linux-old/include/linux/mm.h
<<
>>
Prefs
   1#ifndef _LINUX_MM_H
   2#define _LINUX_MM_H
   3
   4#include <linux/sched.h>
   5#include <linux/errno.h>
   6
   7#ifdef __KERNEL__
   8
   9#include <linux/config.h>
  10#include <linux/string.h>
  11#include <linux/list.h>
  12#include <linux/mmzone.h>
  13#include <linux/swap.h>
  14#include <linux/rbtree.h>
  15
  16extern unsigned long max_mapnr;
  17extern unsigned long num_physpages;
  18extern unsigned long num_mappedpages;
  19extern void * high_memory;
  20extern int page_cluster;
  21/* The inactive_clean lists are per zone. */
  22extern struct list_head active_list;
  23extern struct list_head inactive_list;
  24
  25#include <asm/page.h>
  26#include <asm/pgtable.h>
  27#include <asm/atomic.h>
  28
  29/*
  30 * Linux kernel virtual memory manager primitives.
  31 * The idea being to have a "virtual" mm in the same way
  32 * we have a virtual fs - giving a cleaner interface to the
  33 * mm details, and allowing different kinds of memory mappings
  34 * (from shared memory to executable loading to arbitrary
  35 * mmap() functions).
  36 */
  37
  38/*
  39 * This struct defines a memory VMM memory area. There is one of these
  40 * per VM-area/task.  A VM area is any part of the process virtual memory
  41 * space that has a special rule for the page-fault handlers (ie a shared
  42 * library, the executable area etc).
  43 */
  44struct vm_area_struct {
  45        struct mm_struct * vm_mm;       /* The address space we belong to. */
  46        unsigned long vm_start;         /* Our start address within vm_mm. */
  47        unsigned long vm_end;           /* The first byte after our end address
  48                                           within vm_mm. */
  49
  50        /* linked list of VM areas per task, sorted by address */
  51        struct vm_area_struct *vm_next;
  52
  53        pgprot_t vm_page_prot;          /* Access permissions of this VMA. */
  54        unsigned long vm_flags;         /* Flags, listed below. */
  55
  56        rb_node_t vm_rb;
  57
  58        /*
  59         * For areas with an address space and backing store,
  60         * one of the address_space->i_mmap{,shared} lists,
  61         * for shm areas, the list of attaches, otherwise unused.
  62         */
  63        struct vm_area_struct *vm_next_share;
  64        struct vm_area_struct **vm_pprev_share;
  65
  66        /* Function pointers to deal with this struct. */
  67        struct vm_operations_struct * vm_ops;
  68
  69        /* Information about our backing store: */
  70        unsigned long vm_pgoff;         /* Offset (within vm_file) in PAGE_SIZE
  71                                           units, *not* PAGE_CACHE_SIZE */
  72        struct file * vm_file;          /* File we map to (can be NULL). */
  73        unsigned long vm_raend;         /* XXX: put full readahead info here. */
  74        void * vm_private_data;         /* was vm_pte (shared mem) */
  75};
  76
  77/*
  78 * vm_flags..
