Import 2.3.13pre3

[davej-history.git] / mm / memory.c

/*
 *  linux/mm/memory.c
 *
 *  Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
 */

/*
 * demand-loading started 01.12.91 - seems it is high on the list of
 * things wanted, and it should be easy to implement. - Linus
 */

/*
 * Ok, demand-loading was easy, shared pages a little bit tricker. Shared
 * pages started 02.12.91, seems to work. - Linus.
 *
 * Tested sharing by executing about 30 /bin/sh: under the old kernel it
 * would have taken more than the 6M I have free, but it worked well as
 * far as I could see.
 *
 * Also corrected some "invalidate()"s - I wasn't doing enough of them.
 */

/*
 * Real VM (paging to/from disk) started 18.12.91. Much more work and
 * thought has to go into this. Oh, well..
 * 19.12.91  -  works, somewhat. Sometimes I get faults, don't know why.
 *              Found it. Everything seems to work now.
 * 20.12.91  -  Ok, making the swap-device changeable like the root.
 */

/*
 * 05.04.94  -  Multi-page memory management added for v1.1.
 *              Idea by Alex Bligh (alex@cconcepts.co.uk)
 */

#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/swap.h>
#include <linux/pagemap.h>
#include <linux/smp_lock.h>
#include <linux/swapctl.h>
#include <linux/iobuf.h>

#include <asm/uaccess.h>
#include <asm/pgtable.h>

unsigned long max_mapnr = 0;
unsigned long num_physpages = 0;
void * high_memory = NULL;

/*
 * We special-case the C-O-W ZERO_PAGE, because it's such
 * a common occurrence (no need to read the page to know
 * that it's zero - better for the cache and memory subsystem).
 */
static inline void copy_cow_page(unsigned long from, unsigned long to)
{
        if (from == ZERO_PAGE(to)) {
                clear_page(to);
                return;
        }
        copy_page(to, from);
}

mem_map_t * mem_map = NULL;

/*
 * oom() prints a message (so that the user knows why the process died),
 * and gives the process an untrappable SIGKILL.
 */
void oom(struct task_struct * task)
{
        printk("\nOut of memory for %s.\n", task->comm);
        force_sig(SIGKILL, task);
}

/*
 * Note: this doesn't free the actual pages themselves. That
 * has been handled earlier when unmapping all the memory regions.
 */
static inline void free_one_pmd(pmd_t * dir)
{
        pte_t * pte;

        if (pmd_none(*dir))
                return;
        if (pmd_bad(*dir)) {
                printk("free_one_pmd: bad directory entry %08lx\n", pmd_val(*dir));
                pmd_clear(dir);
                return;
        }
        pte = pte_offset(dir, 0);
        pmd_clear(dir);
        pte_free(pte);
}

static inline void free_one_pgd(pgd_t * dir)
{
        int j;
        pmd_t * pmd;

        if (pgd_none(*dir))
                return;
        if (pgd_bad(*dir)) {
                printk("free_one_pgd: bad directory entry %08lx\n", pgd_val(*dir));
                pgd_clear(dir);
                return;
        }
        pmd = pmd_offset(dir, 0);
        pgd_clear(dir);
        for (j = 0; j < PTRS_PER_PMD ; j++)
                free_one_pmd(pmd+j);
        pmd_free(pmd);
}

/* Low and high watermarks for page table cache.
   The system should try to have pgt_water[0] <= cache elements <= pgt_water[1]
 */
int pgt_cache_water[2] = { 25, 50 };

/* Returns the number of pages freed */
int check_pgt_cache(void)
{
        return do_check_pgt_cache(pgt_cache_water[0], pgt_cache_water[1]);
}


/*
 * This function clears all user-level page tables of a process - this
 * is needed by execve(), so that old pages aren't in the way.
 */
void clear_page_tables(struct mm_struct *mm, unsigned long first, int nr)
{
        pgd_t * page_dir = mm->pgd;

        page_dir += first;
        do {
                free_one_pgd(page_dir);
                page_dir++;
        } while (--nr);

        /* keep the page table cache within bounds */
        check_pgt_cache();
}

#define PTE_TABLE_MASK  ((PTRS_PER_PTE-1) * sizeof(pte_t))
#define PMD_TABLE_MASK  ((PTRS_PER_PMD-1) * sizeof(pmd_t))

