[net-next-2.6.git] / arch / ia64 / mm / contig.c

/*
 * This file is subject to the terms and conditions of the GNU General Public
 * License.  See the file "COPYING" in the main directory of this archive
 * for more details.
 *
 * Copyright (C) 1998-2003 Hewlett-Packard Co
 *	David Mosberger-Tang <davidm@hpl.hp.com>
 *	Stephane Eranian <eranian@hpl.hp.com>
 * Copyright (C) 2000, Rohit Seth <rohit.seth@intel.com>
 * Copyright (C) 1999 VA Linux Systems
 * Copyright (C) 1999 Walt Drummond <drummond@valinux.com>
 * Copyright (C) 2003 Silicon Graphics, Inc. All rights reserved.
 *
 * Routines used by ia64 machines with contiguous (or virtually contiguous)
 * memory.
 */
#include <linux/bootmem.h>
#include <linux/efi.h>
#include <linux/mm.h>
#include <linux/swap.h>

#include <asm/meminit.h>
#include <asm/pgalloc.h>
#include <asm/pgtable.h>
#include <asm/sections.h>
#include <asm/mca.h>

#ifdef CONFIG_VIRTUAL_MEM_MAP
static unsigned long max_gap;
#endif

/**
 * show_mem - give short summary of memory stats
 *
 * Shows a simple page count of reserved and used pages in the system.
 * For discontig machines, it does this on a per-pgdat basis.
 */
void show_mem(void)
{
	int i, total_reserved = 0;
	int total_shared = 0, total_cached = 0;
	unsigned long total_present = 0;
	pg_data_t *pgdat;

	printk(KERN_INFO "Mem-info:\n");
	show_free_areas();
	printk(KERN_INFO "Free swap:       %6ldkB\n",
	       nr_swap_pages<<(PAGE_SHIFT-10));
	printk(KERN_INFO "Node memory in pages:\n");
	for_each_online_pgdat(pgdat) {
		unsigned long present;
		unsigned long flags;
		int shared = 0, cached = 0, reserved = 0;

		pgdat_resize_lock(pgdat, &flags);
		present = pgdat->node_present_pages;
		for(i = 0; i < pgdat->node_spanned_pages; i++) {
			struct page *page;
			if (pfn_valid(pgdat->node_start_pfn + i))
				page = pfn_to_page(pgdat->node_start_pfn + i);
			else {
#ifdef CONFIG_VIRTUAL_MEM_MAP
				if (max_gap < LARGE_GAP)
					continue;
#endif
				i = vmemmap_find_next_valid_pfn(pgdat->node_id,
					 i) - 1;
				continue;
			}
			if (PageReserved(page))
				reserved++;
			else if (PageSwapCache(page))
				cached++;
			else if (page_count(page))
				shared += page_count(page)-1;
		}
		pgdat_resize_unlock(pgdat, &flags);
		total_present += present;
		total_reserved += reserved;
		total_cached += cached;
		total_shared += shared;
		printk(KERN_INFO "Node %4d:  RAM: %11ld, rsvd: %8d, "
		       "shrd: %10d, swpd: %10d\n", pgdat->node_id,
		       present, reserved, shared, cached);
	}
	printk(KERN_INFO "%ld pages of RAM\n", total_present);
	printk(KERN_INFO "%d reserved pages\n", total_reserved);
	printk(KERN_INFO "%d pages shared\n", total_shared);
	printk(KERN_INFO "%d pages swap cached\n", total_cached);
	printk(KERN_INFO "Total of %ld pages in page table cache\n",
	       pgtable_quicklist_total_size());
	printk(KERN_INFO "%d free buffer pages\n", nr_free_buffer_pages());
}


/* physical address where the bootmem map is located */
unsigned long bootmap_start;

