#include <linux/rmap.h>
#include <linux/security.h>
#include <linux/backing-dev.h>
+#include <linux/syscalls.h>
#include <asm/pgtable.h>
#include <asm/tlbflush.h>
#include <linux/swapops.h>
-#include <linux/vs_base.h>
+#include <linux/vs_memory.h>
-spinlock_t swaplock = SPIN_LOCK_UNLOCKED;
+DEFINE_SPINLOCK(swaplock);
unsigned int nr_swapfiles;
-int total_swap_pages;
+long total_swap_pages;
static int swap_overflow;
EXPORT_SYMBOL(total_swap_pages);
WARN_ON(page_count(page) <= 1);
bdi = bdev->bd_inode->i_mapping->backing_dev_info;
- bdi->unplug_io_fn(bdi, page);
+ blk_run_backing_dev(bdi, page);
}
up_read(&swap_unplug_sem);
}
check_next_cluster:
if (offset+SWAPFILE_CLUSTER-1 <= si->highest_bit)
{
- int nr;
+ unsigned long nr;
for (nr = offset; nr < offset+SWAPFILE_CLUSTER; nr++)
if (si->swap_map[nr])
{
/* Is the only swap cache user the cache itself? */
if (p->swap_map[swp_offset(entry)] == 1) {
/* Recheck the page count with the swapcache lock held.. */
- spin_lock_irq(&swapper_space.tree_lock);
+ write_lock_irq(&swapper_space.tree_lock);
if (page_count(page) == 2)
retval = 1;
- spin_unlock_irq(&swapper_space.tree_lock);
+ write_unlock_irq(&swapper_space.tree_lock);
}
swap_info_put(p);
}
retval = 0;
if (p->swap_map[swp_offset(entry)] == 1) {
/* Recheck the page count with the swapcache lock held.. */
- spin_lock_irq(&swapper_space.tree_lock);
+ write_lock_irq(&swapper_space.tree_lock);
if ((page_count(page) == 2) && !PageWriteback(page)) {
__delete_from_swap_cache(page);
SetPageDirty(page);
retval = 1;
}
- spin_unlock_irq(&swapper_space.tree_lock);
+ write_unlock_irq(&swapper_space.tree_lock);
}
swap_info_put(p);
p = swap_info_get(entry);
if (p) {
- if (swap_entry_free(p, swp_offset(entry)) == 1) {
- spin_lock_irq(&swapper_space.tree_lock);
- page = radix_tree_lookup(&swapper_space.page_tree,
- entry.val);
- if (page && TestSetPageLocked(page))
- page = NULL;
- spin_unlock_irq(&swapper_space.tree_lock);
- }
+ if (swap_entry_free(p, swp_offset(entry)) == 1)
+ page = find_trylock_page(&swapper_space, entry.val);
swap_info_put(p);
}
if (page) {
}
/*
- * The swap entry has been read in advance, and we return 1 to indicate
- * that the page has been used or is no longer needed.
- *
* Always set the resulting pte to be nowrite (the same as COW pages
* after one process has exited). We don't know just how many PTEs will
* share this swap entry, so be cautious and let do_wp_page work out
* what to do if a write is requested later.
+ *
+ * vma->vm_mm->page_table_lock is held.
*/
-/* vma->vm_mm->page_table_lock is held */
-static void
-unuse_pte(struct vm_area_struct *vma, unsigned long address, pte_t *dir,
- swp_entry_t entry, struct page *page)
+static void unuse_pte(struct vm_area_struct *vma, pte_t *pte,
+ unsigned long addr, swp_entry_t entry, struct page *page)
{
- // vma->vm_mm->rss++;
- vx_rsspages_inc(vma->vm_mm);
+ inc_mm_counter(vma->vm_mm, rss);
get_page(page);
- set_pte(dir, pte_mkold(mk_pte(page, vma->vm_page_prot)));
- page_add_anon_rmap(page, vma, address);
+ set_pte_at(vma->vm_mm, addr, pte,
+ pte_mkold(mk_pte(page, vma->vm_page_prot)));
+ page_add_anon_rmap(page, vma, addr);
swap_free(entry);
+ /*
+ * Move the page to the active list so it is not
+ * immediately swapped out again after swapon.
