/*
|
/*
|
* linux/mm/vmscan.c
|
* linux/mm/vmscan.c
|
*
|
*
|
* Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds
|
* Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds
|
*
|
*
|
* Swap reorganised 29.12.95, Stephen Tweedie.
|
* Swap reorganised 29.12.95, Stephen Tweedie.
|
* kswapd added: 7.1.96 sct
|
* kswapd added: 7.1.96 sct
|
* Version: $Id: vmscan.c,v 1.1 2005-12-20 11:47:02 jcastillo Exp $
|
* Version: $Id: vmscan.c,v 1.1 2005-12-20 11:47:02 jcastillo Exp $
|
*/
|
*/
|
|
|
#include <linux/mm.h>
|
#include <linux/mm.h>
|
#include <linux/sched.h>
|
#include <linux/sched.h>
|
#include <linux/head.h>
|
#include <linux/head.h>
|
#include <linux/kernel.h>
|
#include <linux/kernel.h>
|
#include <linux/kernel_stat.h>
|
#include <linux/kernel_stat.h>
|
#include <linux/errno.h>
|
#include <linux/errno.h>
|
#include <linux/string.h>
|
#include <linux/string.h>
|
#include <linux/stat.h>
|
#include <linux/stat.h>
|
#include <linux/swap.h>
|
#include <linux/swap.h>
|
#include <linux/fs.h>
|
#include <linux/fs.h>
|
#include <linux/swapctl.h>
|
#include <linux/swapctl.h>
|
#include <linux/pagemap.h>
|
#include <linux/pagemap.h>
|
#include <linux/smp_lock.h>
|
#include <linux/smp_lock.h>
|
|
|
#include <asm/dma.h>
|
#include <asm/dma.h>
|
#include <asm/system.h> /* for cli()/sti() */
|
#include <asm/system.h> /* for cli()/sti() */
|
#include <asm/segment.h> /* for memcpy_to/fromfs */
|
#include <asm/segment.h> /* for memcpy_to/fromfs */
|
#include <asm/bitops.h>
|
#include <asm/bitops.h>
|
#include <asm/pgtable.h>
|
#include <asm/pgtable.h>
|
|
|
/*
|
/*
|
* To check memory consuming code elsewhere set this to 1
|
* To check memory consuming code elsewhere set this to 1
|
*/
|
*/
|
/* #define MM_DEBUG */
|
/* #define MM_DEBUG */
|
|
|
/*
|
/*
|
* When are we next due for a page scan?
|
* When are we next due for a page scan?
|
*/
|
*/
|
static int next_swap_jiffies = 0;
|
static int next_swap_jiffies = 0;
|
|
|
/*
|
/*
|
* Was the last kswapd wakeup caused by
|
* Was the last kswapd wakeup caused by
|
* nr_free_pages < free_pages_low
|
* nr_free_pages < free_pages_low
|
*/
|
*/
|
static int last_wakeup_low = 0;
|
static int last_wakeup_low = 0;
|
|
|
/*
|
/*
|
* How often do we do a pageout scan during normal conditions?
|
* How often do we do a pageout scan during normal conditions?
|
* Default is four times a second.
|
* Default is four times a second.
|
*/
|
*/
|
int swapout_interval = HZ / 4;
|
int swapout_interval = HZ / 4;
|
|
|
/*
|
/*
|
* The wait queue for waking up the pageout daemon:
|
* The wait queue for waking up the pageout daemon:
|
*/
|
*/
|
static struct wait_queue * kswapd_wait = NULL;
|
static struct wait_queue * kswapd_wait = NULL;
|
|
|
/*
|
/*
|
* We avoid doing a reschedule if the pageout daemon is already awake;
|
* We avoid doing a reschedule if the pageout daemon is already awake;
|
*/
|
*/
|
static int kswapd_awake = 0;
|
static int kswapd_awake = 0;
|
|
|
/*
|
/*
|
* sysctl-modifiable parameters to control the aggressiveness of the
|
* sysctl-modifiable parameters to control the aggressiveness of the
|
* page-searching within the kswapd page recovery daemon.