/**
 * find_max_pfn - adjust the maximum page number callback
 * @start: start of range
 * @end: end of range
 * @arg: address of pointer to global max_pfn variable
 *
 * Passed as a callback function to efi_memmap_walk() to determine the highest
 * available page frame number in the system.
 */
int
find_max_pfn (unsigned long start, unsigned long end, void *arg)
{
	unsigned long *max_pfnp = arg, pfn;

	pfn = (PAGE_ALIGN(end - 1) - PAGE_OFFSET) >> PAGE_SHIFT;
	if (pfn > *max_pfnp)
		*max_pfnp = pfn;
	return 0;
}

/**
 * find_bootmap_location - callback to find a memory area for the bootmap
 * @start: start of region
 * @end: end of region
 * @arg: unused callback data
 *
 * Find a place to put the bootmap and return its starting address in
 * bootmap_start.  This address must be page-aligned.
 */
static int __init
find_bootmap_location (unsigned long start, unsigned long end, void *arg)
{
	unsigned long needed = *(unsigned long *)arg;
	unsigned long range_start, range_end, free_start;
	int i;

#if IGNORE_PFN0
	if (start == PAGE_OFFSET) {
		start += PAGE_SIZE;
		if (start >= end)
			return 0;
	}
#endif

	free_start = PAGE_OFFSET;

	for (i = 0; i < num_rsvd_regions; i++) {
		range_start = max(start, free_start);
		range_end   = min(end, rsvd_region[i].start & PAGE_MASK);

		free_start = PAGE_ALIGN(rsvd_region[i].end);

		if (range_end <= range_start)
			continue; /* skip over empty range */

		if (range_end - range_start >= needed) {
			bootmap_start = __pa(range_start);
			return -1;	/* done */
		}

		/* nothing more available in this segment */
		if (range_end == end)
			return 0;
	}
	return 0;
}

/**
 * find_memory - setup memory map
 *
 * Walk the EFI memory map and find usable memory for the system, taking
 * into account reserved areas.
 */
void __init
find_memory (void)
{
	unsigned long bootmap_size;

	reserve_memory();

	/* first find highest page frame number */
	max_pfn = 0;
	efi_memmap_walk(find_max_pfn, &max_pfn);

	/* how many bytes to cover all the pages */
	bootmap_size = bootmem_bootmap_pages(max_pfn) << PAGE_SHIFT;

	/* look for a location to hold the bootmap */
	bootmap_start = ~0UL;
	efi_memmap_walk(find_bootmap_location, &bootmap_size);
	if (bootmap_start == ~0UL)
		panic("Cannot find %ld bytes for bootmap\n", bootmap_size);

	bootmap_size = init_bootmem(bootmap_start >> PAGE_SHIFT, max_pfn);

	/* Free all available memory, then mark bootmem-map as being in use. */
	efi_memmap_walk(filter_rsvd_memory, free_bootmem);
	reserve_bootmem(bootmap_start, bootmap_size);

	find_initrd();

#ifdef CONFIG_CRASH_DUMP
	/* If we are doing a crash dump, we still need to know the real mem
	 * size before original memory map is reset. */
	saved_max_pfn = max_pfn;
#endif
}

#ifdef CONFIG_SMP
/**
 * per_cpu_init - setup per-cpu variables
 *
 * Allocate and setup per-cpu data areas.
 */
void * __cpuinit
per_cpu_init (void)
{
	void *cpu_data;
	int cpu;
	static int first_time=1;

	/*
	 * get_free_pages() cannot be used before cpu_init() done.  BSP
	 * allocates "NR_CPUS" pages for all CPUs to avoid that AP calls
	 * get_zeroed_page().
	 */
	if (first_time) {
		first_time=0;
		cpu_data = __alloc_bootmem(PERCPU_PAGE_SIZE * NR_CPUS,
					   PERCPU_PAGE_SIZE, __pa(MAX_DMA_ADDRESS));
		for (cpu = 0; cpu < NR_CPUS; cpu++) {
			memcpy(cpu_data, __phys_per_cpu_start, __per_cpu_end - __per_cpu_start);
			__per_cpu_offset[cpu] = (char *) cpu_data - __per_cpu_start;
			cpu_data += PERCPU_PAGE_SIZE;
			per_cpu(local_per_cpu_offset, cpu) = __per_cpu_offset[cpu];
		}
	}
	return __per_cpu_start + __per_cpu_offset[smp_processor_id()];
}
#endif /* CONFIG_SMP */

static int
count_pages (u64 start, u64 end, void *arg)
{
	unsigned long *count = arg;