/*
 * copy one vm_area from one task to the other. Assumes the page tables
 * already present in the new task to be cleared in the whole range
 * covered by this vma.
 *
 * 08Jan98 Merged into one routine from several inline routines to reduce
 *         variable count and make things faster. -jj
 */
int copy_page_range(struct mm_struct *dst, struct mm_struct *src,
                        struct vm_area_struct *vma)
{
        pgd_t * src_pgd, * dst_pgd;
        unsigned long address = vma->vm_start;
        unsigned long end = vma->vm_end;
        unsigned long cow = (vma->vm_flags & (VM_SHARED | VM_MAYWRITE)) == VM_MAYWRITE;
        
        src_pgd = pgd_offset(src, address)-1;
        dst_pgd = pgd_offset(dst, address)-1;
        
        for (;;) {
                pmd_t * src_pmd, * dst_pmd;

                src_pgd++; dst_pgd++;
                
                /* copy_pmd_range */
                
                if (pgd_none(*src_pgd))
                        goto skip_copy_pmd_range;
                if (pgd_bad(*src_pgd)) {
                        printk("copy_pmd_range: bad pgd (%08lx)\n", 
                                pgd_val(*src_pgd));
                        pgd_clear(src_pgd);
skip_copy_pmd_range:    address = (address + PGDIR_SIZE) & PGDIR_MASK;
                        if (address >= end)
                                goto out;
                        continue;
                }
                if (pgd_none(*dst_pgd)) {
                        if (!pmd_alloc(dst_pgd, 0))
                                goto nomem;
                }
                
                src_pmd = pmd_offset(src_pgd, address);
                dst_pmd = pmd_offset(dst_pgd, address);

                do {
                        pte_t * src_pte, * dst_pte;
                
                        /* copy_pte_range */
                
                        if (pmd_none(*src_pmd))
                                goto skip_copy_pte_range;
                        if (pmd_bad(*src_pmd)) {
                                printk("copy_pte_range: bad pmd (%08lx)\n", pmd_val(*src_pmd));
                                pmd_clear(src_pmd);
skip_copy_pte_range:            address = (address + PMD_SIZE) & PMD_MASK;
                                if (address >= end)
                                        goto out;
                                goto cont_copy_pmd_range;
                        }
                        if (pmd_none(*dst_pmd)) {
                                if (!pte_alloc(dst_pmd, 0))
                                        goto nomem;
                        }
                        
                        src_pte = pte_offset(src_pmd, address);
                        dst_pte = pte_offset(dst_pmd, address);
                        
                        do {
                                pte_t pte = *src_pte;
                                unsigned long page_nr;
                                
                                /* copy_one_pte */

                                if (pte_none(pte))
                                        goto cont_copy_pte_range;
                                if (!pte_present(pte)) {
                                        swap_duplicate(pte_val(pte));
                                        set_pte(dst_pte, pte);
                                        goto cont_copy_pte_range;
                                }
                                page_nr = MAP_NR(pte_page(pte));
                                if (page_nr >= max_mapnr || 
                                    PageReserved(mem_map+page_nr)) {
                                        set_pte(dst_pte, pte);
                                        goto cont_copy_pte_range;
                                }
                                /* If it's a COW mapping, write protect it both in the parent and the child */
                                if (cow) {
                                        pte = pte_wrprotect(pte);
                                        set_pte(src_pte, pte);
                                }
                                /* If it's a shared mapping, mark it clean in the child */
                                if (vma->vm_flags & VM_SHARED)
                                        pte = pte_mkclean(pte);
                                set_pte(dst_pte, pte_mkold(pte));
                                get_page(mem_map + page_nr);
                        
cont_copy_pte_range:            address += PAGE_SIZE;
                                if (address >= end)
                                        goto out;
                                src_pte++;
                                dst_pte++;
                        } while ((unsigned long)src_pte & PTE_TABLE_MASK);
                
cont_copy_pmd_range:    src_pmd++;
                        dst_pmd++;
                } while ((unsigned long)src_pmd & PMD_TABLE_MASK);
        }
out:
        return 0;

nomem:
        return -ENOMEM;
}

/*
 * Return indicates whether a page was freed so caller can adjust rss
 */
static inline int free_pte(pte_t page)
{
        if (pte_present(page)) {
                unsigned long addr = pte_page(page);
                if (MAP_NR(addr) >= max_mapnr || PageReserved(mem_map+MAP_NR(addr)))
                        return 0;
                /* 
                 * free_page() used to be able to clear swap cache
                 * entries.  We may now have to do it manually.  
                 */
                free_page_and_swap_cache(addr);
                return 1;
        }
        swap_free(pte_val(page));
        return 0;
}

static inline void forget_pte(pte_t page)
{
        if (!pte_none(page)) {
                printk("forget_pte: old mapping existed!\n");
                free_pte(page);
        }
}