|
* page-searching within the kswapd page recovery daemon.
|
*/
|
*/
|
kswapd_control_t kswapd_ctl = {4, -1, -1, -1, -1};
|
kswapd_control_t kswapd_ctl = {4, -1, -1, -1, -1};
|
|
|
static void init_swap_timer(void);
|
static void init_swap_timer(void);
|
|
|
/*
|
/*
|
* The swap-out functions return 1 if they successfully
|
* The swap-out functions return 1 if they successfully
|
* threw something out, and we got a free page. It returns
|
* threw something out, and we got a free page. It returns
|
* zero if it couldn't do anything, and any other value
|
* zero if it couldn't do anything, and any other value
|
* indicates it decreased rss, but the page was shared.
|
* indicates it decreased rss, but the page was shared.
|
*
|
*
|
* NOTE! If it sleeps, it *must* return 1 to make sure we
|
* NOTE! If it sleeps, it *must* return 1 to make sure we
|
* don't continue with the swap-out. Otherwise we may be
|
* don't continue with the swap-out. Otherwise we may be
|
* using a process that no longer actually exists (it might
|
* using a process that no longer actually exists (it might
|
* have died while we slept).
|
* have died while we slept).
|
*/
|
*/
|
static inline int try_to_swap_out(struct task_struct * tsk, struct vm_area_struct* vma,
|
static inline int try_to_swap_out(struct task_struct * tsk, struct vm_area_struct* vma,
|
unsigned long address, pte_t * page_table, int dma, int wait, int can_do_io)
|
unsigned long address, pte_t * page_table, int dma, int wait, int can_do_io)
|
{
|
{
|
pte_t pte;
|
pte_t pte;
|
unsigned long entry;
|
unsigned long entry;
|
unsigned long page;
|
unsigned long page;
|
struct page * page_map;
|
struct page * page_map;
|
|
|
pte = *page_table;
|
pte = *page_table;
|
if (!pte_present(pte))
|
if (!pte_present(pte))
|
return 0;
|
return 0;
|
page = pte_page(pte);
|
page = pte_page(pte);
|
if (MAP_NR(page) >= MAP_NR(high_memory))
|
if (MAP_NR(page) >= MAP_NR(high_memory))
|
return 0;
|
return 0;
|
|
|
page_map = mem_map + MAP_NR(page);
|
page_map = mem_map + MAP_NR(page);
|
if (PageReserved(page_map)
|
if (PageReserved(page_map)
|
|| PageLocked(page_map)
|
|| PageLocked(page_map)
|
|| (dma && !PageDMA(page_map)))
|
|| (dma && !PageDMA(page_map)))
|
return 0;
|
return 0;
|
/* Deal with page aging. Pages age from being unused; they
|
/* Deal with page aging. Pages age from being unused; they
|
* rejuvenate on being accessed. Only swap old pages (age==0
|
* rejuvenate on being accessed. Only swap old pages (age==0
|
* is oldest). */
|
* is oldest). */
|
if ((pte_dirty(pte) && delete_from_swap_cache(MAP_NR(page)))
|
if ((pte_dirty(pte) && delete_from_swap_cache(MAP_NR(page)))
|
|| pte_young(pte)) {
|
|| pte_young(pte)) {
|
set_pte(page_table, pte_mkold(pte));
|
set_pte(page_table, pte_mkold(pte));
|
touch_page(page_map);
|
touch_page(page_map);
|
return 0;
|
return 0;
|
}
|
}
|
age_page(page_map);
|
age_page(page_map);
|
if (page_map->age)
|
if (page_map->age)
|
return 0;
|
return 0;
|
if (pte_dirty(pte)) {
|
if (pte_dirty(pte)) {
|
if(!can_do_io)
|
if(!can_do_io)
|
return 0;
|
return 0;
|
if (vma->vm_ops && vma->vm_ops->swapout) {
|