	*count += (end - start) >> PAGE_SHIFT;
	return 0;
}

/*
 * Set up the page tables.
 */

void __init
paging_init (void)
{
	unsigned long max_dma;
	unsigned long max_zone_pfns[MAX_NR_ZONES];

	num_physpages = 0;
	efi_memmap_walk(count_pages, &num_physpages);

	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
#ifdef CONFIG_ZONE_DMA
	max_dma = virt_to_phys((void *) MAX_DMA_ADDRESS) >> PAGE_SHIFT;
	max_zone_pfns[ZONE_DMA] = max_dma;
#endif
	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;

#ifdef CONFIG_VIRTUAL_MEM_MAP
	efi_memmap_walk(register_active_ranges, NULL);
	efi_memmap_walk(find_largest_hole, (u64 *)&max_gap);
	if (max_gap < LARGE_GAP) {
		vmem_map = (struct page *) 0;
		free_area_init_nodes(max_zone_pfns);
	} else {
		unsigned long map_size;

		/* allocate virtual_mem_map */

		map_size = PAGE_ALIGN(ALIGN(max_low_pfn, MAX_ORDER_NR_PAGES) *
			sizeof(struct page));
		vmalloc_end -= map_size;
		vmem_map = (struct page *) vmalloc_end;
		efi_memmap_walk(create_mem_map_page_table, NULL);

		/*
		 * alloc_node_mem_map makes an adjustment for mem_map
		 * which isn't compatible with vmem_map.
		 */
		NODE_DATA(0)->node_mem_map = vmem_map +
			find_min_pfn_with_active_regions();
		free_area_init_nodes(max_zone_pfns);

