contig.c 8.0 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
/*
 * 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>
20
#include <linux/nmi.h>
L
Linus Torvalds 已提交
21 22 23 24 25 26 27 28 29
#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
30
static unsigned long max_gap;
L
Linus Torvalds 已提交
31 32 33
#endif

/**
34
 * show_mem - give short summary of memory stats
L
Linus Torvalds 已提交
35
 *
36 37
 * Shows a simple page count of reserved and used pages in the system.
 * For discontig machines, it does this on a per-pgdat basis.
L
Linus Torvalds 已提交
38
 */
39
void show_mem(void)
L
Linus Torvalds 已提交
40
{
41 42 43 44
	int i, total_reserved = 0;
	int total_shared = 0, total_cached = 0;
	unsigned long total_present = 0;
	pg_data_t *pgdat;
L
Linus Torvalds 已提交
45

J
Jes Sorensen 已提交
46
	printk(KERN_INFO "Mem-info:\n");
L
Linus Torvalds 已提交
47
	show_free_areas();
48 49 50 51 52 53 54 55 56 57
	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;
58 59
			if (unlikely(i % MAX_ORDER_NR_PAGES == 0))
				touch_nmi_watchdog();
60 61 62
			if (pfn_valid(pgdat->node_start_pfn + i))
				page = pfn_to_page(pgdat->node_start_pfn + i);
			else {
63
#ifdef CONFIG_VIRTUAL_MEM_MAP
64 65
				if (max_gap < LARGE_GAP)
					continue;
66
#endif
67 68 69 70 71 72 73 74 75 76
				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;
77
		}
78 79 80 81 82 83 84 85
		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);
L
Linus Torvalds 已提交
86
	}
87 88 89 90 91
	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",
92
	       quicklist_total_size());
93
	printk(KERN_INFO "%d free buffer pages\n", nr_free_buffer_pages());
L
Linus Torvalds 已提交
94 95
}

96

L
Linus Torvalds 已提交
97 98 99 100 101 102 103 104 105 106 107 108
/* physical address where the bootmem map is located */
unsigned long bootmap_start;

/**
 * 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.
 */
109
static int __init
110
find_bootmap_location (u64 start, u64 end, void *arg)
L
Linus Torvalds 已提交
111
{
112 113
	u64 needed = *(unsigned long *)arg;
	u64 range_start, range_end, free_start;
L
Linus Torvalds 已提交
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146
	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;
}

147 148 149 150 151 152 153 154 155 156
#ifdef CONFIG_SMP
static void *cpu_data;
/**
 * per_cpu_init - setup per-cpu variables
 *
 * Allocate and setup per-cpu data areas.
 */
void * __cpuinit
per_cpu_init (void)
{
157 158 159 160 161 162 163
	static bool first_time = true;
	void *cpu0_data = __cpu0_per_cpu;
	unsigned int cpu;

	if (!first_time)
		goto skip;
	first_time = false;
164 165 166 167 168 169

	/*
	 * 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().
	 */
170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190
	for (cpu = 0; cpu < NR_CPUS; cpu++) {
		void *src = cpu == 0 ? cpu0_data : __phys_per_cpu_start;

		memcpy(cpu_data, src, __per_cpu_end - __per_cpu_start);
		__per_cpu_offset[cpu] = (char *)cpu_data - __per_cpu_start;
		per_cpu(local_per_cpu_offset, cpu) = __per_cpu_offset[cpu];

		/*
		 * percpu area for cpu0 is moved from the __init area
		 * which is setup by head.S and used till this point.
		 * Update ar.k3.  This move is ensures that percpu
		 * area for cpu0 is on the correct node and its
		 * virtual address isn't insanely far from other
		 * percpu areas which is important for congruent
		 * percpu allocator.
		 */
		if (cpu == 0)
			ia64_set_kr(IA64_KR_PER_CPU_DATA, __pa(cpu_data) -
				    (unsigned long)__per_cpu_start);

		cpu_data += PERCPU_PAGE_SIZE;
191
	}
192
skip:
193 194 195 196 197 198
	return __per_cpu_start + __per_cpu_offset[smp_processor_id()];
}

static inline void
alloc_per_cpu_data(void)
{
199
	cpu_data = __alloc_bootmem(PERCPU_PAGE_SIZE * NR_CPUS,
200 201 202 203 204 205
				   PERCPU_PAGE_SIZE, __pa(MAX_DMA_ADDRESS));
}
#else
#define alloc_per_cpu_data() do { } while (0)
#endif /* CONFIG_SMP */