if (vma->vm_ops && vma->vm_ops->swapout) {
|
pid_t pid = tsk->pid;
|
pid_t pid = tsk->pid;
|
vma->vm_mm->rss--;
|
vma->vm_mm->rss--;
|
if (vma->vm_ops->swapout(vma, address - vma->vm_start + vma->vm_offset, page_table))
|
if (vma->vm_ops->swapout(vma, address - vma->vm_start + vma->vm_offset, page_table))
|
kill_proc(pid, SIGBUS, 1);
|
kill_proc(pid, SIGBUS, 1);
|
} else {
|
} else {
|
if (!(entry = get_swap_page()))
|
if (!(entry = get_swap_page()))
|
return 0;
|
return 0;
|
vma->vm_mm->rss--;
|
vma->vm_mm->rss--;
|
flush_cache_page(vma, address);
|
flush_cache_page(vma, address);
|
set_pte(page_table, __pte(entry));
|
set_pte(page_table, __pte(entry));
|
flush_tlb_page(vma, address);
|
flush_tlb_page(vma, address);
|
tsk->nswap++;
|
tsk->nswap++;
|
rw_swap_page(WRITE, entry, (char *) page, wait);
|
rw_swap_page(WRITE, entry, (char *) page, wait);
|
}
|
}
|
free_page(page);
|
free_page(page);
|
return 1; /* we slept: the process may not exist any more */
|
return 1; /* we slept: the process may not exist any more */
|
}
|
}
|
if ((entry = find_in_swap_cache(MAP_NR(page)))) {
|
if ((entry = find_in_swap_cache(MAP_NR(page)))) {
|
if (page_map->count != 1) {
|
if (page_map->count != 1) {
|
set_pte(page_table, pte_mkdirty(pte));
|
set_pte(page_table, pte_mkdirty(pte));
|
printk("Aiee.. duplicated cached swap-cache entry\n");
|
printk("Aiee.. duplicated cached swap-cache entry\n");
|
return 0;
|
return 0;
|
}
|
}
|
vma->vm_mm->rss--;
|
vma->vm_mm->rss--;
|
flush_cache_page(vma, address);
|
flush_cache_page(vma, address);
|
set_pte(page_table, __pte(entry));
|
set_pte(page_table, __pte(entry));
|
flush_tlb_page(vma, address);
|
flush_tlb_page(vma, address);
|
free_page(page);
|
free_page(page);
|
return 1;
|
return 1;
|
}
|
}
|
vma->vm_mm->rss--;
|
vma->vm_mm->rss--;
|
flush_cache_page(vma, address);
|
flush_cache_page(vma, address);
|
pte_clear(page_table);
|
pte_clear(page_table);
|
flush_tlb_page(vma, address);
|
flush_tlb_page(vma, address);
|
entry = page_unuse(page);
|
entry = page_unuse(page);
|
free_page(page);
|
free_page(page);
|
return entry;
|
return entry;
|
}
|
}
|
|
|
/*
|
/*
|
* A new implementation of swap_out(). We do not swap complete processes,
|
* A new implementation of swap_out(). We do not swap complete processes,
|
* but only a small number of blocks, before we continue with the next
|
* but only a small number of blocks, before we continue with the next
|
* process. The number of blocks actually swapped is determined on the
|
* process. The number of blocks actually swapped is determined on the
|
* number of page faults, that this process actually had in the last time,
|
* number of page faults, that this process actually had in the last time,
|
* so we won't swap heavily used processes all the time ...
|
* so we won't swap heavily used processes all the time ...
|
*
|
*
|
* Note: the priority argument is a hint on much CPU to waste with the
|
* Note: the priority argument is a hint on much CPU to waste with the
|
* swap block search, not a hint, of how much blocks to swap with
|
* swap block search, not a hint, of how much blocks to swap with
|
* each process.
|
* each process.