		printk("Virtual mem_map starts at 0x%p\n", mem_map);
	}
#else /* !CONFIG_VIRTUAL_MEM_MAP */
	add_active_range(0, 0, max_low_pfn);
	free_area_init_nodes(max_zone_pfns);
#endif /* !CONFIG_VIRTUAL_MEM_MAP */
	zero_page_memmap_ptr = virt_to_page(ia64_imva(empty_zero_page));
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* This file is subject to the terms and conditions of the GNU General Public
	3	* License. See the file "COPYING" in the main directory of this archive
	4	* for more details.
	5	*
	6	* Copyright (C) 1998-2003 Hewlett-Packard Co
	7	* David Mosberger-Tang <davidm@hpl.hp.com>
	8	* Stephane Eranian <eranian@hpl.hp.com>
	9	* Copyright (C) 2000, Rohit Seth <rohit.seth@intel.com>
	10	* Copyright (C) 1999 VA Linux Systems
	11	* Copyright (C) 1999 Walt Drummond <drummond@valinux.com>
	12	* Copyright (C) 2003 Silicon Graphics, Inc. All rights reserved.
	13	*
	14	* Routines used by ia64 machines with contiguous (or virtually contiguous)
	15	* memory.
	16	*/
1da177e4 LT	17	#include <linux/bootmem.h>
	18	#include <linux/efi.h>
	19	#include <linux/mm.h>
	20	#include <linux/swap.h>
	21
	22	#include <asm/meminit.h>
	23	#include <asm/pgalloc.h>
	24	#include <asm/pgtable.h>
	25	#include <asm/sections.h>
	26	#include <asm/mca.h>
	27
	28	#ifdef CONFIG_VIRTUAL_MEM_MAP
e44e41d0	29	static unsigned long max_gap;
1da177e4 LT	30	#endif
	31
	32	/**
f1c0afa2	33	* show_mem - give short summary of memory stats
1da177e4	34	*
f1c0afa2 GB	35	* Shows a simple page count of reserved and used pages in the system.
f1c0afa2 GB	36	* For discontig machines, it does this on a per-pgdat basis.
1da177e4	37	*/
f1c0afa2	38	void show_mem(void)
1da177e4	39	{
f1c0afa2 GB	40	int i, total_reserved = 0;
	41	int total_shared = 0, total_cached = 0;
	42	unsigned long total_present = 0;
	43	pg_data_t *pgdat;
1da177e4	44
709a6c1c	45	printk(KERN_INFO "Mem-info:\n");
1da177e4	46	show_free_areas();
709a6c1c JS	47	printk(KERN_INFO "Free swap: %6ldkB\n",
709a6c1c JS	48	nr_swap_pages<<(PAGE_SHIFT-10));
f1c0afa2 GB	49	printk(KERN_INFO "Node memory in pages:\n");
	50	for_each_online_pgdat(pgdat) {
	51	unsigned long present;
	52	unsigned long flags;
	53	int shared = 0, cached = 0, reserved = 0;
	54
	55	pgdat_resize_lock(pgdat, &flags);
	56	present = pgdat->node_present_pages;
	57	for(i = 0; i < pgdat->node_spanned_pages; i++) {
	58	struct page *page;
	59	if (pfn_valid(pgdat->node_start_pfn + i))
	60	page = pfn_to_page(pgdat->node_start_pfn + i);
	61	else {
e44e41d0	62	#ifdef CONFIG_VIRTUAL_MEM_MAP
f1c0afa2 GB	63	if (max_gap < LARGE_GAP)
f1c0afa2 GB	64	continue;
e44e41d0	65	#endif
f1c0afa2 GB	66	i = vmemmap_find_next_valid_pfn(pgdat->node_id,
	67	i) - 1;
	68	continue;
	69	}
	70	if (PageReserved(page))
	71	reserved++;
	72	else if (PageSwapCache(page))
	73	cached++;
	74	else if (page_count(page))
	75	shared += page_count(page)-1;
e44e41d0	76	}
f1c0afa2 GB	77	pgdat_resize_unlock(pgdat, &flags);
	78	total_present += present;
	79	total_reserved += reserved;
	80	total_cached += cached;
	81	total_shared += shared;
	82	printk(KERN_INFO "Node %4d: RAM: %11ld, rsvd: %8d, "
	83	"shrd: %10d, swpd: %10d\n", pgdat->node_id,
	84	present, reserved, shared, cached);
1da177e4	85	}
f1c0afa2 GB	86	printk(KERN_INFO "%ld pages of RAM\n", total_present);
	87	printk(KERN_INFO "%d reserved pages\n", total_reserved);
	88	printk(KERN_INFO "%d pages shared\n", total_shared);