+ */
+ activate_page(page);
}
-/* vma->vm_mm->page_table_lock is held */
-static unsigned long unuse_pmd(struct vm_area_struct * vma, pmd_t *dir,
- unsigned long address, unsigned long size, unsigned long offset,
- swp_entry_t entry, struct page *page)
+static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
+ unsigned long addr, unsigned long end,
+ swp_entry_t entry, struct page *page)
{
- pte_t * pte;
- unsigned long end;
+ pte_t *pte;
pte_t swp_pte = swp_entry_to_pte(entry);
- if (pmd_none(*dir))
- return 0;
- if (pmd_bad(*dir)) {
- pmd_ERROR(*dir);
- pmd_clear(dir);
- return 0;
- }
- pte = pte_offset_map(dir, address);
- offset += address & PMD_MASK;
- address &= ~PMD_MASK;
- end = address + size;
- if (end > PMD_SIZE)
- end = PMD_SIZE;
+ pte = pte_offset_map(pmd, addr);
do {
/*
* swapoff spends a _lot_ of time in this loop!
* Test inline before going to call unuse_pte.
*/
if (unlikely(pte_same(*pte, swp_pte))) {
- unuse_pte(vma, offset + address, pte, entry, page);
+ unuse_pte(vma, pte, addr, entry, page);
pte_unmap(pte);
- /* add 1 since address may be 0 */
- return 1 + offset + address;
+ return 1;
}
- address += PAGE_SIZE;
- pte++;
- } while (address && (address < end));
+ } while (pte++, addr += PAGE_SIZE, addr != end);
pte_unmap(pte - 1);
return 0;
}
-/* vma->vm_mm->page_table_lock is held */
-static unsigned long unuse_pgd(struct vm_area_struct * vma, pgd_t *dir,
- unsigned long address, unsigned long size,
- swp_entry_t entry, struct page *page)
+static inline int unuse_pmd_range(struct vm_area_struct *vma, pud_t *pud,
+ unsigned long addr, unsigned long end,
+ swp_entry_t entry, struct page *page)
{
- pmd_t * pmd;
- unsigned long offset, end;
- unsigned long foundaddr;
+ pmd_t *pmd;
+ unsigned long next;
- if (pgd_none(*dir))
- return 0;
- if (pgd_bad(*dir)) {
- pgd_ERROR(*dir);
- pgd_clear(dir);
- return 0;
- }
- pmd = pmd_offset(dir, address);
- offset = address & PGDIR_MASK;
- address &= ~PGDIR_MASK;
- end = address + size;
- if (end > PGDIR_SIZE)
- end = PGDIR_SIZE;
- if (address >= end)
- BUG();
+ pmd = pmd_offset(pud, addr);
do {
- foundaddr = unuse_pmd(vma, pmd, address, end - address,
- offset, entry, page);
- if (foundaddr)
- return foundaddr;
- address = (address + PMD_SIZE) & PMD_MASK;
- pmd++;
- } while (address && (address < end));
+ next = pmd_addr_end(addr, end);
+ if (pmd_none_or_clear_bad(pmd))
+ continue;
+ if (unuse_pte_range(vma, pmd, addr, next, entry, page))
+ return 1;
+ } while (pmd++, addr = next, addr != end);
return 0;
}
-/* vma->vm_mm->page_table_lock is held */
-static unsigned long unuse_vma(struct vm_area_struct * vma, pgd_t *pgdir,
- swp_entry_t entry, struct page *page)
+static inline int unuse_pud_range(struct vm_area_struct *vma, pgd_t *pgd,
+ unsigned long addr, unsigned long end,
+ swp_entry_t entry, struct page *page)
{
- unsigned long start = vma->vm_start, end = vma->vm_end;
- unsigned long foundaddr;
+ pud_t *pud;
+ unsigned long next;
- if (start >= end)
- BUG();
+ pud = pud_offset(pgd, addr);
do {
- foundaddr = unuse_pgd(vma, pgdir, start, end - start,
- entry, page);
- if (foundaddr)
- return foundaddr;
- start = (start + PGDIR_SIZE) & PGDIR_MASK;
- pgdir++;
- } while (start && (start < end));
+ next = pud_addr_end(addr, end);
+ if (pud_none_or_clear_bad(pud))
+ continue;
+ if (unuse_pmd_range(vma, pud, addr, next, entry, page))
+ return 1;
+ } while (pud++, addr = next, addr != end);
return 0;
}
-static int unuse_process(struct mm_struct * mm,
- swp_entry_t entry, struct page* page)
+static int unuse_vma(struct vm_area_struct *vma,
+ swp_entry_t entry, struct page *page)
{
- struct vm_area_struct* vma;
- unsigned long foundaddr = 0;
+ pgd_t *pgd;
+ unsigned long addr, end, next;
- /*
- * Go through process' page directory.