|
*
|
*
|
* (C) 1993 Kai Petzke, wpp@marie.physik.tu-berlin.de
|
* (C) 1993 Kai Petzke, wpp@marie.physik.tu-berlin.de
|
*/
|
*/
|
|
|
static inline int swap_out_pmd(struct task_struct * tsk, struct vm_area_struct * vma,
|
static inline int swap_out_pmd(struct task_struct * tsk, struct vm_area_struct * vma,
|
pmd_t *dir, unsigned long address, unsigned long end, int dma, int wait,
|
pmd_t *dir, unsigned long address, unsigned long end, int dma, int wait,
|
int can_do_io)
|
int can_do_io)
|
{
|
{
|
pte_t * pte;
|
pte_t * pte;
|
unsigned long pmd_end;
|
unsigned long pmd_end;
|
|
|
if (pmd_none(*dir))
|
if (pmd_none(*dir))
|
return 0;
|
return 0;
|
if (pmd_bad(*dir)) {
|
if (pmd_bad(*dir)) {
|
printk("swap_out_pmd: bad pmd (%08lx)\n", pmd_val(*dir));
|
printk("swap_out_pmd: bad pmd (%08lx)\n", pmd_val(*dir));
|
pmd_clear(dir);
|
pmd_clear(dir);
|
return 0;
|
return 0;
|
}
|
}
|
|
|
pte = pte_offset(dir, address);
|
pte = pte_offset(dir, address);
|
|
|
pmd_end = (address + PMD_SIZE) & PMD_MASK;
|
pmd_end = (address + PMD_SIZE) & PMD_MASK;
|
if (end > pmd_end)
|
if (end > pmd_end)
|
end = pmd_end;
|
end = pmd_end;
|
|
|
do {
|
do {
|
int result;
|
int result;
|
tsk->swap_address = address + PAGE_SIZE;
|
tsk->swap_address = address + PAGE_SIZE;
|
result = try_to_swap_out(tsk, vma, address, pte, dma, wait,
|
result = try_to_swap_out(tsk, vma, address, pte, dma, wait,
|
can_do_io);
|
can_do_io);
|
if (result)
|
if (result)
|
return result;
|
return result;
|
address += PAGE_SIZE;
|
address += PAGE_SIZE;
|
pte++;
|
pte++;
|
} while (address < end);
|
} while (address < end);
|
return 0;
|
return 0;
|
}
|
}
|
|
|
static inline int swap_out_pgd(struct task_struct * tsk, struct vm_area_struct * vma,
|
static inline int swap_out_pgd(struct task_struct * tsk, struct vm_area_struct * vma,
|
pgd_t *dir, unsigned long address, unsigned long end, int dma, int wait,
|
pgd_t *dir, unsigned long address, unsigned long end, int dma, int wait,
|
int can_do_io)
|
int can_do_io)
|
{
|
{
|
pmd_t * pmd;
|
pmd_t * pmd;
|
unsigned long pgd_end;
|
unsigned long pgd_end;
|
|
|
if (pgd_none(*dir))
|
if (pgd_none(*dir))
|
return 0;
|
return 0;
|
if (pgd_bad(*dir)) {
|
if (pgd_bad(*dir)) {
|
printk("swap_out_pgd: bad pgd (%08lx)\n", pgd_val(*dir));
|
printk("swap_out_pgd: bad pgd (%08lx)\n", pgd_val(*dir));
|
pgd_clear(dir);
|
pgd_clear(dir);
|
return 0;
|
return 0;
|
}
|
}
|
|
|
pmd = pmd_offset(dir, address);
|
pmd = pmd_offset(dir, address);
|
|
|
pgd_end = (address + PGDIR_SIZE) & PGDIR_MASK;
|
pgd_end = (address + PGDIR_SIZE) & PGDIR_MASK;
|
if (end > pgd_end)
|
if (end > pgd_end)
|
end = pgd_end;
|
end = pgd_end;
|
|
|
do {
|
do {
|
int result = swap_out_pmd(tsk, vma, pmd, address, end, dma, wait,
|
int result = swap_out_pmd(tsk, vma, pmd, address, end, dma, wait,
|
can_do_io);
|
can_do_io);
|
if (result)
|
if (result)
|
return result;
|
return result;
|
address = (address + PMD_SIZE) & PMD_MASK;
|
address = (address + PMD_SIZE) & PMD_MASK;
|
pmd++;
|
pmd++;
|
} while (address < end);
|
} while (address < end);
|
return 0;
|
return 0;
|
}
|
}
|
|
|
static int swap_out_vma(struct task_struct * tsk, struct vm_area_struct * vma,
