2005-04-16 16:20:36 -06:00
|
|
|
/*
|
|
|
|
* linux/arch/x86_64/mm/init.c
|
|
|
|
*
|
|
|
|
* Copyright (C) 1995 Linus Torvalds
|
|
|
|
* Copyright (C) 2000 Pavel Machek <pavel@suse.cz>
|
|
|
|
* Copyright (C) 2002,2003 Andi Kleen <ak@suse.de>
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/signal.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/ptrace.h>
|
|
|
|
#include <linux/mman.h>
|
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/swap.h>
|
|
|
|
#include <linux/smp.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/pagemap.h>
|
|
|
|
#include <linux/bootmem.h>
|
|
|
|
#include <linux/proc_fs.h>
|
2005-11-05 09:25:53 -07:00
|
|
|
#include <linux/pci.h>
|
2006-06-27 03:53:52 -06:00
|
|
|
#include <linux/poison.h>
|
2006-01-11 14:44:42 -07:00
|
|
|
#include <linux/dma-mapping.h>
|
2006-01-16 23:03:41 -07:00
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/memory_hotplug.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
#include <asm/processor.h>
|
|
|
|
#include <asm/system.h>
|
|
|
|
#include <asm/uaccess.h>
|
|
|
|
#include <asm/pgtable.h>
|
|
|
|
#include <asm/pgalloc.h>
|
|
|
|
#include <asm/dma.h>
|
|
|
|
#include <asm/fixmap.h>
|
|
|
|
#include <asm/e820.h>
|
|
|
|
#include <asm/apic.h>
|
|
|
|
#include <asm/tlb.h>
|
|
|
|
#include <asm/mmu_context.h>
|
|
|
|
#include <asm/proto.h>
|
|
|
|
#include <asm/smp.h>
|
2005-11-05 09:25:53 -07:00
|
|
|
#include <asm/sections.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
#ifndef Dprintk
|
|
|
|
#define Dprintk(x...)
|
|
|
|
#endif
|
|
|
|
|
2006-01-11 14:44:42 -07:00
|
|
|
struct dma_mapping_ops* dma_ops;
|
|
|
|
EXPORT_SYMBOL(dma_ops);
|
|
|
|
|
2005-11-05 09:25:53 -07:00
|
|
|
static unsigned long dma_reserve __initdata;
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* NOTE: pagetable_init alloc all the fixmap pagetables contiguous on the
|
|
|
|
* physical space so we can cache the place of the first one and move
|
|
|
|
* around without checking the pgd every time.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void show_mem(void)
|
|
|
|
{
|
2005-09-12 10:49:24 -06:00
|
|
|
long i, total = 0, reserved = 0;
|
|
|
|
long shared = 0, cached = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
pg_data_t *pgdat;
|
|
|
|
struct page *page;
|
|
|
|
|
2005-09-12 10:49:24 -06:00
|
|
|
printk(KERN_INFO "Mem-info:\n");
|
2005-04-16 16:20:36 -06:00
|
|
|
show_free_areas();
|
2005-09-12 10:49:24 -06:00
|
|
|
printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2006-03-27 02:15:59 -07:00
|
|
|
for_each_online_pgdat(pgdat) {
|
2005-04-16 16:20:36 -06:00
|
|
|
for (i = 0; i < pgdat->node_spanned_pages; ++i) {
|
|
|
|
page = pfn_to_page(pgdat->node_start_pfn + i);
|
|
|
|
total++;
|
2005-09-12 10:49:24 -06:00
|
|
|
if (PageReserved(page))
|
|
|
|
reserved++;
|
|
|
|
else if (PageSwapCache(page))
|
|
|
|
cached++;
|
|
|
|
else if (page_count(page))
|
|
|
|
shared += page_count(page) - 1;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
}
|
2005-09-12 10:49:24 -06:00
|
|
|
printk(KERN_INFO "%lu pages of RAM\n", total);
|
|
|
|
printk(KERN_INFO "%lu reserved pages\n",reserved);
|
|
|
|
printk(KERN_INFO "%lu pages shared\n",shared);
|
|
|
|
printk(KERN_INFO "%lu pages swap cached\n",cached);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
int after_bootmem;
|
|
|
|
|
2006-03-25 08:30:25 -07:00
|
|
|
static __init void *spp_getpage(void)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
void *ptr;
|
|
|
|
if (after_bootmem)
|
|
|
|
ptr = (void *) get_zeroed_page(GFP_ATOMIC);
|
|
|
|
else
|
|
|
|
ptr = alloc_bootmem_pages(PAGE_SIZE);
|
|
|
|
if (!ptr || ((unsigned long)ptr & ~PAGE_MASK))
|
|
|
|
panic("set_pte_phys: cannot allocate page data %s\n", after_bootmem?"after bootmem":"");
|
|
|
|
|
|
|
|
Dprintk("spp_getpage %p\n", ptr);
|
|
|
|
return ptr;
|
|
|
|
}
|
|
|
|
|
2006-03-25 08:30:25 -07:00
|
|
|
static __init void set_pte_phys(unsigned long vaddr,
|
2005-04-16 16:20:36 -06:00
|
|
|
unsigned long phys, pgprot_t prot)
|
|
|
|
{
|
|
|
|
pgd_t *pgd;
|
|
|
|
pud_t *pud;
|
|
|
|
pmd_t *pmd;
|
|
|
|
pte_t *pte, new_pte;
|
|
|
|
|
|
|
|