- */
- down_read(&mm->mmap_sem);
+ if (page->mapping) {
+ addr = page_address_in_vma(page, vma);
+ if (addr == -EFAULT)
+ return 0;
+ else
+ end = addr + PAGE_SIZE;
+ } else {
+ addr = vma->vm_start;
+ end = vma->vm_end;
+ }
+
+ pgd = pgd_offset(vma->vm_mm, addr);
+ do {
+ next = pgd_addr_end(addr, end);
+ if (pgd_none_or_clear_bad(pgd))
+ continue;
+ if (unuse_pud_range(vma, pgd, addr, next, entry, page))
+ return 1;
+ } while (pgd++, addr = next, addr != end);
+ return 0;
+}
+
+static int unuse_mm(struct mm_struct *mm,
+ swp_entry_t entry, struct page *page)
+{
+ struct vm_area_struct *vma;
+
+ if (!down_read_trylock(&mm->mmap_sem)) {
+ /*
+ * Our reference to the page stops try_to_unmap_one from
+ * unmapping its ptes, so swapoff can make progress.
+ */
+ unlock_page(page);
+ down_read(&mm->mmap_sem);
+ lock_page(page);
+ }
spin_lock(&mm->page_table_lock);
for (vma = mm->mmap; vma; vma = vma->vm_next) {
- if (!is_vm_hugetlb_page(vma)) {
- pgd_t * pgd = pgd_offset(mm, vma->vm_start);
- foundaddr = unuse_vma(vma, pgd, entry, page);
- if (foundaddr)
- break;
- }
+ if (vma->anon_vma && unuse_vma(vma, entry, page))
+ break;
}
spin_unlock(&mm->page_table_lock);
up_read(&mm->mmap_sem);
/*
- * Currently unuse_process cannot fail, but leave error handling
+ * Currently unuse_mm cannot fail, but leave error handling
* at call sites for now, since we change it from time to time.
*/
return 0;
*
* A simpler strategy would be to start at the last mm we
* freed the previous entry from; but that would take less
- * advantage of mmlist ordering (now preserved by swap_out()),
- * which clusters forked address spaces together, most recent
- * child immediately after parent. If we race with dup_mmap(),
- * we very much want to resolve parent before child, otherwise
- * we may miss some entries: using last mm would invert that.
+ * advantage of mmlist ordering, which clusters forked mms
+ * together, child after parent. If we race with dup_mmap(), we
+ * prefer to resolve parent before child, lest we miss entries
+ * duplicated after we scanned child: using last mm would invert
+ * that. Though it's only a serious concern when an overflowed
+ * swap count is reset from SWAP_MAP_MAX, preventing a rescan.
*/
start_mm = &init_mm;
atomic_inc(&init_mm.mm_users);
/*
* Keep on scanning until all entries have gone. Usually,
* one pass through swap_map is enough, but not necessarily:
- * mmput() removes mm from mmlist before exit_mmap() and its
- * zap_page_range(). That's not too bad, those entries are
- * on their way out, and handled faster there than here.
- * do_munmap() behaves similarly, taking the range out of mm's
- * vma list before zap_page_range(). But unfortunately, when
- * unmapping a part of a vma, it takes the whole out first,
- * then reinserts what's left after (might even reschedule if
- * open() method called) - so swap entries may be invisible
- * to swapoff for a while, then reappear - but that is rare.
+ * there are races when an instance of an entry might be missed.
*/
- while ((i = find_next_to_unuse(si, i))) {
+ while ((i = find_next_to_unuse(si, i)) != 0) {
if (signal_pending(current)) {
retval = -EINTR;
break;
wait_on_page_writeback(page);
/*
- * Remove all references to entry, without blocking.
+ * Remove all references to entry.
* Whenever we reach init_mm, there's no address space
* to search, but use it as a reminder to search shmem.
*/
if (start_mm == &init_mm)
shmem = shmem_unuse(entry, page);
else
- retval = unuse_process(start_mm, entry, page);
+ retval = unuse_mm(start_mm, entry, page);
}
if (*swap_map > 1) {
int set_start_mm = (*swap_map >= swcount);
while (*swap_map > 1 && !retval &&
(p = p->next) != &start_mm->mmlist) {
mm = list_entry(p, struct mm_struct, mmlist);
- atomic_inc(&mm->mm_users);
+ if (atomic_inc_return(&mm->mm_users) == 1) {
+ atomic_dec(&mm->mm_users);
+ continue;
+ }
spin_unlock(&mmlist_lock);
mmput(prev_mm);
prev_mm = mm;
set_start_mm = 1;
shmem = shmem_unuse(entry, page);
} else
- retval = unuse_process(mm, entry, page);
+ retval = unuse_mm(mm, entry, page);
if (set_start_mm && *swap_map < swcount) {
mmput(new_start_mm);
atomic_inc(&mm->mm_users);
return retval;
}
+/*
+ * After a successful try_to_unuse, if no swap is now in use, we know we
+ * can empty the mmlist. swap_list_lock must be held on entry and exit.