|
static int swap_out_vma(struct task_struct * tsk, struct vm_area_struct * vma,
|
pgd_t *pgdir, unsigned long start, int dma, int wait, int can_do_io)
|
pgd_t *pgdir, unsigned long start, int dma, int wait, int can_do_io)
|
{
|
{
|
unsigned long end;
|
unsigned long end;
|
|
|
/* Don't swap out areas like shared memory which have their
|
/* Don't swap out areas like shared memory which have their
|
own separate swapping mechanism or areas which are locked down */
|
own separate swapping mechanism or areas which are locked down */
|
if (vma->vm_flags & (VM_SHM | VM_LOCKED))
|
if (vma->vm_flags & (VM_SHM | VM_LOCKED))
|
return 0;
|
return 0;
|
|
|
end = vma->vm_end;
|
end = vma->vm_end;
|
while (start < end) {
|
while (start < end) {
|
int result = swap_out_pgd(tsk, vma, pgdir, start, end, dma, wait,
|
int result = swap_out_pgd(tsk, vma, pgdir, start, end, dma, wait,
|
can_do_io);
|
can_do_io);
|
if (result)
|
if (result)
|
return result;
|
return result;
|
start = (start + PGDIR_SIZE) & PGDIR_MASK;
|
start = (start + PGDIR_SIZE) & PGDIR_MASK;
|
pgdir++;
|
pgdir++;
|
}
|
}
|
return 0;
|
return 0;
|
}
|
}
|
|
|
static int swap_out_process(struct task_struct * p, int dma, int wait, int can_do_io)
|
static int swap_out_process(struct task_struct * p, int dma, int wait, int can_do_io)
|
{
|
{
|
unsigned long address;
|
unsigned long address;
|
struct vm_area_struct* vma;
|
struct vm_area_struct* vma;
|
|
|
/*
|
/*
|
* Go through process' page directory.
|
* Go through process' page directory.
|
*/
|
*/
|
address = p->swap_address;
|
address = p->swap_address;
|
|
|
/*
|
/*
|
* Find the proper vm-area
|
* Find the proper vm-area
|
*/
|
*/
|
vma = find_vma(p->mm, address);
|
vma = find_vma(p->mm, address);
|
if (!vma) {
|
if (!vma) {
|
p->swap_address = 0;
|
p->swap_address = 0;
|
return 0;
|
return 0;
|
}
|
}
|
if (address < vma->vm_start)
|
if (address < vma->vm_start)
|
address = vma->vm_start;
|
address = vma->vm_start;
|
|
|
for (;;) {
|
for (;;) {
|
int result = swap_out_vma(p, vma, pgd_offset(p->mm, address), address, dma, wait,
|
int result = swap_out_vma(p, vma, pgd_offset(p->mm, address), address, dma, wait,
|
can_do_io);
|
can_do_io);
|
if (result)
|
if (result)
|
return result;
|
return result;
|
vma = vma->vm_next;
|
vma = vma->vm_next;
|
if (!vma)
|
if (!vma)
|
break;
|
break;
|
address = vma->vm_start;
|
address = vma->vm_start;
|
}
|
}
|
p->swap_address = 0;
|
p->swap_address = 0;
|
return 0;
|
return 0;
|
}
|
}
|
|
|
static int swap_out(unsigned int priority, int dma, int wait, int can_do_io)
|
static int swap_out(unsigned int priority, int dma, int wait, int can_do_io)
|
{
|
{
|
static int swap_task;
|
static int swap_task;
|
int loop, counter, shfrv;
|
int loop, counter, shfrv;
|
struct task_struct *p;
|
struct task_struct *p;
|
|
|
#ifdef MM_DEBUG
|
#ifdef MM_DEBUG
|
shfrv = 10;
|
shfrv = 10;
|
#else
|
#else
|
/*
|
/*
|
* Trouble due ageing pages: In some situations it is possible that we cross only tasks
|
* Trouble due ageing pages: In some situations it is possible that we cross only tasks
|
* which are swapped out or which have only physical pages with age >= 3.