static inline int zap_pte_range(struct mm_struct *mm, pmd_t * pmd, unsigned long address, unsigned long size)
{
        pte_t * pte;
        int freed;

        if (pmd_none(*pmd))
                return 0;
        if (pmd_bad(*pmd)) {
                printk("zap_pte_range: bad pmd (%08lx)\n", pmd_val(*pmd));
                pmd_clear(pmd);
                return 0;
        }
        pte = pte_offset(pmd, address);
        address &= ~PMD_MASK;
        if (address + size > PMD_SIZE)
                size = PMD_SIZE - address;
        size >>= PAGE_SHIFT;
        freed = 0;
        for (;;) {
                pte_t page;
                if (!size)
                        break;
                page = *pte;
                pte++;
                size--;
                pte_clear(pte-1);
                if (pte_none(page))
                        continue;
                freed += free_pte(page);
        }
        return freed;
}

static inline int zap_pmd_range(struct mm_struct *mm, pgd_t * dir, unsigned long address, unsigned long size)
{
        pmd_t * pmd;
        unsigned long end;
        int freed;

        if (pgd_none(*dir))
                return 0;
        if (pgd_bad(*dir)) {
                printk("zap_pmd_range: bad pgd (%08lx)\n", pgd_val(*dir));
                pgd_clear(dir);
                return 0;
        }
        pmd = pmd_offset(dir, address);
        address &= ~PGDIR_MASK;
        end = address + size;
        if (end > PGDIR_SIZE)
                end = PGDIR_SIZE;
        freed = 0;
        do {
                freed += zap_pte_range(mm, pmd, address, end - address);
                address = (address + PMD_SIZE) & PMD_MASK; 
                pmd++;
        } while (address < end);
        return freed;
}

/*
 * remove user pages in a given range.
 */
void zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size)
{
        pgd_t * dir;
        unsigned long end = address + size;
        int freed = 0;

        dir = pgd_offset(mm, address);

        /*
         * This is a long-lived spinlock. That's fine.
         * There's no contention, because the page table
         * lock only protects against kswapd anyway, and
         * even if kswapd happened to be looking at this
         * process we _want_ it to get stuck.
         */
        spin_lock(&mm->page_table_lock);
        while (address < end) {
                freed += zap_pmd_range(mm, dir, address, end - address);
                address = (address + PGDIR_SIZE) & PGDIR_MASK;
                dir++;
        }
        spin_unlock(&mm->page_table_lock);
        /*
         * Update rss for the mm_struct (not necessarily current->mm)
         */
        if (mm->rss > 0) {
                mm->rss -= freed;
                if (mm->rss < 0)
                        mm->rss = 0;
        }
}


/*
 * Do a quick page-table lookup for a single page. 
 */
static unsigned long follow_page(unsigned long address) 
{
        pgd_t *pgd;
        pmd_t *pmd;

        pgd = pgd_offset(current->mm, address);
        pmd = pmd_offset(pgd, address);
        if (pmd) {
                pte_t * pte = pte_offset(pmd, address);
                if (pte && pte_present(*pte)) {
                        return pte_page(*pte);
                }
        }
        
        printk(KERN_ERR "Missing page in follow_page\n");
        return 0;
}

/* 
 * Given a physical address, is there a useful struct page pointing to it?
 */

static struct page * get_page_map(unsigned long page)
{
        struct page *map;
        
        if (MAP_NR(page) >= max_mapnr)
                return 0;
        if (page == ZERO_PAGE(page))
                return 0;
        map = mem_map + MAP_NR(page);
        if (PageReserved(map))
                return 0;
        return map;
}

/*
 * Force in an entire range of pages from the current process's user VA,
 * and pin and lock the pages for IO.  
 */

#define dprintk(x...)
int map_user_kiobuf(int rw, struct kiobuf *iobuf, unsigned long va, size_t len)
{
        unsigned long           ptr, end;
        int                     err;
        struct mm_struct *      mm;
        struct vm_area_struct * vma = 0;
        unsigned long           page;
        struct page *           map;
        int                     doublepage = 0;
        int                     repeat = 0;
        int                     i;
        
        /* Make sure the iobuf is not already mapped somewhere. */
        if (iobuf->nr_pages)
                return -EINVAL;

        mm = current->mm;
        dprintk ("map_user_kiobuf: begin\n");
        
        ptr = va & PAGE_MASK;
        end = (va + len + PAGE_SIZE - 1) & PAGE_MASK;
        err = expand_kiobuf(iobuf, (end - ptr) >> PAGE_SHIFT);
        if (err)
                return err;

 repeat:
        down(&mm->mmap_sem);

        err = -EFAULT;
        iobuf->locked = 1;
        iobuf->offset = va & ~PAGE_MASK;
        iobuf->length = len;
        
        i = 0;
        