+ * Note that mmlist_lock nests inside swap_list_lock, and an mm must be
+ * added to the mmlist just after page_duplicate - before would be racy.
+ */
+static void drain_mmlist(void)
+{
+ struct list_head *p, *next;
+ unsigned int i;
+
+ for (i = 0; i < nr_swapfiles; i++)
+ if (swap_info[i].inuse_pages)
+ return;
+ spin_lock(&mmlist_lock);
+ list_for_each_safe(p, next, &init_mm.mmlist)
+ list_del_init(p);
+ spin_unlock(&mmlist_lock);
+}
+
/*
* Use this swapdev's extent info to locate the (PAGE_SIZE) block which
* corresponds to page offset `offset'.
unsigned short *swap_map;
struct file *swap_file, *victim;
struct address_space *mapping;
+ struct inode *inode;
char * pathname;
int i, type, prev;
int err;
if (IS_ERR(pathname))
goto out;
- victim = filp_open(pathname, O_RDWR, 0);
+ victim = filp_open(pathname, O_RDWR|O_LARGEFILE, 0);
putname(pathname);
err = PTR_ERR(victim);
if (IS_ERR(victim))
}
down(&swapon_sem);
swap_list_lock();
+ drain_mmlist();
swap_device_lock(p);
swap_file = p->swap_file;
p->swap_file = NULL;
swap_list_unlock();
up(&swapon_sem);
vfree(swap_map);
- if (S_ISBLK(mapping->host->i_mode)) {
- struct block_device *bdev = I_BDEV(mapping->host);
+ inode = mapping->host;
+ if (S_ISBLK(inode->i_mode)) {
+ struct block_device *bdev = I_BDEV(inode);
set_blocksize(bdev, p->old_block_size);
bd_release(bdev);
} else {
- up(&mapping->host->i_sem);
+ down(&inode->i_sem);
+ inode->i_flags &= ~S_SWAPFILE;
+ up(&inode->i_sem);
}
filp_close(swap_file, NULL);
err = 0;
int i, prev;
int error;
static int least_priority;
- union swap_header *swap_header = 0;
+ union swap_header *swap_header = NULL;
int swap_header_version;
int nr_good_pages = 0;
unsigned long maxpages = 1;
p->highest_bit = 0;
p->cluster_nr = 0;
p->inuse_pages = 0;
- p->sdev_lock = SPIN_LOCK_UNLOCKED;
+ spin_lock_init(&p->sdev_lock);
p->next = -1;
if (swap_flags & SWAP_FLAG_PREFER) {
p->prio =
name = NULL;
goto bad_swap_2;
}
- swap_file = filp_open(name, O_RDWR, 0);
+ swap_file = filp_open(name, O_RDWR|O_LARGEFILE, 0);
error = PTR_ERR(swap_file);
if (IS_ERR(swap_file)) {
swap_file = NULL;
p->bdev = inode->i_sb->s_bdev;
down(&inode->i_sem);
did_down = 1;
+ if (IS_SWAPFILE(inode)) {
+ error = -EBUSY;
+ goto bad_swap;
+ }
} else {
goto bad_swap;
}
++least_priority;
swap_list_unlock();
destroy_swap_extents(p);
- if (swap_map)
- vfree(swap_map);
+ vfree(swap_map);
if (swap_file)
filp_close(swap_file, NULL);
out:
}
if (name)
putname(name);
- if (error && did_down)
+ if (did_down) {
+ if (!error)
+ inode->i_flags |= S_SWAPFILE;
up(&inode->i_sem);
+ }
return error;
}
val->freeswap = nr_swap_pages + nr_to_be_unused;
val->totalswap = total_swap_pages + nr_to_be_unused;
swap_list_unlock();
- if (vx_flags(VXF_VIRT_MEM, 0))
- vx_vsi_swapinfo(val);
+ if (vx_flags(VXF_VIRT_MEM, 0))
+ vx_vsi_swapinfo(val);
}
/*