|
* which are swapped out or which have only physical pages with age >= 3.
|
* High values of swap_cnt for memory consuming tasks do aggravate such situations.
|
* High values of swap_cnt for memory consuming tasks do aggravate such situations.
|
*
|
*
|
* If PAGEOUT_WEIGHT has a value of 8192 a right shift value of 10 leads to
|
* If PAGEOUT_WEIGHT has a value of 8192 a right shift value of 10 leads to
|
* (8 * nr_tasks) >> priority
|
* (8 * nr_tasks) >> priority
|
* Together with a high number of tasks, say 100, we have counters (due priority)
|
* Together with a high number of tasks, say 100, we have counters (due priority)
|
* 12(6) + 25(5) + 50(4) + 100(3) + 200(2) + 400(1) + 800(0)
|
* 12(6) + 25(5) + 50(4) + 100(3) + 200(2) + 400(1) + 800(0)
|
* and as total result 1587 scans of swap_out() to swap out a task page.
|
* and as total result 1587 scans of swap_out() to swap out a task page.
|
*
|
*
|
* Just assume 80 tasks are swapped out and the remaining tasks have a swap_cnt value >= 40
|
* Just assume 80 tasks are swapped out and the remaining tasks have a swap_cnt value >= 40
|
* together with pages with age >= 3. Then we need approx 20*40*2 = 1600 scans to get a
|
* together with pages with age >= 3. Then we need approx 20*40*2 = 1600 scans to get a
|
* free page.
|
* free page.
|
* And now assume that the amount of cached pages, buffers, and ipc pages are really low.
|
* And now assume that the amount of cached pages, buffers, and ipc pages are really low.
|
*/
|
*/
|
switch (priority) {
|
switch (priority) {
|
case 6: case 5: case 4: /* be friendly */
|
case 6: case 5: case 4: /* be friendly */
|
shfrv = 10;
|
shfrv = 10;
|
break;
|
break;
|
case 3: case 2: case 1: /* more intensive */
|
case 3: case 2: case 1: /* more intensive */
|
shfrv = 9;
|
shfrv = 9;
|
break;
|
break;
|
case 0: default: /* sorry we need a page */
|
case 0: default: /* sorry we need a page */
|
shfrv = 8;
|
shfrv = 8;
|
break;
|
break;
|
}
|
}
|
/*
|
/*
|
* kswapd should be more friendly to other processes.
|
* kswapd should be more friendly to other processes.
|
*/
|
*/
|
if (kswapd_awake)
|
if (kswapd_awake)
|
shfrv = 10;
|
shfrv = 10;
|
#endif
|
#endif
|
|
|
counter = ((PAGEOUT_WEIGHT * nr_tasks) >> shfrv) >> priority;
|
counter = ((PAGEOUT_WEIGHT * nr_tasks) >> shfrv) >> priority;
|
for(; counter >= 0; counter--) {
|
for(; counter >= 0; counter--) {
|
/*
|
/*
|
* Check that swap_task is suitable for swapping. If not, look for
|
* Check that swap_task is suitable for swapping. If not, look for
|
* the next suitable process.
|
* the next suitable process.
|
*/
|
*/
|
loop = 0;
|
loop = 0;
|
while(1) {
|
while(1) {
|
if (swap_task >= NR_TASKS) {
|
if (swap_task >= NR_TASKS) {
|
swap_task = 1;
|
swap_task = 1;
|
if (loop)
|
if (loop)
|
/* all processes are unswappable or already swapped out */
|
/* all processes are unswappable or already swapped out */
|
return 0;
|
return 0;
|
loop = 1;
|
loop = 1;
|
}
|
}
|
|
|
p = task[swap_task];
|
p = task[swap_task];
|
if (p && p->swappable && p->mm->rss)
|
if (p && p->swappable && p->mm->rss)
|
break;
|
break;
|
|
|
swap_task++;
|
swap_task++;
|
}
|
}
|
|
|
/*
|
/*
|
* Determine the number of pages to swap from this process.