        /* 
         * First of all, try to fault in all of the necessary pages
         */
        while (ptr < end) {
                if (!vma || ptr >= vma->vm_end) {
                        vma = find_vma(current->mm, ptr);
                        if (!vma) 
                                goto out_unlock;
                }
                if (handle_mm_fault(current, vma, ptr, (rw==READ)) <= 0) 
                        goto out_unlock;
                spin_lock(&mm->page_table_lock);
                page = follow_page(ptr);
                if (!page) {
                        dprintk (KERN_ERR "Missing page in map_user_kiobuf\n");
                        map = NULL;
                        goto retry;
                }
                map = get_page_map(page);
                if (map) {
                        if (TryLockPage(map)) {
                                goto retry;
                        }
                        atomic_inc(&map->count);
                }
                spin_unlock(&mm->page_table_lock);
                dprintk ("Installing page %p %p: %d\n", (void *)page, map, i);
                iobuf->pagelist[i] = page;
                iobuf->maplist[i] = map;
                iobuf->nr_pages = ++i;
                
                ptr += PAGE_SIZE;
        }

        up(&mm->mmap_sem);
        dprintk ("map_user_kiobuf: end OK\n");
        return 0;

 out_unlock:
        up(&mm->mmap_sem);
        unmap_kiobuf(iobuf);
        dprintk ("map_user_kiobuf: end %d\n", err);
        return err;

 retry:

        /* 
         * Undo the locking so far, wait on the page we got to, and try again.
         */
        spin_unlock(&mm->page_table_lock);
        unmap_kiobuf(iobuf);
        up(&mm->mmap_sem);

        /* 
         * Did the release also unlock the page we got stuck on?
         */
        if (map) {
                if (!PageLocked(map)) {
                        /* If so, we may well have the page mapped twice
                         * in the IO address range.  Bad news.  Of
                         * course, it _might_ * just be a coincidence,
                         * but if it happens more than * once, chances
                         * are we have a double-mapped page. */
                        if (++doublepage >= 3) {
                                return -EINVAL;
                        }
                }
        
                /*
                 * Try again...
                 */
                wait_on_page(map);
        }
        
        if (++repeat < 16)
                goto repeat;
        return -EAGAIN;
}


/*
 * Unmap all of the pages referenced by a kiobuf.  We release the pages,
 * and unlock them if they were locked. 
 */

void unmap_kiobuf (struct kiobuf *iobuf) 
{
        int i;
        struct page *map;
        
        for (i = 0; i < iobuf->nr_pages; i++) {
                map = iobuf->maplist[i];
                
                if (map && iobuf->locked) {
                        __free_page(map);
                        UnlockPage(map);
                }
        }
        
        iobuf->nr_pages = 0;
        iobuf->locked = 0;
}

static inline void zeromap_pte_range(pte_t * pte, unsigned long address,
                                     unsigned long size, pgprot_t prot)
{
        unsigned long end;

        address &= ~PMD_MASK;
        end = address + size;
        if (end > PMD_SIZE)
                end = PMD_SIZE;
        do {
                pte_t zero_pte = pte_wrprotect(mk_pte(ZERO_PAGE(address),
                                               prot));
                pte_t oldpage = *pte;
                set_pte(pte, zero_pte);
                forget_pte(oldpage);
                address += PAGE_SIZE;
                pte++;
        } while (address < end);
}

static inline int zeromap_pmd_range(pmd_t * pmd, unsigned long address,
                                    unsigned long size, pgprot_t prot)
{
        unsigned long end;

        address &= ~PGDIR_MASK;
        end = address + size;
        if (end > PGDIR_SIZE)
                end = PGDIR_SIZE;
        do {
                pte_t * pte = pte_alloc(pmd, address);
                if (!pte)
                        return -ENOMEM;
                zeromap_pte_range(pte, address, end - address, prot);
                address = (address + PMD_SIZE) & PMD_MASK;
                pmd++;
        } while (address < end);
        return 0;
}

int zeromap_page_range(unsigned long address, unsigned long size, pgprot_t prot)
{
        int error = 0;
        pgd_t * dir;
        unsigned long beg = address;
        unsigned long end = address + size;

        dir = pgd_offset(current->mm, address);
        flush_cache_range(current->mm, beg, end);
        while (address < end) {
                pmd_t *pmd = pmd_alloc(dir, address);
                error = -ENOMEM;
                if (!pmd)
                        break;
                error = zeromap_pmd_range(pmd, address, end - address, prot);
                if (error)
                        break;
                address = (address + PGDIR_SIZE) & PGDIR_MASK;
                dir++;
        }
        flush_tlb_range(current->mm, beg, end);
        return error;
}