	89	printk(KERN_INFO "%d pages swap cached\n", total_cached);
	90	printk(KERN_INFO "Total of %ld pages in page table cache\n",
709a6c1c	91	pgtable_quicklist_total_size());
f1c0afa2	92	printk(KERN_INFO "%d free buffer pages\n", nr_free_buffer_pages());
1da177e4 LT	93	}
1da177e4 LT	94
f1c0afa2	95
1da177e4 LT	96	/* physical address where the bootmem map is located */
	97	unsigned long bootmap_start;
	98
	99	/**
	100	* find_max_pfn - adjust the maximum page number callback
	101	* @start: start of range
	102	* @end: end of range
	103	* @arg: address of pointer to global max_pfn variable
	104	*
	105	* Passed as a callback function to efi_memmap_walk() to determine the highest
	106	* available page frame number in the system.
	107	*/
	108	int
	109	find_max_pfn (unsigned long start, unsigned long end, void *arg)
	110	{
	111	unsigned long *max_pfnp = arg, pfn;
	112
	113	pfn = (PAGE_ALIGN(end - 1) - PAGE_OFFSET) >> PAGE_SHIFT;
	114	if (pfn > *max_pfnp)
	115	*max_pfnp = pfn;
	116	return 0;
	117	}
	118
	119	/**
	120	* find_bootmap_location - callback to find a memory area for the bootmap
	121	* @start: start of region
	122	* @end: end of region
	123	* @arg: unused callback data
	124	*
	125	* Find a place to put the bootmap and return its starting address in
	126	* bootmap_start. This address must be page-aligned.
	127	*/
dae28066	128	static int __init
1da177e4 LT	129	find_bootmap_location (unsigned long start, unsigned long end, void *arg)
	130	{
	131	unsigned long needed = (unsigned long )arg;
	132	unsigned long range_start, range_end, free_start;
	133	int i;
	134
	135	#if IGNORE_PFN0
	136	if (start == PAGE_OFFSET) {
	137	start += PAGE_SIZE;
	138	if (start >= end)
	139	return 0;
	140	}
	141	#endif
	142
	143	free_start = PAGE_OFFSET;
	144
	145	for (i = 0; i < num_rsvd_regions; i++) {
	146	range_start = max(start, free_start);
	147	range_end = min(end, rsvd_region[i].start & PAGE_MASK);
	148
	149	free_start = PAGE_ALIGN(rsvd_region[i].end);
	150
	151	if (range_end <= range_start)
	152	continue; /* skip over empty range */
	153
	154	if (range_end - range_start >= needed) {
	155	bootmap_start = __pa(range_start);
	156	return -1; /* done */
	157	}
	158
	159	/* nothing more available in this segment */
	160	if (range_end == end)
	161	return 0;
	162	}
	163	return 0;
	164	}
	165
	166	/**
	167	* find_memory - setup memory map
	168	*
	169	* Walk the EFI memory map and find usable memory for the system, taking
	170	* into account reserved areas.
	171	*/
dae28066	172	void __init
1da177e4 LT	173	find_memory (void)
	174	{
	175	unsigned long bootmap_size;
	176
	177	reserve_memory();
	178
	179	/* first find highest page frame number */
	180	max_pfn = 0;
	181	efi_memmap_walk(find_max_pfn, &max_pfn);
	182
	183	/* how many bytes to cover all the pages */
	184	bootmap_size = bootmem_bootmap_pages(max_pfn) << PAGE_SHIFT;
	185
	186	/* look for a location to hold the bootmap */
	187	bootmap_start = ~0UL;
	188	efi_memmap_walk(find_bootmap_location, &bootmap_size);
	189	if (bootmap_start == ~0UL)
	190	panic("Cannot find %ld bytes for bootmap\n", bootmap_size);
	191
	192	bootmap_size = init_bootmem(bootmap_start >> PAGE_SHIFT, max_pfn);
	193
	194	/* Free all available memory, then mark bootmem-map as being in use. */
	195	efi_memmap_walk(filter_rsvd_memory, free_bootmem);
	196	reserve_bootmem(bootmap_start, bootmap_size);
	197
	198	find_initrd();
45a98fc6 H	199
	200	#ifdef CONFIG_CRASH_DUMP