Dprintk("set_pte_phys %lx to %lx\n", vaddr, phys);
|
|
|
|
|
|
|
|
pgd = pgd_offset_k(vaddr);
|
|
|
|
if (pgd_none(*pgd)) {
|
|
|
|
printk("PGD FIXMAP MISSING, it should be setup in head.S!\n");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
pud = pud_offset(pgd, vaddr);
|
|
|
|
if (pud_none(*pud)) {
|
|
|
|
pmd = (pmd_t *) spp_getpage();
|
|
|
|
set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
|
|
|
|
if (pmd != pmd_offset(pud, 0)) {
|
|
|
|
printk("PAGETABLE BUG #01! %p <-> %p\n", pmd, pmd_offset(pud,0));
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
pmd = pmd_offset(pud, vaddr);
|
|
|
|
if (pmd_none(*pmd)) {
|
|
|
|
pte = (pte_t *) spp_getpage();
|
|
|
|
set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
|
|
|
|
if (pte != pte_offset_kernel(pmd, 0)) {
|
|
|
|
printk("PAGETABLE BUG #02!\n");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
new_pte = pfn_pte(phys >> PAGE_SHIFT, prot);
|
|
|
|
|
|
|
|
pte = pte_offset_kernel(pmd, vaddr);
|
|
|
|
if (!pte_none(*pte) &&
|
|
|
|
pte_val(*pte) != (pte_val(new_pte) & __supported_pte_mask))
|
|
|
|
pte_ERROR(*pte);
|
|
|
|
set_pte(pte, new_pte);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* It's enough to flush this one mapping.
|
|
|
|
* (PGE mappings get flushed as well)
|
|
|
|
*/
|
|
|
|
__flush_tlb_one(vaddr);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* NOTE: this is meant to be run only at boot */
|
2006-03-25 08:30:25 -07:00
|
|
|
void __init
|
|
|
|
__set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t prot)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
unsigned long address = __fix_to_virt(idx);
|
|
|
|
|
|
|
|
if (idx >= __end_of_fixed_addresses) {
|
|
|
|
printk("Invalid __set_fixmap\n");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
set_pte_phys(address, phys, prot);
|
|
|
|
}
|
|
|
|
|
|
|
|
unsigned long __initdata table_start, table_end;
|
|
|
|
|
|
|
|
extern pmd_t temp_boot_pmds[];
|
|
|
|
|
|
|
|
static struct temp_map {
|
|
|
|
pmd_t *pmd;
|
|
|
|
void *address;
|
|
|
|
int allocated;
|
|
|
|
} temp_mappings[] __initdata = {
|
|
|
|
{ &temp_boot_pmds[0], (void *)(40UL * 1024 * 1024) },
|
|
|
|
{ &temp_boot_pmds[1], (void *)(42UL * 1024 * 1024) },
|
|
|
|
{}
|
|
|
|
};
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
static __meminit void *alloc_low_page(int *index, unsigned long *phys)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
struct temp_map *ti;
|
|
|
|
int i;
|
|
|
|
unsigned long pfn = table_end++, paddr;
|
|
|
|
void *adr;
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
if (after_bootmem) {
|
|
|
|
adr = (void *)get_zeroed_page(GFP_ATOMIC);
|
|
|
|
*phys = __pa(adr);
|
|
|
|
return adr;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
if (pfn >= end_pfn)
|
|
|
|
panic("alloc_low_page: ran out of memory");
|
|
|
|
for (i = 0; temp_mappings[i].allocated; i++) {
|
|
|
|
if (!temp_mappings[i].pmd)
|
|
|
|
panic("alloc_low_page: ran out of temp mappings");
|
|
|
|
}
|
|
|
|
ti = &temp_mappings[i];
|
|
|
|
paddr = (pfn << PAGE_SHIFT) & PMD_MASK;
|
|
|
|
set_pmd(ti->pmd, __pmd(paddr | _KERNPG_TABLE | _PAGE_PSE));
|
|
|
|
ti->allocated = 1;
|
|
|
|
__flush_tlb();
|
|
|
|
adr = ti->address + ((pfn << PAGE_SHIFT) & ~PMD_MASK);
|
2006-01-16 23:03:41 -07:00
|
|
|
memset(adr, 0, PAGE_SIZE);
|
2005-04-16 16:20:36 -06:00
|
|
|
*index = i;
|
|
|
|
*phys = pfn * PAGE_SIZE;
|
|
|
|
return adr;
|
|
|
|
}
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
static __meminit void unmap_low_page(int i)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-01-16 23:03:41 -07:00
|
|
|
struct temp_map *ti;
|
|
|
|
|
|
|
|
if (after_bootmem)
|
|
|
|
return;
|
|
|
|
|
|
|
|
ti = &temp_mappings[i];
|
2005-04-16 16:20:36 -06:00
|
|
|
set_pmd(ti->pmd, __pmd(0));
|
|
|
|
ti->allocated = 0;
|
|
|
|
}
|
|
|
|
|
2006-03-25 08:30:22 -07:00
|
|
|
/* Must run before zap_low_mappings */
|
|
|
|
__init void *early_ioremap(unsigned long addr, unsigned long size)
|
|
|
|
{
|
|
|
|
unsigned long map = round_down(addr, LARGE_PAGE_SIZE);
|
|
|
|
|
|
|
|
/* actually usually some more */
|
|
|
|
if (size >= LARGE_PAGE_SIZE) {