/*
 * maps a range of physical memory into the requested pages. the old
 * mappings are removed. any references to nonexistent pages results
 * in null mappings (currently treated as "copy-on-access")
 */
static inline void remap_pte_range(pte_t * pte, unsigned long address, unsigned long size,
        unsigned long phys_addr, pgprot_t prot)
{
        unsigned long end;

        address &= ~PMD_MASK;
        end = address + size;
        if (end > PMD_SIZE)
                end = PMD_SIZE;
        do {
                unsigned long mapnr;
                pte_t oldpage = *pte;
                pte_clear(pte);

                mapnr = MAP_NR(__va(phys_addr));
                if (mapnr >= max_mapnr || PageReserved(mem_map+mapnr))
                        set_pte(pte, mk_pte_phys(phys_addr, prot));
                forget_pte(oldpage);
                address += PAGE_SIZE;
                phys_addr += PAGE_SIZE;
                pte++;
        } while (address < end);
}

static inline int remap_pmd_range(pmd_t * pmd, unsigned long address, unsigned long size,
        unsigned long phys_addr, pgprot_t prot)
{
        unsigned long end;

        address &= ~PGDIR_MASK;
        end = address + size;
        if (end > PGDIR_SIZE)
                end = PGDIR_SIZE;
        phys_addr -= address;
        do {
                pte_t * pte = pte_alloc(pmd, address);
                if (!pte)
                        return -ENOMEM;
                remap_pte_range(pte, address, end - address, address + phys_addr, prot);
                address = (address + PMD_SIZE) & PMD_MASK;
                pmd++;
        } while (address < end);
        return 0;
}

int remap_page_range(unsigned long from, unsigned long phys_addr, unsigned long size, pgprot_t prot)
{
        int error = 0;
        pgd_t * dir;
        unsigned long beg = from;
        unsigned long end = from + size;

        phys_addr -= from;
        dir = pgd_offset(current->mm, from);
        flush_cache_range(current->mm, beg, end);
        while (from < end) {
                pmd_t *pmd = pmd_alloc(dir, from);
                error = -ENOMEM;
                if (!pmd)
                        break;
                error = remap_pmd_range(pmd, from, end - from, phys_addr + from, prot);
                if (error)
                        break;
                from = (from + PGDIR_SIZE) & PGDIR_MASK;
                dir++;
        }
        flush_tlb_range(current->mm, beg, end);
        return error;
}

/*
 * This routine is used to map in a page into an address space: needed by
 * execve() for the initial stack and environment pages.
 */
unsigned long put_dirty_page(struct task_struct * tsk, unsigned long page, unsigned long address)
{
        pgd_t * pgd;
        pmd_t * pmd;
        pte_t * pte;

        if (MAP_NR(page) >= max_mapnr)
                printk("put_dirty_page: trying to put page %08lx at %08lx\n",page,address);
        if (page_count(mem_map + MAP_NR(page)) != 1)
                printk("mem_map disagrees with %08lx at %08lx\n",page,address);
        pgd = pgd_offset(tsk->mm,address);
        pmd = pmd_alloc(pgd, address);
        if (!pmd) {
                free_page(page);
                oom(tsk);
                return 0;
        }
        pte = pte_alloc(pmd, address);
        if (!pte) {
                free_page(page);
                oom(tsk);
                return 0;
        }
        if (!pte_none(*pte)) {
                printk("put_dirty_page: pte %08lx already exists\n",
                       pte_val(*pte));
                free_page(page);
                return 0;
        }
        flush_page_to_ram(page);
        set_pte(pte, pte_mkwrite(pte_mkdirty(mk_pte(page, PAGE_COPY))));
/* no need for flush_tlb */
        return page;
}

/*
 * This routine handles present pages, when users try to write
 * to a shared page. It is done by copying the page to a new address
 * and decrementing the shared-page counter for the old page.
 *
 * Goto-purists beware: the only reason for goto's here is that it results
 * in better assembly code.. The "default" path will see no jumps at all.
 *
 * Note that this routine assumes that the protection checks have been
 * done by the caller (the low-level page fault routine in most cases).
 * Thus we can safely just mark it writable once we've done any necessary
 * COW.
 *
 * We also mark the page dirty at this point even though the page will
 * change only once the write actually happens. This avoids a few races,
 * and potentially makes it more efficient.
 *
 * We enter with the page table read-lock held, and need to exit without
 * it.
 */
static int do_wp_page(struct task_struct * tsk, struct vm_area_struct * vma,
        unsigned long address, pte_t *page_table, pte_t pte)
{
        unsigned long old_page, new_page;
        struct page * page;

        old_page = pte_page(pte);
        if (MAP_NR(old_page) >= max_mapnr)
                goto bad_wp_page;
        tsk->min_flt++;
        page = mem_map + MAP_NR(old_page);
        