  79 */
  80#define VM_READ         0x00000001      /* currently active flags */
  81#define VM_WRITE        0x00000002
  82#define VM_EXEC         0x00000004
  83#define VM_SHARED       0x00000008
  84
  85#define VM_MAYREAD      0x00000010      /* limits for mprotect() etc */
  86#define VM_MAYWRITE     0x00000020
  87#define VM_MAYEXEC      0x00000040
  88#define VM_MAYSHARE     0x00000080
  89
  90#define VM_GROWSDOWN    0x00000100      /* general info on the segment */
  91#define VM_GROWSUP      0x00000200
  92#define VM_SHM          0x00000400      /* shared memory area, don't swap out */
  93#define VM_DENYWRITE    0x00000800      /* ETXTBSY on write attempts.. */
  94
  95#define VM_EXECUTABLE   0x00001000
  96#define VM_LOCKED       0x00002000
  97#define VM_IO           0x00004000      /* Memory mapped I/O or similar */
  98
  99                                        /* Used by sys_madvise() */
 100#define VM_SEQ_READ     0x00008000      /* App will access data sequentially */
 101#define VM_RAND_READ    0x00010000      /* App will not benefit from clustered reads */
 102
 103#define VM_DONTCOPY     0x00020000      /* Do not copy this vma on fork */
 104#define VM_DONTEXPAND   0x00040000      /* Cannot expand with mremap() */
 105#define VM_RESERVED     0x00080000      /* Don't unmap it from swap_out */
 106
 107#ifndef VM_STACK_FLAGS
 108#define VM_STACK_FLAGS  0x00000177
 109#endif
 110
 111#define VM_READHINTMASK                 (VM_SEQ_READ | VM_RAND_READ)
 112#define VM_ClearReadHint(v)             (v)->vm_flags &= ~VM_READHINTMASK
 113#define VM_NormalReadHint(v)            (!((v)->vm_flags & VM_READHINTMASK))
 114#define VM_SequentialReadHint(v)        ((v)->vm_flags & VM_SEQ_READ)
 115#define VM_RandomReadHint(v)            ((v)->vm_flags & VM_RAND_READ)
 116
 117/* read ahead limits */
 118extern int vm_min_readahead;
 119extern int vm_max_readahead;
 120
 121/*
 122 * mapping from the currently active vm_flags protection bits (the
 123 * low four bits) to a page protection mask..
 124 */
 125extern pgprot_t protection_map[16];
 126
 127
 128/*
 129 * These are the virtual MM functions - opening of an area, closing and
 130 * unmapping it (needed to keep files on disk up-to-date etc), pointer
 131 * to the functions called when a no-page or a wp-page exception occurs. 
 132 */
 133struct vm_operations_struct {
 134        void (*open)(struct vm_area_struct * area);
 135        void (*close)(struct vm_area_struct * area);
 136        struct page * (*nopage)(struct vm_area_struct * area, unsigned long address, int unused);
 137};
 138
 139/*
 140 * Each physical page in the system has a struct page associated with
 141 * it to keep track of whatever it is we are using the page for at the
 142 * moment. Note that we have no way to track which tasks are using
 143 * a page.
 144 *
 145 * Try to keep the most commonly accessed fields in single cache lines
 146 * here (16 bytes or greater).  This ordering should be particularly
 147 * beneficial on 32-bit processors.
 148 *
 149 * The first line is data used in page cache lookup, the second line
 150 * is used for linear searches (eg. clock algorithm scans). 
 151 *
 152 * TODO: make this structure smaller, it could be as small as 32 bytes.
 153 */
 154typedef struct page {
 155        struct list_head list;          /* ->mapping has some page lists. */
 156        struct address_space *mapping;  /* The inode (or ...) we belong to. */
 157        unsigned long index;            /* Our offset within mapping. */
 158        struct page *next_hash;         /* Next page sharing our hash bucket in
 159                                           the pagecache hash table. */
 160        atomic_t count;                 /* Usage count, see below. */
 161        unsigned long flags;            /* atomic flags, some possibly
 162                                           updated asynchronously */
 163        struct list_head lru;           /* Pageout list, eg. active_list;
 164                                           protected by pagemap_lru_lock !! */
 165        struct page **pprev_hash;       /* Complement to *next_hash. */
 166        struct buffer_head * buffers;   /* Buffer maps us to a disk block. */
 167
 168        /*
 169         * On machines where all RAM is mapped into kernel address space,
 170         * we can simply calculate the virtual address. On machines with
 171         * highmem some memory is mapped into kernel virtual memory
 172         * dynamically, so we need a place to store that address.
 173         * Note that this field could be 16 bits on x86 ... ;)
 174         *
 175         * Architectures with slow multiplication can define
 176         * WANT_PAGE_VIRTUAL in asm/page.h
 177         */
 178#if defined(CONFIG_HIGHMEM) || defined(WANT_PAGE_VIRTUAL)
 179        void *virtual;                  /* Kernel virtual address (NULL if
 180                                           not kmapped, ie. highmem) */
 181#endif /* CONFIG_HIGMEM || WANT_PAGE_VIRTUAL */
 182} mem_map_t;
 183
 184/*
 185 * Methods to modify the page usage count.
 186 *
 187 * What counts for a page usage:
 188 * - cache mapping   (page->mapping)
 189 * - disk mapping    (page->buffers)
 190 * - page mapped in a task's page tables, each mapping
 191 *   is counted separately
 192 *
 193 * Also, many kernel routines increase the page count before a critical
 194 * routine so they can be sure the page doesn't go away from under them.