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
set_pmd(temp_mappings[0].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE));
|
|
|
|
map += LARGE_PAGE_SIZE;
|
|
|
|
set_pmd(temp_mappings[1].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE));
|
|
|
|
__flush_tlb();
|
|
|
|
return temp_mappings[0].address + (addr & (LARGE_PAGE_SIZE-1));
|
|
|
|
}
|
|
|
|
|
|
|
|
/* To avoid virtual aliases later */
|
|
|
|
__init void early_iounmap(void *addr, unsigned long size)
|
|
|
|
{
|
|
|
|
if ((void *)round_down((unsigned long)addr, LARGE_PAGE_SIZE) != temp_mappings[0].address)
|
|
|
|
printk("early_iounmap: bad address %p\n", addr);
|
|
|
|
set_pmd(temp_mappings[0].pmd, __pmd(0));
|
|
|
|
set_pmd(temp_mappings[1].pmd, __pmd(0));
|
|
|
|
__flush_tlb();
|
|
|
|
}
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
static void __meminit
|
2006-09-26 02:52:36 -06:00
|
|
|
phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end)
|
2006-01-16 23:03:41 -07:00
|
|
|
{
|
2006-09-26 02:52:36 -06:00
|
|
|
int i = pmd_index(address);
|
2006-01-16 23:03:41 -07:00
|
|
|
|
2006-09-26 02:52:36 -06:00
|
|
|
for (; i < PTRS_PER_PMD; i++, address += PMD_SIZE) {
|
2006-01-16 23:03:41 -07:00
|
|
|
unsigned long entry;
|
2006-09-26 02:52:36 -06:00
|
|
|
pmd_t *pmd = pmd_page + pmd_index(address);
|
2006-01-16 23:03:41 -07:00
|
|
|
|
2006-06-26 05:59:02 -06:00
|
|
|
if (address >= end) {
|
|
|
|
if (!after_bootmem)
|
|
|
|
for (; i < PTRS_PER_PMD; i++, pmd++)
|
|
|
|
set_pmd(pmd, __pmd(0));
|
2006-01-16 23:03:41 -07:00
|
|
|
break;
|
|
|
|
}
|
2006-09-26 02:52:36 -06:00
|
|
|
|
|
|
|
if (pmd_val(*pmd))
|
|
|
|
continue;
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
entry = _PAGE_NX|_PAGE_PSE|_KERNPG_TABLE|_PAGE_GLOBAL|address;
|
|
|
|
entry &= __supported_pte_mask;
|
|
|
|
set_pmd(pmd, __pmd(entry));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __meminit
|
|
|
|
phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end)
|
|
|
|
{
|
2006-09-26 02:52:36 -06:00
|
|
|
pmd_t *pmd = pmd_offset(pud,0);
|
|
|
|
spin_lock(&init_mm.page_table_lock);
|
|
|
|
phys_pmd_init(pmd, address, end);
|
|
|
|
spin_unlock(&init_mm.page_table_lock);
|
|
|
|
__flush_tlb_all();
|
2006-01-16 23:03:41 -07:00
|
|
|
}
|
|
|
|
|
2006-09-26 02:52:36 -06:00
|
|
|
static void __meminit phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-09-26 02:52:36 -06:00
|
|
|
int i = pud_index(addr);
|
2006-01-16 23:03:41 -07:00
|
|
|
|
|
|
|
|
2006-09-26 02:52:36 -06:00
|
|
|
for (; i < PTRS_PER_PUD; i++, addr = (addr & PUD_MASK) + PUD_SIZE ) {
|
2005-04-16 16:20:36 -06:00
|
|
|
int map;
|
2006-09-26 02:52:36 -06:00
|
|
|
unsigned long pmd_phys;
|
|
|
|
pud_t *pud = pud_page + pud_index(addr);
|
2005-04-16 16:20:36 -06:00
|
|
|
pmd_t *pmd;
|
|
|
|
|
2006-09-26 02:52:36 -06:00
|
|
|
if (addr >= end)
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
|
2006-09-26 02:52:36 -06:00
|
|
|
if (!after_bootmem && !e820_any_mapped(addr,addr+PUD_SIZE,0)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
set_pud(pud, __pud(0));
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2006-09-26 02:52:36 -06:00
|
|
|
if (pud_val(*pud)) {
|
|
|
|
phys_pmd_update(pud, addr, end);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
pmd = alloc_low_page(&map, &pmd_phys);
|
2006-01-16 23:03:41 -07:00
|
|
|
spin_lock(&init_mm.page_table_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
set_pud(pud, __pud(pmd_phys | _KERNPG_TABLE));
|
2006-09-26 02:52:36 -06:00
|
|
|
phys_pmd_init(pmd, addr, end);
|
2006-01-16 23:03:41 -07:00
|
|
|
spin_unlock(&init_mm.page_table_lock);
|
2005-04-16 16:20:36 -06:00
|
|
|
unmap_low_page(map);
|
|
|
|
}
|
|
|
|
__flush_tlb();
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __init find_early_table_space(unsigned long end)
|
|
|
|
{
|
2006-01-11 14:46:57 -07:00
|
|
|
unsigned long puds, pmds, tables, start;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
puds = (end + PUD_SIZE - 1) >> PUD_SHIFT;
|
|
|
|
pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT;
|
|
|
|
tables = round_up(puds * sizeof(pud_t), PAGE_SIZE) +
|
|
|
|
round_up(pmds * sizeof(pmd_t), PAGE_SIZE);