        /*
         * We can avoid the copy if:
         * - we're the only user (count == 1)
         * - the only other user is the swap cache,
         *   and the only swap cache user is itself,
         *   in which case we can remove the page
         *   from the swap cache.
         */
        switch (page_count(page)) {
        case 2:
                if (!PageSwapCache(page))
                        break;
                if (swap_count(page->offset) != 1)
                        break;
                delete_from_swap_cache(page);
                /* FallThrough */
        case 1:
                flush_cache_page(vma, address);
                set_pte(page_table, pte_mkyoung(pte_mkdirty(pte_mkwrite(pte))));
                flush_tlb_page(vma, address);
                spin_unlock(&tsk->mm->page_table_lock);
                return 1;
        }

        /*
         * Ok, we need to copy. Oh, well..
         */
        spin_unlock(&tsk->mm->page_table_lock);
        new_page = __get_free_page(GFP_USER);
        if (!new_page)
                return -1;
        spin_lock(&tsk->mm->page_table_lock);

        /*
         * Re-check the pte - we dropped the lock
         */
        if (pte_val(*page_table) == pte_val(pte)) {
                if (PageReserved(page))
                        ++vma->vm_mm->rss;
                copy_cow_page(old_page,new_page);
                flush_page_to_ram(old_page);
                flush_page_to_ram(new_page);
                flush_cache_page(vma, address);
                set_pte(page_table, pte_mkwrite(pte_mkdirty(mk_pte(new_page, vma->vm_page_prot))));
                flush_tlb_page(vma, address);

                /* Free the old page.. */
                new_page = old_page;
        }
        spin_unlock(&tsk->mm->page_table_lock);
        free_page(new_page);
        return 1;

bad_wp_page:
        spin_unlock(&tsk->mm->page_table_lock);
        printk("do_wp_page: bogus page at address %08lx (%08lx)\n",address,old_page);
        return -1;
}

/*
 * This function zeroes out partial mmap'ed pages at truncation time..
 */
static void partial_clear(struct vm_area_struct *vma, unsigned long address)
{
        pgd_t *page_dir;
        pmd_t *page_middle;
        pte_t *page_table, pte;

        page_dir = pgd_offset(vma->vm_mm, address);
        if (pgd_none(*page_dir))
                return;
        if (pgd_bad(*page_dir)) {
                printk("bad page table directory entry %p:[%lx]\n", page_dir, pgd_val(*page_dir));
                pgd_clear(page_dir);
                return;
        }
        page_middle = pmd_offset(page_dir, address);
        if (pmd_none(*page_middle))
                return;
        if (pmd_bad(*page_middle)) {
                printk("bad page table directory entry %p:[%lx]\n", page_dir, pgd_val(*page_dir));
                pmd_clear(page_middle);
                return;
        }
        page_table = pte_offset(page_middle, address);
        pte = *page_table;
        if (!pte_present(pte))
                return;
        flush_cache_page(vma, address);
        address &= ~PAGE_MASK;
        address += pte_page(pte);
        if (MAP_NR(address) >= max_mapnr)
                return;
        memset((void *) address, 0, PAGE_SIZE - (address & ~PAGE_MASK));
        flush_page_to_ram(pte_page(pte));
}

/*
 * Handle all mappings that got truncated by a "truncate()"
 * system call.
 *
 * NOTE! We have to be ready to update the memory sharing
 * between the file and the memory map for a potential last
 * incomplete page.  Ugly, but necessary.
 */
void vmtruncate(struct inode * inode, unsigned long offset)
{
        struct vm_area_struct * mpnt;

        truncate_inode_pages(inode, offset);
        spin_lock(&inode->i_shared_lock);
        if (!inode->i_mmap)
                goto out_unlock;
        mpnt = inode->i_mmap;
        do {
                struct mm_struct *mm = mpnt->vm_mm;
                unsigned long start = mpnt->vm_start;
                unsigned long end = mpnt->vm_end;
                unsigned long len = end - start;
                unsigned long diff;

