcache_icache_page(struct page *page)
	427	{
0895ecda DG	428	#ifdef CONFIG_HUGETLB_PAGE
	429	if (PageCompound(page)) {
	430	flush_dcache_icache_hugepage(page);
	431	return;
	432	}
	433	#endif
14cf11af	434	#ifdef CONFIG_BOOKE
0895ecda DG	435	{
	436	void *start = kmap_atomic(page, KM_PPC_SYNC_ICACHE);
	437	__flush_dcache_icache(start);
	438	kunmap_atomic(start, KM_PPC_SYNC_ICACHE);
	439	}
ab1f9dac	440	#elif defined(CONFIG_8xx) \|\| defined(CONFIG_PPC64)
14cf11af PM	441	/* On 8xx there is no need to kmap since highmem is not supported */
	442	__flush_dcache_icache(page_address(page));
	443	#else
	444	__flush_dcache_icache_phys(page_to_pfn(page) << PAGE_SHIFT);
	445	#endif
14cf11af	446	}
0895ecda	447
14cf11af PM	448	void clear_user_page(void page, unsigned long vaddr, struct page pg)
	449	{
	450	clear_page(page);
	451
14cf11af PM	452	/*
	453	* We shouldnt have to do this, but some versions of glibc
	454	* require it (ld.so assumes zero filled pages are icache clean)
	455	* - Anton
	456	*/
09f5dc44	457	flush_dcache_page(pg);
14cf11af PM	458	}
	459	EXPORT_SYMBOL(clear_user_page);
	460
	461	void copy_user_page(void vto, void vfrom, unsigned long vaddr,
	462	struct page *pg)
	463	{
	464	copy_page(vto, vfrom);
	465
	466	/*
	467	* We should be able to use the following optimisation, however
	468	* there are two problems.
	469	* Firstly a bug in some versions of binutils meant PLT sections
	470	* were not marked executable.
	471	* Secondly the first word in the GOT section is blrl, used
	472	* to establish the GOT address. Until recently the GOT was
	473	* not marked executable.
	474	* - Anton
	475	*/
	476	#if 0
	477	if (!vma->vm_file && ((vma->vm_flags & VM_EXEC) == 0))
	478	return;
	479	#endif
	480
09f5dc44	481	flush_dcache_page(pg);
14cf11af PM	482	}
	483
	484	void flush_icache_user_range(struct vm_area_struct vma, struct page page,
	485	unsigned long addr, int len)
	486	{
	487	unsigned long maddr;
	488
	489	maddr = (unsigned long) kmap(page) + (addr & ~PAGE_MASK);
	490	flush_icache_range(maddr, maddr + len);
	491	kunmap(page);
	492	}
	493	EXPORT_SYMBOL(flush_icache_user_range);
	494
	495	/*
	496	* This is called at the end of handling a user page fault, when the
	497	* fault has been handled by updating a PTE in the linux page tables.
	498	* We use it to preload an HPTE into the hash table corresponding to
	499	* the updated linux PTE.
	500	*
01edcd89	501	* This must always be called with the pte lock held.
14cf11af PM	502	*/
14cf11af PM	503	void update_mmu_cache(struct vm_area_struct *vma, unsigned long address,
4b3073e1	504	pte_t *ptep)
14cf11af	505	{
3c726f8d BH	506	#ifdef CONFIG_PPC_STD_MMU
3c726f8d BH	507	unsigned long access = 0, trap;
14cf11af	508
14cf11af	509	/* We only want HPTEs for linux PTEs that have _PAGE_ACCESSED set */
4b3073e1	510	if (!pte_young(*ptep) \|\| address >= TASK_SIZE)
14cf11af	511	return;
14cf11af	512
3c726f8d BH	513	/* We try to figure out if we are coming from an instruction
	514	* access fault and pass that down to __hash_page so we avoid
	515	* double-faulting on execution of fresh text. We have to test
	516	* for regs NULL since init will get here first thing at boot
	517	*
	518	* We also avoid filling the hash if not coming from a fault
	519	*/
	520	if (current->thread.regs == NULL)
14cf11af	521	return;
3c726f8d BH	522	trap = TRAP(current->thread.regs);
	523	if (trap == 0x400)
	524	access \|= _PAGE_EXEC;
	525	else if (trap != 0x300)
	526	return;
	527	hash_preload(vma->vm_mm, address, access, trap);
	528	#endif /* CONFIG_PPC_STD_MMU */
14cf11af	529	}