|
|
|
|
|
2006-01-15 17:56:51 -07:00
|
|
|
/* RED-PEN putting page tables only on node 0 could
|
|
|
|
cause a hotspot and fill up ZONE_DMA. The page tables
|
|
|
|
need roughly 0.5KB per GB. */
|
|
|
|
start = 0x8000;
|
|
|
|
table_start = find_e820_area(start, end, tables);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (table_start == -1UL)
|
|
|
|
panic("Cannot find space for the kernel page tables");
|
|
|
|
|
|
|
|
table_start >>= PAGE_SHIFT;
|
|
|
|
table_end = table_start;
|
2006-01-16 23:03:41 -07:00
|
|
|
|
|
|
|
early_printk("kernel direct mapping tables up to %lx @ %lx-%lx\n",
|
2006-06-26 05:59:02 -06:00
|
|
|
end, table_start << PAGE_SHIFT,
|
|
|
|
(table_start << PAGE_SHIFT) + tables);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Setup the direct mapping of the physical memory at PAGE_OFFSET.
|
|
|
|
This runs before bootmem is initialized and gets pages directly from the
|
|
|
|
physical memory. To access them they are temporarily mapped. */
|
2006-01-16 23:03:41 -07:00
|
|
|
void __meminit init_memory_mapping(unsigned long start, unsigned long end)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
unsigned long next;
|
|
|
|
|
|
|
|
Dprintk("init_memory_mapping\n");
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Find space for the kernel direct mapping tables.
|
|
|
|
* Later we should allocate these tables in the local node of the memory
|
|
|
|
* mapped. Unfortunately this is done currently before the nodes are
|
|
|
|
* discovered.
|
|
|
|
*/
|
2006-01-16 23:03:41 -07:00
|
|
|
if (!after_bootmem)
|
|
|
|
find_early_table_space(end);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
start = (unsigned long)__va(start);
|
|
|
|
end = (unsigned long)__va(end);
|
|
|
|
|
|
|
|
for (; start < end; start = next) {
|
|
|
|
int map;
|
|
|
|
unsigned long pud_phys;
|
2006-01-16 23:03:41 -07:00
|
|
|
pgd_t *pgd = pgd_offset_k(start);
|
|
|
|
pud_t *pud;
|
|
|
|
|
|
|
|
if (after_bootmem)
|
2006-06-26 05:57:56 -06:00
|
|
|
pud = pud_offset(pgd, start & PGDIR_MASK);
|
2006-01-16 23:03:41 -07:00
|
|
|
else
|
|
|
|
pud = alloc_low_page(&map, &pud_phys);
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
next = start + PGDIR_SIZE;
|
|
|
|
if (next > end)
|
|
|
|
next = end;
|
|
|
|
phys_pud_init(pud, __pa(start), __pa(next));
|
2006-01-16 23:03:41 -07:00
|
|
|
if (!after_bootmem)
|
|
|
|
set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys));
|
2005-04-16 16:20:36 -06:00
|
|
|
unmap_low_page(map);
|
|
|
|
}
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
if (!after_bootmem)
|
|
|
|
asm volatile("movq %%cr4,%0" : "=r" (mmu_cr4_features));
|
2005-04-16 16:20:36 -06:00
|
|
|
__flush_tlb_all();
|
|
|
|
}
|
|
|
|
|
2005-11-05 09:25:53 -07:00
|
|
|
void __cpuinit zap_low_mappings(int cpu)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2005-11-05 09:25:53 -07:00
|
|
|
if (cpu == 0) {
|
|
|
|
pgd_t *pgd = pgd_offset_k(0UL);
|
|
|
|
pgd_clear(pgd);
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* For AP's, zap the low identity mappings by changing the cr3
|
|
|
|
* to init_level4_pgt and doing local flush tlb all
|
|
|
|
*/
|
|
|
|
asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
|
|
|
|
}
|
|
|
|
__flush_tlb_all();
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2005-06-23 01:08:06 -06:00
|
|
|
#ifndef CONFIG_NUMA
|
2005-04-16 16:20:36 -06:00
|
|
|
void __init paging_init(void)
|
|
|
|
{
|
2006-10-11 02:20:39 -06:00
|
|
|
unsigned long max_zone_pfns[MAX_NR_ZONES];
|
|
|
|
memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
|
|
|
|
max_zone_pfns[ZONE_DMA] = MAX_DMA_PFN;
|
|
|
|
max_zone_pfns[ZONE_DMA32] = MAX_DMA32_PFN;
|
|
|
|
max_zone_pfns[ZONE_NORMAL] = end_pfn;
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
memory_present(0, 0, end_pfn);
|
|
|
|
sparse_init();
|
2006-09-27 02:49:52 -06:00
|
|
|
free_area_init_nodes(max_zone_pfns);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/* Unmap a kernel mapping if it exists. This is useful to avoid prefetches
|
|
|
|
from the CPU leading to inconsistent cache lines. address and size
|
|
|
|
must be aligned to 2MB boundaries.