 195 */
 196#define get_page(p)             atomic_inc(&(p)->count)
 197#define put_page(p)             __free_page(p)
 198#define put_page_testzero(p)    atomic_dec_and_test(&(p)->count)
 199#define page_count(p)           atomic_read(&(p)->count)
 200#define set_page_count(p,v)     atomic_set(&(p)->count, v)
 201
 202static inline struct page *nth_page(struct page *page, int n)
 203{
 204        return page + n;
 205}
 206
 207/*
 208 * Various page->flags bits:
 209 *
 210 * PG_reserved is set for special pages, which can never be swapped
 211 * out. Some of them might not even exist (eg empty_bad_page)...
 212 *
 213 * Multiple processes may "see" the same page. E.g. for untouched
 214 * mappings of /dev/null, all processes see the same page full of
 215 * zeroes, and text pages of executables and shared libraries have
 216 * only one copy in memory, at most, normally.
 217 *
 218 * For the non-reserved pages, page->count denotes a reference count.
 219 *   page->count == 0 means the page is free.
 220 *   page->count == 1 means the page is used for exactly one purpose
 221 *   (e.g. a private data page of one process).
 222 *
 223 * A page may be used for kmalloc() or anyone else who does a
 224 * __get_free_page(). In this case the page->count is at least 1, and
 225 * all other fields are unused but should be 0 or NULL. The
 226 * management of this page is the responsibility of the one who uses
 227 * it.
 228 *
 229 * The other pages (we may call them "process pages") are completely
 230 * managed by the Linux memory manager: I/O, buffers, swapping etc.
 231 * The following discussion applies only to them.
 232 *
 233 * A page may belong to an inode's memory mapping. In this case,
 234 * page->mapping is the pointer to the inode, and page->index is the
 235 * file offset of the page, in units of PAGE_CACHE_SIZE.
 236 *
 237 * A page may have buffers allocated to it. In this case,
 238 * page->buffers is a circular list of these buffer heads. Else,
 239 * page->buffers == NULL.
 240 *
 241 * For pages belonging to inodes, the page->count is the number of
 242 * attaches, plus 1 if buffers are allocated to the page, plus one
 243 * for the page cache itself.
 244 *
 245 * All pages belonging to an inode are in these doubly linked lists:
 246 * mapping->clean_pages, mapping->dirty_pages and mapping->locked_pages;
 247 * using the page->list list_head. These fields are also used for
 248 * freelist managemet (when page->count==0).
 249 *
 250 * There is also a hash table mapping (mapping,index) to the page
 251 * in memory if present. The lists for this hash table use the fields
 252 * page->next_hash and page->pprev_hash.
 253 *
 254 * All process pages can do I/O:
 255 * - inode pages may need to be read from disk,
 256 * - inode pages which have been modified and are MAP_SHARED may need
 257 *   to be written to disk,
 258 * - private pages which have been modified may need to be swapped out
 259 *   to swap space and (later) to be read back into memory.
 260 * During disk I/O, PG_locked is used. This bit is set before I/O
 261 * and reset when I/O completes. page_waitqueue(page) is a wait queue of all
 262 * tasks waiting for the I/O on this page to complete.
 263 * PG_uptodate tells whether the page's contents is valid.
 264 * When a read completes, the page becomes uptodate, unless a disk I/O
 265 * error happened.
 266 *
 267 * For choosing which pages to swap out, inode pages carry a
 268 * PG_referenced bit, which is set any time the system accesses
 269 * that page through the (mapping,index) hash table. This referenced
 270 * bit, together with the referenced bit in the page tables, is used
 271 * to manipulate page->age and move the page across the active,
 272 * inactive_dirty and inactive_clean lists.
 273 *
 274 * Note that the referenced bit, the page->lru list_head and the
 275 * active, inactive_dirty and inactive_clean lists are protected by
 276 * the pagemap_lru_lock, and *NOT* by the usual PG_locked bit!
 277 *
 278 * PG_skip is used on sparc/sparc64 architectures to "skip" certain
 279 * parts of the address space.
 280 *
 281 * PG_error is set to indicate that an I/O error occurred on this page.