L
Linus Torvalds 已提交
206 207 208 209 210 211
/**
 * find_memory - setup memory map
 *
 * Walk the EFI memory map and find usable memory for the system, taking
 * into account reserved areas.
 */
212
void __init
L
Linus Torvalds 已提交
213 214 215 216 217 218 219
find_memory (void)
{
	unsigned long bootmap_size;

	reserve_memory();

	/* first find highest page frame number */
220 221 222 223
	min_low_pfn = ~0UL;
	max_low_pfn = 0;
	efi_memmap_walk(find_max_min_low_pfn, NULL);
	max_pfn = max_low_pfn;
L
Linus Torvalds 已提交
224 225 226 227 228 229 230 231 232
	/* 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);

233 234
	bootmap_size = init_bootmem_node(NODE_DATA(0),
			(bootmap_start >> PAGE_SHIFT), 0, max_pfn);
L
Linus Torvalds 已提交
235 236 237

	/* Free all available memory, then mark bootmem-map as being in use. */
	efi_memmap_walk(filter_rsvd_memory, free_bootmem);
238
	reserve_bootmem(bootmap_start, bootmap_size, BOOTMEM_DEFAULT);
L
Linus Torvalds 已提交
239 240

	find_initrd();
241

242
	alloc_per_cpu_data();
L
Linus Torvalds 已提交
243 244
}

245
static int count_pages(u64 start, u64 end, void *arg)
L
Linus Torvalds 已提交
246 247 248 249 250 251 252 253 254 255 256
{
	unsigned long *count = arg;

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

/*
 * Set up the page tables.
 */

257
void __init
L
Linus Torvalds 已提交
258 259 260
paging_init (void)
{
	unsigned long max_dma;
261
	unsigned long max_zone_pfns[MAX_NR_ZONES];
L
Linus Torvalds 已提交
262 263 264 265

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

266
	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
267 268
#ifdef CONFIG_ZONE_DMA
	max_dma = virt_to_phys((void *) MAX_DMA_ADDRESS) >> PAGE_SHIFT;
269
	max_zone_pfns[ZONE_DMA] = max_dma;
270
#endif
271
	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
L
Linus Torvalds 已提交
272 273

#ifdef CONFIG_VIRTUAL_MEM_MAP
274
	efi_memmap_walk(filter_memory, register_active_ranges);
L
Linus Torvalds 已提交
275 276 277
	efi_memmap_walk(find_largest_hole, (u64 *)&max_gap);
	if (max_gap < LARGE_GAP) {
		vmem_map = (struct page *) 0;
278
		free_area_init_nodes(max_zone_pfns);
L
Linus Torvalds 已提交
279 280 281 282 283
	} else {
		unsigned long map_size;

		/* allocate virtual_mem_map */

284 285
		map_size = PAGE_ALIGN(ALIGN(max_low_pfn, MAX_ORDER_NR_PAGES) *
			sizeof(struct page));
286 287
		VMALLOC_END -= map_size;
		vmem_map = (struct page *) VMALLOC_END;
L
Linus Torvalds 已提交
288 289
		efi_memmap_walk(create_mem_map_page_table, NULL);

290 291 292 293 294 295 296
		/*
		 * 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);
L
Linus Torvalds 已提交
297 298 299 300

		printk("Virtual mem_map starts at 0x%p\n", mem_map);
	}
#else /* !CONFIG_VIRTUAL_MEM_MAP */
301 302
	add_active_range(0, 0, max_low_pfn);
	free_area_init_nodes(max_zone_pfns);
L
Linus Torvalds 已提交
303 304 305
#endif /* !CONFIG_VIRTUAL_MEM_MAP */
	zero_page_memmap_ptr = virt_to_page(ia64_imva(empty_zero_page));
}