|
|
|
|
Does nothing when the mapping doesn't exist. */
|
|
|
|
void __init clear_kernel_mapping(unsigned long address, unsigned long size)
|
|
|
|
{
|
|
|
|
unsigned long end = address + size;
|
|
|
|
|
|
|
|
BUG_ON(address & ~LARGE_PAGE_MASK);
|
|
|
|
BUG_ON(size & ~LARGE_PAGE_MASK);
|
|
|
|
|
|
|
|
for (; address < end; address += LARGE_PAGE_SIZE) {
|
|
|
|
pgd_t *pgd = pgd_offset_k(address);
|
|
|
|
pud_t *pud;
|
|
|
|
pmd_t *pmd;
|
|
|
|
if (pgd_none(*pgd))
|
|
|
|
continue;
|
|
|
|
pud = pud_offset(pgd, address);
|
|
|
|
if (pud_none(*pud))
|
|
|
|
continue;
|
|
|
|
pmd = pmd_offset(pud, address);
|
|
|
|
if (!pmd || pmd_none(*pmd))
|
|
|
|
continue;
|
|
|
|
if (0 == (pmd_val(*pmd) & _PAGE_PSE)) {
|
|
|
|
/* Could handle this, but it should not happen currently. */
|
|
|
|
printk(KERN_ERR
|
|
|
|
"clear_kernel_mapping: mapping has been split. will leak memory\n");
|
|
|
|
pmd_ERROR(*pmd);
|
|
|
|
}
|
|
|
|
set_pmd(pmd, __pmd(0));
|
|
|
|
}
|
|
|
|
__flush_tlb_all();
|
|
|
|
}
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
/*
|
|
|
|
* Memory hotplug specific functions
|
|
|
|
*/
|
|
|
|
void online_page(struct page *page)
|
|
|
|
{
|
|
|
|
ClearPageReserved(page);
|
2006-03-22 01:08:40 -07:00
|
|
|
init_page_count(page);
|
2006-01-16 23:03:41 -07:00
|
|
|
__free_page(page);
|
|
|
|
totalram_pages++;
|
|
|
|
num_physpages++;
|
|
|
|
}
|
|
|
|
|
2006-06-27 03:53:30 -06:00
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
2006-04-07 11:49:15 -06:00
|
|
|
/*
|
|
|
|
* Memory is added always to NORMAL zone. This means you will never get
|
|
|
|
* additional DMA/DMA32 memory.