 282 *
 283 * PG_arch_1 is an architecture specific page state bit.  The generic
 284 * code guarantees that this bit is cleared for a page when it first
 285 * is entered into the page cache.
 286 *
 287 * PG_highmem pages are not permanently mapped into the kernel virtual
 288 * address space, they need to be kmapped separately for doing IO on
 289 * the pages. The struct page (these bits with information) are always
 290 * mapped into kernel address space...
 291 */
 292#define PG_locked                0      /* Page is locked. Don't touch. */
 293#define PG_error                 1
 294#define PG_referenced            2
 295#define PG_uptodate              3
 296#define PG_dirty                 4
 297#define PG_unused                5
 298#define PG_lru                   6
 299#define PG_active                7
 300#define PG_slab                  8
 301#define PG_skip                 10
 302#define PG_highmem              11
 303#define PG_checked              12      /* kill me in 2.5.<early>. */
 304#define PG_arch_1               13
 305#define PG_reserved             14
 306#define PG_launder              15      /* written out by VM pressure.. */
 307#define PG_fs_1                 16      /* Filesystem specific */
 308
 309#ifndef arch_set_page_uptodate
 310#define arch_set_page_uptodate(page)
 311#endif
 312
 313/* Make it prettier to test the above... */
 314#define UnlockPage(page)        unlock_page(page)
 315#define Page_Uptodate(page)     test_bit(PG_uptodate, &(page)->flags)
 316#ifndef SetPageUptodate
 317#define SetPageUptodate(page)   set_bit(PG_uptodate, &(page)->flags)
 318#endif
 319#define ClearPageUptodate(page) clear_bit(PG_uptodate, &(page)->flags)
 320#define PageDirty(page)         test_bit(PG_dirty, &(page)->flags)
 321#define SetPageDirty(page)      set_bit(PG_dirty, &(page)->flags)
 322#define ClearPageDirty(page)    clear_bit(PG_dirty, &(page)->flags)
 323#define PageLocked(page)        test_bit(PG_locked, &(page)->flags)
 324#define LockPage(page)          set_bit(PG_locked, &(page)->flags)
 325#define TryLockPage(page)       test_and_set_bit(PG_locked, &(page)->flags)
 326#define PageChecked(page)       test_bit(PG_checked, &(page)->flags)
 327#define SetPageChecked(page)    set_bit(PG_checked, &(page)->flags)
 328#define ClearPageChecked(page)  clear_bit(PG_checked, &(page)->flags)
 329#define PageLaunder(page)       test_bit(PG_launder, &(page)->flags)
 330#define SetPageLaunder(page)    set_bit(PG_launder, &(page)->flags)
 331#define ClearPageLaunder(page)  clear_bit(PG_launder, &(page)->flags)
 332#define ClearPageArch1(page)    clear_bit(PG_arch_1, &(page)->flags)
 333
 334/*
 335 * The zone field is never updated after free_area_init_core()
 336 * sets it, so none of the operations on it need to be atomic.
 337 */
 338#define NODE_SHIFT 4
 339#define ZONE_SHIFT (BITS_PER_LONG - 8)
 340
 341struct zone_struct;
 342extern struct zone_struct *zone_table[];
 343
 344static inline zone_t *page_zone(struct page *page)
 345{
 346        return zone_table[page->flags >> ZONE_SHIFT];
 347}
 348
 349static inline void set_page_zone(struct page *page, unsigned long zone_num)
 350{
 351        page->flags &= ~(~0UL << ZONE_SHIFT);
 352        page->flags |= zone_num << ZONE_SHIFT;
 353}
 354
 355/*
 356 * In order to avoid #ifdefs within C code itself, we define
 357 * set_page_address to a noop for non-highmem machines, where
 358 * the field isn't useful.
 359 * The same is true for page_address() in arch-dependent code.
 360 */
 361#if defined(CONFIG_HIGHMEM) || defined(WANT_PAGE_VIRTUAL)
 362
 363#define set_page_address(page, address)                 \
 364        do {                                            \
 365                (page)->virtual = (address);            \
 366        } while(0)
 367
 368#else /* CONFIG_HIGHMEM || WANT_PAGE_VIRTUAL */
 369#define set_page_address(page, address)  do { } while(0)
 370#endif /* CONFIG_HIGHMEM || WANT_PAGE_VIRTUAL */
 371
 372/*
 373 * Permanent address of a page. Obviously must never be
 374 * called on a highmem page.