	201	/* If we are doing a crash dump, we still need to know the real mem
233c2f99	202	* size before original memory map is reset. */
45a98fc6 H	203	saved_max_pfn = max_pfn;
45a98fc6 H	204	#endif
1da177e4 LT	205	}
	206
	207	#ifdef CONFIG_SMP
	208	/**
	209	* per_cpu_init - setup per-cpu variables
	210	*
	211	* Allocate and setup per-cpu data areas.
	212	*/
244fd545	213	void * __cpuinit
1da177e4 LT	214	per_cpu_init (void)
	215	{
	216	void *cpu_data;
	217	int cpu;
ff741906	218	static int first_time=1;
1da177e4 LT	219
	220	/*
	221	* get_free_pages() cannot be used before cpu_init() done. BSP
	222	* allocates "NR_CPUS" pages for all CPUs to avoid that AP calls
	223	* get_zeroed_page().
	224	*/
ff741906 AR	225	if (first_time) {
ff741906 AR	226	first_time=0;
1da177e4 LT	227	cpu_data = __alloc_bootmem(PERCPU_PAGE_SIZE * NR_CPUS,
	228	PERCPU_PAGE_SIZE, __pa(MAX_DMA_ADDRESS));
	229	for (cpu = 0; cpu < NR_CPUS; cpu++) {
	230	memcpy(cpu_data, __phys_per_cpu_start, __per_cpu_end - __per_cpu_start);
	231	__per_cpu_offset[cpu] = (char *) cpu_data - __per_cpu_start;
	232	cpu_data += PERCPU_PAGE_SIZE;
	233	per_cpu(local_per_cpu_offset, cpu) = __per_cpu_offset[cpu];
	234	}
	235	}
	236	return __per_cpu_start + __per_cpu_offset[smp_processor_id()];
	237	}
	238	#endif /* CONFIG_SMP */
	239
	240	static int
	241	count_pages (u64 start, u64 end, void *arg)
	242	{
	243	unsigned long *count = arg;
	244
	245	*count += (end - start) >> PAGE_SHIFT;
	246	return 0;
	247	}
	248
1da177e4 LT	249	/*
	250	* Set up the page tables.
	251	*/
	252
dae28066	253	void __init
1da177e4 LT	254	paging_init (void)
	255	{
	256	unsigned long max_dma;
05e0caad	257	unsigned long max_zone_pfns[MAX_NR_ZONES];
1da177e4 LT	258
	259	num_physpages = 0;
	260	efi_memmap_walk(count_pages, &num_physpages);
	261
6391af17	262	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
09ae1f58 CL	263	#ifdef CONFIG_ZONE_DMA
09ae1f58 CL	264	max_dma = virt_to_phys((void *) MAX_DMA_ADDRESS) >> PAGE_SHIFT;
05e0caad	265	max_zone_pfns[ZONE_DMA] = max_dma;
09ae1f58	266	#endif
05e0caad	267	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
1da177e4 LT	268
1da177e4 LT	269	#ifdef CONFIG_VIRTUAL_MEM_MAP
8b9c1068	270	efi_memmap_walk(register_active_ranges, NULL);
1da177e4 LT	271	efi_memmap_walk(find_largest_hole, (u64 *)&max_gap);
	272	if (max_gap < LARGE_GAP) {
	273	vmem_map = (struct page *) 0;
05e0caad	274	free_area_init_nodes(max_zone_pfns);
1da177e4 LT	275	} else {
	276	unsigned long map_size;
	277
	278	/* allocate virtual_mem_map */
	279
921eea1c BP	280	map_size = PAGE_ALIGN(ALIGN(max_low_pfn, MAX_ORDER_NR_PAGES) *
921eea1c BP	281	sizeof(struct page));
1da177e4 LT	282	vmalloc_end -= map_size;
	283	vmem_map = (struct page *) vmalloc_end;
	284	efi_memmap_walk(create_mem_map_page_table, NULL);
	285
05e0caad MG	286	/*
	287	* alloc_node_mem_map makes an adjustment for mem_map
	288	* which isn't compatible with vmem_map.
	289	*/
	290	NODE_DATA(0)->node_mem_map = vmem_map +
	291	find_min_pfn_with_active_regions();
	292	free_area_init_nodes(max_zone_pfns);
1da177e4 LT	293
	294	printk("Virtual mem_map starts at 0x%p\n", mem_map);
	295	}
	296	#else /* !CONFIG_VIRTUAL_MEM_MAP */
05e0caad MG	297	add_active_range(0, 0, max_low_pfn);
05e0caad MG	298	free_area_init_nodes(max_zone_pfns);
1da177e4 LT	299	#endif /* !CONFIG_VIRTUAL_MEM_MAP */
	300	zero_page_memmap_ptr = virt_to_page(ia64_imva(empty_zero_page));
	301	}