|
|
|
|
*/
|
2006-06-27 03:53:30 -06:00
|
|
|
int arch_add_memory(int nid, u64 start, u64 size)
|
2006-01-16 23:03:41 -07:00
|
|
|
{
|
2006-06-27 03:53:30 -06:00
|
|
|
struct pglist_data *pgdat = NODE_DATA(nid);
|
2006-09-26 00:31:09 -06:00
|
|
|
struct zone *zone = pgdat->node_zones + ZONE_NORMAL;
|
2006-01-16 23:03:41 -07:00
|
|
|
unsigned long start_pfn = start >> PAGE_SHIFT;
|
|
|
|
unsigned long nr_pages = size >> PAGE_SHIFT;
|
|
|
|
int ret;
|
|
|
|
|
2006-10-01 00:27:09 -06:00
|
|
|
init_memory_mapping(start, (start + size -1));
|
|
|
|
|
2006-01-16 23:03:41 -07:00
|
|
|
ret = __add_pages(zone, start_pfn, nr_pages);
|
|
|
|
if (ret)
|
|
|
|
goto error;
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
error:
|
|
|
|
printk("%s: Problem encountered in __add_pages!\n", __func__);
|
|
|
|
return ret;
|
|
|
|
}
|
2006-06-27 03:53:30 -06:00
|
|
|
EXPORT_SYMBOL_GPL(arch_add_memory);
|
2006-01-16 23:03:41 -07:00
|
|
|
|
|
|
|
int remove_memory(u64 start, u64 size)
|
|
|
|
{
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(remove_memory);
|
|
|
|
|
2006-11-18 23:19:40 -07:00
|
|
|
#if !defined(CONFIG_ACPI_NUMA) && defined(CONFIG_NUMA)
|
2006-10-01 00:27:06 -06:00
|
|
|
int memory_add_physaddr_to_nid(u64 start)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
2006-10-01 00:27:07 -06:00
|
|
|
EXPORT_SYMBOL_GPL(memory_add_physaddr_to_nid);
|
2006-10-01 00:27:06 -06:00
|
|
|
#endif
|
|
|
|
|
2006-10-01 00:27:09 -06:00
|
|
|
#endif /* CONFIG_MEMORY_HOTPLUG */
|
|
|
|
|
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE
|
2006-06-27 03:53:30 -06:00
|
|
|
/*
|
|
|
|
* Memory Hotadd without sparsemem. The mem_maps have been allocated in advance,
|
|
|
|
* just online the pages.
|
|
|
|
*/
|
|
|
|
int __add_pages(struct zone *z, unsigned long start_pfn, unsigned long nr_pages)
|
|
|
|
{
|
|
|
|
int err = -EIO;
|
|
|
|
unsigned long pfn;
|
|
|
|
unsigned long total = 0, mem = 0;
|
|
|
|
for (pfn = start_pfn; pfn < start_pfn + nr_pages; pfn++) {
|
|
|
|
if (pfn_valid(pfn)) {
|
|
|
|
online_page(pfn_to_page(pfn));
|
|
|
|
err = 0;
|
|
|
|
mem++;
|
|
|
|
}
|
|
|
|
total++;
|
|
|
|
}
|
|
|
|
if (!err) {
|
|
|
|
z->spanned_pages += total;
|
|
|
|
z->present_pages += mem;
|
|
|
|
z->zone_pgdat->node_spanned_pages += total;
|
|
|
|
z->zone_pgdat->node_present_pages += mem;
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
2006-10-01 00:27:09 -06:00
|
|
|
#endif
|
2006-01-16 23:03:41 -07:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel, kcore_modules,
|
|
|
|
kcore_vsyscall;
|
|
|
|
|
|
|
|
void __init mem_init(void)
|
|
|
|
{
|
2005-09-12 10:49:24 -06:00
|
|
|
long codesize, reservedpages, datasize, initsize;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2006-06-26 05:58:11 -06:00
|
|
|
pci_iommu_alloc();
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/* clear the zero-page */
|
|
|
|
memset(empty_zero_page, 0, PAGE_SIZE);
|
|
|
|
|
|
|
|
reservedpages = 0;
|
|
|
|
|
|
|
|
/* this will put all low memory onto the freelists */
|
2005-06-23 01:08:06 -06:00
|
|
|
#ifdef CONFIG_NUMA
|
2005-09-12 10:49:24 -06:00
|
|
|
totalram_pages = numa_free_all_bootmem();
|
2005-04-16 16:20:36 -06:00
|
|
|
#else
|
2005-09-12 10:49:24 -06:00
|
|
|
totalram_pages = free_all_bootmem();
|
2005-04-16 16:20:36 -06:00
|
|
|
#endif
|
2006-09-27 02:49:52 -06:00
|
|
|
reservedpages = end_pfn - totalram_pages -
|
|
|
|
absent_pages_in_range(0, end_pfn);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
after_bootmem = 1;
|
|
|
|
|
|
|
|
codesize = (unsigned long) &_etext - (unsigned long) &_text;
|
|
|
|
datasize = (unsigned long) &_edata - (unsigned long) &_etext;