 375 */
 376#if defined(CONFIG_HIGHMEM) || defined(WANT_PAGE_VIRTUAL)
 377
 378#define page_address(page) ((page)->virtual)
 379
 380#else /* CONFIG_HIGHMEM || WANT_PAGE_VIRTUAL */
 381
 382#define page_address(page)                                              \
 383        __va( (((page) - page_zone(page)->zone_mem_map) << PAGE_SHIFT)  \
 384                        + page_zone(page)->zone_start_paddr)
 385
 386#endif /* CONFIG_HIGHMEM || WANT_PAGE_VIRTUAL */
 387
 388extern void FASTCALL(set_page_dirty(struct page *));
 389
 390/*
 391 * The first mb is necessary to safely close the critical section opened by the
 392 * TryLockPage(), the second mb is necessary to enforce ordering between
 393 * the clear_bit and the read of the waitqueue (to avoid SMP races with a
 394 * parallel wait_on_page).
 395 */
 396#define PageError(page)         test_bit(PG_error, &(page)->flags)
 397#define SetPageError(page)      set_bit(PG_error, &(page)->flags)
 398#define ClearPageError(page)    clear_bit(PG_error, &(page)->flags)
 399#define PageReferenced(page)    test_bit(PG_referenced, &(page)->flags)
 400#define SetPageReferenced(page) set_bit(PG_referenced, &(page)->flags)
 401#define ClearPageReferenced(page)       clear_bit(PG_referenced, &(page)->flags)
 402#define PageTestandClearReferenced(page)        test_and_clear_bit(PG_referenced, &(page)->flags)
 403#define PageSlab(page)          test_bit(PG_slab, &(page)->flags)
 404#define PageSetSlab(page)       set_bit(PG_slab, &(page)->flags)
 405#define PageClearSlab(page)     clear_bit(PG_slab, &(page)->flags)
 406#define PageReserved(page)      test_bit(PG_reserved, &(page)->flags)
 407
 408#define PageActive(page)        test_bit(PG_active, &(page)->flags)
 409#define SetPageActive(page)     set_bit(PG_active, &(page)->flags)
 410#define ClearPageActive(page)   clear_bit(PG_active, &(page)->flags)
 411
 412#define PageLRU(page)           test_bit(PG_lru, &(page)->flags)
 413#define TestSetPageLRU(page)    test_and_set_bit(PG_lru, &(page)->flags)
 414#define TestClearPageLRU(page)  test_and_clear_bit(PG_lru, &(page)->flags)
 415
 416#ifdef CONFIG_HIGHMEM
 417#define PageHighMem(page)               test_bit(PG_highmem, &(page)->flags)
 418#else
 419#define PageHighMem(page)               0 /* needed to optimize away at compile time */
 420#endif
 421
 422#define SetPageReserved(page)           set_bit(PG_reserved, &(page)->flags)
 423#define ClearPageReserved(page)         clear_bit(PG_reserved, &(page)->flags)
 424
 425/*
 426 * Error return values for the *_nopage functions
 427 */
 428#define NOPAGE_SIGBUS   (NULL)
 429#define NOPAGE_OOM      ((struct page *) (-1))
 430
 431/* The array of struct pages */
 432extern mem_map_t * mem_map;
 433
 434/*
 435 * There is only one page-allocator function, and two main namespaces to
 436 * it. The alloc_page*() variants return 'struct page *' and as such
 437 * can allocate highmem pages, the *get*page*() variants return
 438 * virtual kernel addresses to the allocated page(s).
 439 */
 440extern struct page * FASTCALL(_alloc_pages(unsigned int gfp_mask, unsigned int order));
 441extern struct page * FASTCALL(__alloc_pages(unsigned int gfp_mask, unsigned int order, zonelist_t *zonelist));
 442extern struct page * alloc_pages_node(int nid, unsigned int gfp_mask, unsigned int order);
 443
 444static inline struct page * alloc_pages(unsigned int gfp_mask, unsigned int order)
 445{
 446        /*
 447         * Gets optimized away by the compiler.