                /* mapping wholly truncated? */
                if (mpnt->vm_offset >= offset) {
                        flush_cache_range(mm, start, end);
                        zap_page_range(mm, start, len);
                        flush_tlb_range(mm, start, end);
                        continue;
                }
                /* mapping wholly unaffected? */
                diff = offset - mpnt->vm_offset;
                if (diff >= len)
                        continue;
                /* Ok, partially affected.. */
                start += diff;
                len = (len - diff) & PAGE_MASK;
                if (start & ~PAGE_MASK) {
                        partial_clear(mpnt, start);
                        start = (start + ~PAGE_MASK) & PAGE_MASK;
                }
                flush_cache_range(mm, start, end);
                zap_page_range(mm, start, len);
                flush_tlb_range(mm, start, end);
        } while ((mpnt = mpnt->vm_next_share) != NULL);
out_unlock:
        spin_unlock(&inode->i_shared_lock);
}



/* 
 * Primitive swap readahead code. We simply read an aligned block of
 * (1 << page_cluster) entries in the swap area. This method is chosen
 * because it doesn't cost us any seek time.  We also make sure to queue
 * the 'original' request together with the readahead ones...  
 */
void swapin_readahead(unsigned long entry)
{
        int i;
        struct page *new_page;
        unsigned long offset = SWP_OFFSET(entry);
        struct swap_info_struct *swapdev = SWP_TYPE(entry) + swap_info;
        
        offset = (offset >> page_cluster) << page_cluster;

        i = 1 << page_cluster;
        do {
                /* Don't read-ahead past the end of the swap area */
                if (offset >= swapdev->max)
                        break;
                /* Don't block on I/O for read-ahead */
                if (atomic_read(&nr_async_pages) >= pager_daemon.swap_cluster)
                        break;
                /* Don't read in bad or busy pages */
                if (!swapdev->swap_map[offset])
                        break;
                if (swapdev->swap_map[offset] == SWAP_MAP_BAD)
                        break;

                /* Ok, do the async read-ahead now */
                new_page = read_swap_cache_async(SWP_ENTRY(SWP_TYPE(entry), offset), 0);
                if (new_page != NULL)
                        __free_page(new_page);
                offset++;
        } while (--i);
        return;
}

static int do_swap_page(struct task_struct * tsk,
        struct vm_area_struct * vma, unsigned long address,
        pte_t * page_table, unsigned long entry, int write_access)
{
        struct page *page = lookup_swap_cache(entry);
        pte_t pte;

        if (!page) {
                lock_kernel();
                swapin_readahead(entry);
                page = read_swap_cache(entry);
                unlock_kernel();
                if (!page)
                        return -1;

                flush_page_to_ram(page_address(page));
        }

        vma->vm_mm->rss++;
        tsk->min_flt++;
        lock_kernel();
        swap_free(entry);
        unlock_kernel();

        pte = mk_pte(page_address(page), vma->vm_page_prot);

        if (write_access && !is_page_shared(page)) {
                delete_from_swap_cache(page);
                pte = pte_mkwrite(pte_mkdirty(pte));
        }
        set_pte(page_table, pte);
        /* No need to invalidate - it was non-present before */
        update_mmu_cache(vma, address, pte);
        return 1;
}

/*
 * This only needs the MM semaphore
 */
static int do_anonymous_page(struct task_struct * tsk, struct vm_area_struct * vma, pte_t *page_table, int write_access, unsigned long addr)
{
        pte_t entry = pte_wrprotect(mk_pte(ZERO_PAGE(addr), vma->vm_page_prot));
        if (write_access) {
                unsigned long page = __get_free_page(GFP_USER);
                if (!page)
                        return -1;
                clear_page(page);
                entry = pte_mkwrite(pte_mkdirty(mk_pte(page, vma->vm_page_prot)));
                vma->vm_mm->rss++;
                tsk->min_flt++;
                flush_page_to_ram(page);
        }
        set_pte(page_table, entry);
        /* No need to invalidate - it was non-present before */
        update_mmu_cache(vma, addr, entry);
        return 1;
}

/*
 * do_no_page() tries to create a new page mapping. It aggressively
 * tries to share with existing pages, but makes a separate copy if
 * the "write_access" parameter is true in order to avoid the next
 * page fault.
 *
 * As this is called only for pages that do not currently exist, we
 * do not need to flush old virtual caches or the TLB.
 *
 * This is called with the MM semaphore and the kernel lock held.
 * We need to release the kernel lock as soon as possible..
 */
static int do_no_page(struct task_struct * tsk, struct vm_area_struct * vma,
        unsigned long address, int write_access, pte_t *page_table)
{
        unsigned long page;
        pte_t entry;

        if (!vma->vm_ops || !vma->vm_ops->nopage)
                return do_anonymous_page(tsk, vma, page_table, write_access, address);