|
|
|
|
initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
|
|
|
|
|
|
|
|
/* Register memory areas for /proc/kcore */
|
|
|
|
kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
|
|
|
|
kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
|
|
|
|
VMALLOC_END-VMALLOC_START);
|
|
|
|
kclist_add(&kcore_kernel, &_stext, _end - _stext);
|
|
|
|
kclist_add(&kcore_modules, (void *)MODULES_VADDR, MODULES_LEN);
|
|
|
|
kclist_add(&kcore_vsyscall, (void *)VSYSCALL_START,
|
|
|
|
VSYSCALL_END - VSYSCALL_START);
|
|
|
|
|
2005-09-12 10:49:24 -06:00
|
|
|
printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, %ldk data, %ldk init)\n",
|
2005-04-16 16:20:36 -06:00
|
|
|
(unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
|
|
|
|
end_pfn << (PAGE_SHIFT-10),
|
|
|
|
codesize >> 10,
|
|
|
|
reservedpages << (PAGE_SHIFT-10),
|
|
|
|
datasize >> 10,
|
|
|
|
initsize >> 10);
|
|
|
|
|
2005-11-05 09:25:53 -07:00
|
|
|
#ifdef CONFIG_SMP
|
2005-04-16 16:20:36 -06:00
|
|
|
/*
|
2005-11-05 09:25:53 -07:00
|
|
|
* Sync boot_level4_pgt mappings with the init_level4_pgt
|
|
|
|
* except for the low identity mappings which are already zapped
|
|
|
|
* in init_level4_pgt. This sync-up is essential for AP's bringup
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
2005-11-05 09:25:53 -07:00
|
|
|
memcpy(boot_level4_pgt+1, init_level4_pgt+1, (PTRS_PER_PGD-1)*sizeof(pgd_t));
|
2005-04-16 16:20:36 -06:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2006-06-26 05:56:16 -06:00
|
|
|
void free_init_pages(char *what, unsigned long begin, unsigned long end)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
unsigned long addr;
|
|
|
|
|
2006-06-26 05:56:16 -06:00
|
|
|
if (begin >= end)
|
|
|
|
return;
|
|
|
|
|
|
|
|
printk(KERN_INFO "Freeing %s: %ldk freed\n", what, (end - begin) >> 10);
|
|
|
|
for (addr = begin; addr < end; addr += PAGE_SIZE) {
|
2005-04-16 16:20:36 -06:00
|
|
|
ClearPageReserved(virt_to_page(addr));
|
2006-03-22 01:08:40 -07:00
|
|
|
init_page_count(virt_to_page(addr));
|
2006-06-27 03:53:52 -06:00
|
|
|
memset((void *)(addr & ~(PAGE_SIZE-1)),
|
|
|
|
POISON_FREE_INITMEM, PAGE_SIZE);
|
2005-04-16 16:20:36 -06:00
|
|
|
free_page(addr);
|
|
|
|
totalram_pages++;
|
|
|
|
}
|
2006-06-26 05:56:16 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
void free_initmem(void)
|
|
|
|
{
|
2006-06-27 03:53:52 -06:00
|
|
|
memset(__initdata_begin, POISON_FREE_INITDATA,
|
|
|
|
__initdata_end - __initdata_begin);
|
2006-06-26 05:56:16 -06:00
|
|
|
free_init_pages("unused kernel memory",
|
|
|
|
(unsigned long)(&__init_begin),
|
|
|
|
(unsigned long)(&__init_end));
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2006-01-06 01:12:04 -07:00
|
|
|
#ifdef CONFIG_DEBUG_RODATA
|
|
|
|
|
|
|
|
void mark_rodata_ro(void)
|
|
|
|
{
|
2006-07-01 05:36:30 -06:00
|
|
|
unsigned long addr = (unsigned long)__start_rodata;
|
2006-01-06 01:12:04 -07:00
|
|
|
|
2006-07-01 05:36:30 -06:00
|
|
|
for (; addr < (unsigned long)__end_rodata; addr += PAGE_SIZE)
|
2006-01-06 01:12:04 -07:00
|
|
|
change_page_attr_addr(addr, 1, PAGE_KERNEL_RO);
|
|
|
|
|
|
|
|
printk ("Write protecting the kernel read-only data: %luk\n",
|
2006-07-01 05:36:30 -06:00
|
|
|
(__end_rodata - __start_rodata) >> 10);
|
2006-01-06 01:12:04 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* change_page_attr_addr() requires a global_flush_tlb() call after it.
|
|
|
|
* We do this after the printk so that if something went wrong in the
|
|
|
|
* change, the printk gets out at least to give a better debug hint
|
|
|
|
* of who is the culprit.