 448         */
 449        if (order >= MAX_ORDER)
 450                return NULL;
 451        return _alloc_pages(gfp_mask, order);
 452}
 453
 454#define alloc_page(gfp_mask) alloc_pages(gfp_mask, 0)
 455
 456extern unsigned long FASTCALL(__get_free_pages(unsigned int gfp_mask, unsigned int order));
 457extern unsigned long FASTCALL(get_zeroed_page(unsigned int gfp_mask));
 458
 459#define __get_free_page(gfp_mask) \
 460                __get_free_pages((gfp_mask),0)
 461
 462#define __get_dma_pages(gfp_mask, order) \
 463                __get_free_pages((gfp_mask) | GFP_DMA,(order))
 464
 465/*
 466 * The old interface name will be removed in 2.5:
 467 */
 468#define get_free_page get_zeroed_page
 469
 470/*
 471 * There is only one 'core' page-freeing function.
 472 */
 473extern void FASTCALL(__free_pages(struct page *page, unsigned int order));
 474extern void FASTCALL(free_pages(unsigned long addr, unsigned int order));
 475
 476#define __free_page(page) __free_pages((page), 0)
 477#define free_page(addr) free_pages((addr),0)
 478
 479extern void show_free_areas(void);
 480extern void show_free_areas_node(pg_data_t *pgdat);
 481
 482extern void clear_page_tables(struct mm_struct *, unsigned long, int);
 483
 484extern int fail_writepage(struct page *);
 485struct page * shmem_nopage(struct vm_area_struct * vma, unsigned long address, int unused);
 486struct file *shmem_file_setup(char * name, loff_t size);
 487extern void shmem_lock(struct file * file, int lock);
 488extern int shmem_zero_setup(struct vm_area_struct *);
 489
 490extern void zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size);
 491extern int copy_page_range(struct mm_struct *dst, struct mm_struct *src, struct vm_area_struct *vma);
 492extern int remap_page_range(unsigned long from, unsigned long to, unsigned long size, pgprot_t prot);
 493extern int zeromap_page_range(unsigned long from, unsigned long size, pgprot_t prot);
 494
 495extern int vmtruncate(struct inode * inode, loff_t offset);
 496extern pmd_t *FASTCALL(__pmd_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address));
 497extern pte_t *FASTCALL(pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address));
 498extern int handle_mm_fault(struct mm_struct *mm,struct vm_area_struct *vma, unsigned long address, int write_access);
 499extern int make_pages_present(unsigned long addr, unsigned long end);
 500extern int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write);
 501extern int ptrace_readdata(struct task_struct *tsk, unsigned long src, char *dst, int len);
 502extern int ptrace_writedata(struct task_struct *tsk, char * src, unsigned long dst, int len);
 503extern int ptrace_attach(struct task_struct *tsk);
 504extern int ptrace_detach(struct task_struct *, unsigned int);
 505extern void ptrace_disable(struct task_struct *);
 506extern int ptrace_check_attach(struct task_struct *task, int kill);
 507
 508int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, unsigned long start,
 509                int len, int write, int force, struct page **pages, struct vm_area_struct **vmas);
 510
 511/*
 512 * On a two-level page table, this ends up being trivial. Thus the
 513 * inlining and the symmetry break with pte_alloc() that does all
 514 * of this out-of-line.