        /*
         * The third argument is "no_share", which tells the low-level code
         * to copy, not share the page even if sharing is possible.  It's
         * essentially an early COW detection.
         */
        page = vma->vm_ops->nopage(vma, address & PAGE_MASK, (vma->vm_flags & VM_SHARED)?0:write_access);
        if (!page)
                return 0;       /* SIGBUS - but we _really_ should know whether it is OOM or SIGBUS */
        if (page == -1)
                return -1;      /* OOM */

        ++tsk->maj_flt;
        ++vma->vm_mm->rss;
        /*
         * This silly early PAGE_DIRTY setting removes a race
         * due to the bad i386 page protection. But it's valid
         * for other architectures too.
         *
         * Note that if write_access is true, we either now have
         * an exclusive copy of the page, or this is a shared mapping,
         * so we can make it writable and dirty to avoid having to
         * handle that later.
         */
        flush_page_to_ram(page);
        entry = mk_pte(page, vma->vm_page_prot);
        if (write_access) {
                entry = pte_mkwrite(pte_mkdirty(entry));
        } else if (page_count(mem_map+MAP_NR(page)) > 1 &&
                   !(vma->vm_flags & VM_SHARED))
                entry = pte_wrprotect(entry);
        set_pte(page_table, entry);
        /* no need to invalidate: a not-present page shouldn't be cached */
        update_mmu_cache(vma, address, entry);
        return 1;
}

/*
 * These routines also need to handle stuff like marking pages dirty
 * and/or accessed for architectures that don't do it in hardware (most
 * RISC architectures).  The early dirtying is also good on the i386.
 *
 * There is also a hook called "update_mmu_cache()" that architectures
 * with external mmu caches can use to update those (ie the Sparc or
 * PowerPC hashed page tables that act as extended TLBs).
 *
 * Note the "page_table_lock". It is to protect against kswapd removing
 * pages from under us. Note that kswapd only ever _removes_ pages, never
 * adds them. As such, once we have noticed that the page is not present,
 * we can drop the lock early.
 *
 * The adding of pages is protected by the MM semaphore (which we hold),
 * so we don't need to worry about a page being suddenly been added into
 * our VM.
 */
static inline int handle_pte_fault(struct task_struct *tsk,
        struct vm_area_struct * vma, unsigned long address,
        int write_access, pte_t * pte)
{
        pte_t entry;

        entry = *pte;
        if (!pte_present(entry)) {
                if (pte_none(entry))
                        return do_no_page(tsk, vma, address, write_access, pte);
                return do_swap_page(tsk, vma, address, pte, pte_val(entry), write_access);
        }

        /*
         * Ok, the entry was present, we need to get the page table
         * lock to synchronize with kswapd, and verify that the entry
         * didn't change from under us..
         */
        spin_lock(&tsk->mm->page_table_lock);
        if (pte_val(entry) == pte_val(*pte)) {
                if (write_access) {
                        if (!pte_write(entry))
                                return do_wp_page(tsk, vma, address, pte, entry);

                        entry = pte_mkdirty(entry);
                }
                entry = pte_mkyoung(entry);
                set_pte(pte, entry);
                flush_tlb_page(vma, address);
                update_mmu_cache(vma, address, entry);
        }
        spin_unlock(&tsk->mm->page_table_lock);
        return 1;
}

/*
 * By the time we get here, we already hold the mm semaphore
 */
int handle_mm_fault(struct task_struct *tsk, struct vm_area_struct * vma,
        unsigned long address, int write_access)
{
        pgd_t *pgd;
        pmd_t *pmd;

        pgd = pgd_offset(vma->vm_mm, address);
        pmd = pmd_alloc(pgd, address);
        if (pmd) {
                pte_t * pte = pte_alloc(pmd, address);
                if (pte)
                        return handle_pte_fault(tsk, vma, address, write_access, pte);
        }
        return -1;
}

/*
 * Simplistic page force-in..
 */
int make_pages_present(unsigned long addr, unsigned long end)
{
        int write;
        struct task_struct *tsk = current;
        struct vm_area_struct * vma;

        vma = find_vma(tsk->mm, addr);
        write = (vma->vm_flags & VM_WRITE) != 0;
        while (addr < end) {
                if (handle_mm_fault(tsk, vma, addr, write) < 0)
                        return -1;
                addr += PAGE_SIZE;
        }
        return 0;
}