|
|
|
|
*/
|
|
|
|
global_flush_tlb();
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
#ifdef CONFIG_BLK_DEV_INITRD
|
|
|
|
void free_initrd_mem(unsigned long start, unsigned long end)
|
|
|
|
{
|
2006-06-26 05:56:16 -06:00
|
|
|
free_init_pages("initrd memory", start, end);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
void __init reserve_bootmem_generic(unsigned long phys, unsigned len)
|
|
|
|
{
|
2005-06-23 01:08:06 -06:00
|
|
|
#ifdef CONFIG_NUMA
|
2005-04-16 16:20:36 -06:00
|
|
|
int nid = phys_to_nid(phys);
|
2006-11-14 08:57:46 -07:00
|
|
|
#endif
|
|
|
|
unsigned long pfn = phys >> PAGE_SHIFT;
|
|
|
|
if (pfn >= end_pfn) {
|
|
|
|
/* This can happen with kdump kernels when accessing firmware
|
|
|
|
tables. */
|
|
|
|
if (pfn < end_pfn_map)
|
|
|
|
return;
|
|
|
|
printk(KERN_ERR "reserve_bootmem: illegal reserve %lx %u\n",
|
|
|
|
phys, len);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Should check here against the e820 map to avoid double free */
|
|
|
|
#ifdef CONFIG_NUMA
|
2005-04-16 16:20:36 -06:00
|
|
|
reserve_bootmem_node(NODE_DATA(nid), phys, len);
|
|
|
|
#else
|
|
|
|
reserve_bootmem(phys, len);
|
|
|
|
#endif
|
2006-09-27 02:49:56 -06:00
|
|
|
if (phys+len <= MAX_DMA_PFN*PAGE_SIZE) {
|
2005-11-05 09:25:53 -07:00
|
|
|
dma_reserve += len / PAGE_SIZE;
|
2006-09-27 02:49:56 -06:00
|
|
|
set_dma_reserve(dma_reserve);
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
int kern_addr_valid(unsigned long addr)
|
|
|
|
{
|
|
|
|
unsigned long above = ((long)addr) >> __VIRTUAL_MASK_SHIFT;
|
|
|
|
pgd_t *pgd;
|
|
|
|
pud_t *pud;
|
|
|
|
pmd_t *pmd;
|
|
|
|
pte_t *pte;
|
|
|
|
|
|
|
|
if (above != 0 && above != -1UL)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
pgd = pgd_offset_k(addr);
|
|
|
|
if (pgd_none(*pgd))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
pud = pud_offset(pgd, addr);
|
|
|
|
if (pud_none(*pud))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
pmd = pmd_offset(pud, addr);
|
|
|
|
if (pmd_none(*pmd))
|
|
|
|
return 0;
|
|
|
|
if (pmd_large(*pmd))
|
|
|
|
return pfn_valid(pmd_pfn(*pmd));
|
|
|
|
|
|
|
|
pte = pte_offset_kernel(pmd, addr);
|
|
|
|
if (pte_none(*pte))
|
|
|
|
return 0;
|
|
|
|
return pfn_valid(pte_pfn(*pte));
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_SYSCTL
|
|
|
|
#include <linux/sysctl.h>
|
|
|
|
|
|
|
|
extern int exception_trace, page_fault_trace;
|
|
|
|
|
|
|
|
static ctl_table debug_table2[] = {
|
|
|
|
{ 99, "exception-trace", &exception_trace, sizeof(int), 0644, NULL,
|
|
|
|
proc_dointvec },
|
|
|
|
{ 0, }
|
|
|
|
};
|
|
|
|
|
|
|
|
static ctl_table debug_root_table2[] = {
|
|
|
|
{ .ctl_name = CTL_DEBUG, .procname = "debug", .mode = 0555,
|
|
|
|
.child = debug_table2 },
|
|
|
|
{ 0 },
|
|
|
|
};
|
|
|
|
|
|
|
|
static __init int x8664_sysctl_init(void)
|
|
|
|
{
|
|
|
|
register_sysctl_table(debug_root_table2, 1);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
__initcall(x8664_sysctl_init);
|
|
|
|
#endif
|
|
|
|
|
2006-12-06 18:14:09 -07:00
|
|
|
/* A pseudo VMA to allow ptrace access for the vsyscall page. This only
|
2005-04-16 16:24:55 -06:00
|
|
|
covers the 64bit vsyscall page now. 32bit has a real VMA now and does
|
|
|
|
not need special handling anymore. */
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
static struct vm_area_struct gate_vma = {
|
|
|
|
.vm_start = VSYSCALL_START,
|
2006-12-06 18:14:09 -07:00
|
|
|
.vm_end = VSYSCALL_START + (VSYSCALL_MAPPED_PAGES << PAGE_SHIFT),
|
|
|
|
.vm_page_prot = PAGE_READONLY_EXEC,
|
|
|
|
.vm_flags = VM_READ | VM_EXEC
|
2005-04-16 16:20:36 -06:00
|
|
|
};
|
|
|
|
|
|
|
|
struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_IA32_EMULATION
|
2005-04-16 16:24:55 -06:00
|
|
|
if (test_tsk_thread_flag(tsk, TIF_IA32))
|
|
|
|
return NULL;
|
2005-04-16 16:20:36 -06:00
|
|
|
#endif
|
|
|
|
return &gate_vma;
|
|
|
|
}
|
|
|
|
|
|
|
|
int in_gate_area(struct task_struct *task, unsigned long addr)
|
|
|
|
{
|
|
|
|
struct vm_area_struct *vma = get_gate_vma(task);
|
2005-04-16 16:24:55 -06:00
|
|
|
if (!vma)
|
|
|
|
return 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
return (addr >= vma->vm_start) && (addr < vma->vm_end);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Use this when you have no reliable task/vma, typically from interrupt
|
|
|
|
* context. It is less reliable than using the task's vma and may give
|
|
|
|
* false positives.
|
|
|
|
*/
|
|
|
|
int in_gate_area_no_task(unsigned long addr)
|
|
|
|
{
|
2005-04-16 16:24:55 -06:00
|
|
|
return (addr >= VSYSCALL_START) && (addr < VSYSCALL_END);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|