 515 */
 516static inline pmd_t *pmd_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address)
 517{
 518        if (pgd_none(*pgd))
 519                return __pmd_alloc(mm, pgd, address);
 520        return pmd_offset(pgd, address);
 521}
 522
 523extern int pgt_cache_water[2];
 524extern int check_pgt_cache(void);
 525
 526extern void free_area_init(unsigned long * zones_size);
 527extern void free_area_init_node(int nid, pg_data_t *pgdat, struct page *pmap,
 528        unsigned long * zones_size, unsigned long zone_start_paddr, 
 529        unsigned long *zholes_size);
 530extern void mem_init(void);
 531extern void show_mem(void);
 532extern void si_meminfo(struct sysinfo * val);
 533extern void swapin_readahead(swp_entry_t);
 534
 535extern struct address_space swapper_space;
 536#define PageSwapCache(page) ((page)->mapping == &swapper_space)
 537
 538static inline int is_page_cache_freeable(struct page * page)
 539{
 540        return page_count(page) - !!page->buffers == 1;
 541}
 542
 543extern int FASTCALL(can_share_swap_page(struct page *));
 544extern int FASTCALL(remove_exclusive_swap_page(struct page *));
 545
 546extern void __free_pte(pte_t);
 547
 548/* mmap.c */
 549extern void lock_vma_mappings(struct vm_area_struct *);
 550extern void unlock_vma_mappings(struct vm_area_struct *);
 551extern int insert_vm_struct(struct mm_struct *, struct vm_area_struct *);
 552extern void __insert_vm_struct(struct mm_struct *, struct vm_area_struct *);
 553extern void build_mmap_rb(struct mm_struct *);
 554extern void exit_mmap(struct mm_struct *);
 555
 556extern unsigned long get_unmapped_area(struct file *, unsigned long, unsigned long, unsigned long, unsigned long);
 557
 558extern unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
 559        unsigned long len, unsigned long prot,
 560        unsigned long flag, unsigned long pgoff);
 561
 562static inline unsigned long do_mmap(struct file *file, unsigned long addr,
 563        unsigned long len, unsigned long prot,
 564        unsigned long flag, unsigned long offset)
 565{
 566        unsigned long ret = -EINVAL;
 567        if ((offset + PAGE_ALIGN(len)) < offset)
 568                goto out;
 569        if (!(offset & ~PAGE_MASK))
 570                ret = do_mmap_pgoff(file, addr, len, prot, flag, offset >> PAGE_SHIFT);
 571out:
 572        return ret;
 573}
 574
 575extern int do_munmap(struct mm_struct *, unsigned long, size_t);
 576
 577extern unsigned long do_brk(unsigned long, unsigned long);
 578
 579static inline void __vma_unlink(struct mm_struct * mm, struct vm_area_struct * vma, struct vm_area_struct * prev)
 580{
 581        prev->vm_next = vma->vm_next;
 582        rb_erase(&vma->vm_rb, &mm->mm_rb);
 583        if (mm->mmap_cache == vma)
 584                mm->mmap_cache = prev;
 585}
 586
 587static inline int can_vma_merge(struct vm_area_struct * vma, unsigned long vm_flags)
 588{
 589        if (!vma->vm_file && vma->vm_flags == vm_flags)
 590                return 1;
 591        else
 592                return 0;
 593}
 594
 595struct zone_t;
 596/* filemap.c */
 597extern void remove_inode_page(struct page *);
 598extern unsigned long page_unuse(struct page *);
 599extern void truncate_inode_pages(struct address_space *, loff_t);
 600
 601/* generic vm_area_ops exported for stackable file systems */
 602extern int filemap_sync(struct vm_area_struct *, unsigned long, size_t, unsigned int);
 603extern struct page *filemap_nopage(struct vm_area_struct *, unsigned long, int);
 604
 605/*
 606 * GFP bitmasks..
 607 */
 608/* Zone modifiers in GFP_ZONEMASK (see linux/mmzone.h - low four bits) */
 609#define __GFP_DMA       0x01
 610#define __GFP_HIGHMEM   0x02
 611
 612/* Action modifiers - doesn't change the zoning */
 613#define __GFP_WAIT      0x10    /* Can wait and reschedule? */
 614#define __GFP_HIGH      0x20    /* Should access emergency pools? */
 615#define __GFP_IO        0x40    /* Can start low memory physical IO? */
 616#define __GFP_HIGHIO    0x80    /* Can start high mem physical IO? */
 617#define __GFP_FS        0x100   /* Can call down to low-level FS? */
 618
 619#define GFP_NOHIGHIO    (__GFP_HIGH | __GFP_WAIT | __GFP_IO)
 620#define GFP_NOIO        (__GFP_HIGH | __GFP_WAIT)
 621#define GFP_NOFS        (__GFP_HIGH | __GFP_WAIT | __GFP_IO | __GFP_HIGHIO)
 622#define GFP_ATOMIC      (__GFP_HIGH)
 623#define GFP_USER        (             __GFP_WAIT | __GFP_IO | __GFP_HIGHIO | __GFP_FS)
 624#define GFP_HIGHUSER    (             __GFP_WAIT | __GFP_IO | __GFP_HIGHIO | __GFP_FS | __GFP_HIGHMEM)
 625#define GFP_KERNEL      (__GFP_HIGH | __GFP_WAIT | __GFP_IO | __GFP_HIGHIO | __GFP_FS)
 626#define GFP_NFS         (__GFP_HIGH | __GFP_WAIT | __GFP_IO | __GFP_HIGHIO | __GFP_FS)
 627#define GFP_KSWAPD      (             __GFP_WAIT | __GFP_IO | __GFP_HIGHIO | __GFP_FS)
 628
 629/* Flag - indicates that the buffer will be suitable for DMA.  Ignored on some
 630   platforms, used as appropriate on others */
 631
 632#define GFP_DMA         __GFP_DMA
 633
 634static inline unsigned int pf_gfp_mask(unsigned int gfp_mask)
 635{
 636        /* avoid all memory balancing I/O methods if this task cannot block on I/O */
 637        if (current->flags & PF_NOIO)
 638                gfp_mask &= ~(__GFP_IO | __GFP_HIGHIO | __GFP_FS);
 639
 640        return gfp_mask;
 641}
 642        
 643/* vma is the first one with  address < vma->vm_end,
 644 * and even  address < vma->vm_start. Have to extend vma. */
 645static inline int expand_stack(struct vm_area_struct * vma, unsigned long address)
 646{
 647        unsigned long grow;
 648
 649        /*
 650         * vma->vm_start/vm_end cannot change under us because the caller
 651         * is required to hold the mmap_sem in read mode.  We need the
 652         * page_table_lock lock to serialize against concurrent expand_stacks.
 653         */
 654        address &= PAGE_MASK;
 655        spin_lock(&vma->vm_mm->page_table_lock);
 656
 657        /* already expanded while we were spinning? */
 658        if (vma->vm_start <= address) {
 659                spin_unlock(&vma->vm_mm->page_table_lock);
 660                return 0;
 661        }
 662
 663        grow = (vma->vm_start - address) >> PAGE_SHIFT;
 664        if (vma->vm_end - address > current->rlim[RLIMIT_STACK].rlim_cur ||
 665            ((vma->vm_mm->total_vm + grow) << PAGE_SHIFT) > current->rlim[RLIMIT_AS].rlim_cur) {
 666                spin_unlock(&vma->vm_mm->page_table_lock);
 667                return -ENOMEM;
 668        }
 669
 670        if ((vma->vm_flags & VM_LOCKED) &&
 671            ((vma->vm_mm->locked_vm + grow) << PAGE_SHIFT) > current->rlim[RLIMIT_MEMLOCK].rlim_cur) {
 672                spin_unlock(&vma->vm_mm->page_table_lock);
 673                return -ENOMEM;
 674        }
 675
 676
 677        vma->vm_start = address;
 678        vma->vm_pgoff -= grow;
 679        vma->vm_mm->total_vm += grow;
 680        if (vma->vm_flags & VM_LOCKED)
 681                vma->vm_mm->locked_vm += grow;
 682        spin_unlock(&vma->vm_mm->page_table_lock);
 683        return 0;
 684}
 685
 686/* Look up the first VMA which satisfies  addr < vm_end,  NULL if none. */
 687extern struct vm_area_struct * find_vma(struct mm_struct * mm, unsigned long addr);
 688extern struct vm_area_struct * find_vma_prev(struct mm_struct * mm, unsigned long addr,
 689                                             struct vm_area_struct **pprev);
 690
 691/* Look up the first VMA which intersects the interval start_addr..end_addr-1,
 692   NULL if none.  Assume start_addr < end_addr. */
 693static inline struct vm_area_struct * find_vma_intersection(struct mm_struct * mm, unsigned long start_addr, unsigned long end_addr)
 694{
 695        struct vm_area_struct * vma = find_vma(mm,start_addr);
 696
 697        if (vma && end_addr <= vma->vm_start)
 698                vma = NULL;
 699        return vma;
 700}
 701
 702extern struct vm_area_struct *find_extend_vma(struct mm_struct *mm, unsigned long addr);
 703
 704extern struct page * vmalloc_to_page(void *addr);
 705
 706#endif /* __KERNEL__ */
 707
 708#endif
 709
lxr.linux.no kindly hosted by Redpill Linpro AS, provider of Linux consulting and operations services since 1995.