|
* Determine the number of pages to swap from this process.
|
*/
|
*/
|
if (!p->swap_cnt) {
|
if (!p->swap_cnt) {
|
/*
|
/*
|
* Normalise the number of pages swapped by
|
* Normalise the number of pages swapped by
|
* multiplying by (RSS / 1MB)
|
* multiplying by (RSS / 1MB)
|
*/
|
*/
|
p->swap_cnt = AGE_CLUSTER_SIZE(p->mm->rss);
|
p->swap_cnt = AGE_CLUSTER_SIZE(p->mm->rss);
|
}
|
}
|
if (!--p->swap_cnt)
|
if (!--p->swap_cnt)
|
swap_task++;
|
swap_task++;
|
switch (swap_out_process(p, dma, wait, can_do_io)) {
|
switch (swap_out_process(p, dma, wait, can_do_io)) {
|
case 0:
|
case 0:
|
if (p->state == TASK_STOPPED)
|
if (p->state == TASK_STOPPED)
|
/* Stopped task occupy nonused ram */
|
/* Stopped task occupy nonused ram */
|
break;
|
break;
|
if (p->swap_cnt)
|
if (p->swap_cnt)
|
swap_task++;
|
swap_task++;
|
break;
|
break;
|
case 1:
|
case 1:
|
return 1;
|
return 1;
|
default:
|
default:
|
break;
|
break;
|
}
|
}
|
}
|
}
|
#ifdef MM_DEBUG
|
#ifdef MM_DEBUG
|
if (!priority) {
|
if (!priority) {
|
printk("swap_out: physical ram %6dkB, min pages %6dkB\n",
|
printk("swap_out: physical ram %6dkB, min pages %6dkB\n",
|
(int)(high_memory>>10), min_free_pages<<(PAGE_SHIFT-10));
|
(int)(high_memory>>10), min_free_pages<<(PAGE_SHIFT-10));
|
printk("swap_out: free pages %6dkB, async pages %6dkB\n",
|
printk("swap_out: free pages %6dkB, async pages %6dkB\n",
|
nr_free_pages<<(PAGE_SHIFT-10), nr_async_pages<<(PAGE_SHIFT-10));
|
nr_free_pages<<(PAGE_SHIFT-10), nr_async_pages<<(PAGE_SHIFT-10));
|
}
|
}
|
#endif
|
#endif
|
return 0;
|
return 0;
|
}
|
}
|
|
|
/*
|
/*
|
* We are much more aggressive about trying to swap out than we used
|
* We are much more aggressive about trying to swap out than we used
|
* to be. This works out OK, because we now do proper aging on page
|
* to be. This works out OK, because we now do proper aging on page
|
* contents.
|
* contents.
|
*/
|
*/
|
int try_to_free_page(int priority, int dma, int wait)
|
int try_to_free_page(int priority, int dma, int wait)
|
{
|
{
|
static int state = 0;
|
static int state = 0;
|
int i=6;
|
int i=6;
|
int stop, can_do_io;
|
int stop, can_do_io;
|
|
|
/* we don't try as hard if we're not waiting.. */
|
/* we don't try as hard if we're not waiting.. */
|
stop = 3;
|
stop = 3;
|
can_do_io = 1;
|
can_do_io = 1;
|
if (wait)
|
if (wait)
|
stop = 0;
|
stop = 0;
|
if (priority == GFP_IO)
|
if (priority == GFP_IO)
|
can_do_io = 0;
|
can_do_io = 0;
|
switch (state) {
|
switch (state) {
|
do {
|
do {
|
case 0:
|
case 0:
|
if (shrink_mmap(i, dma, 1))
|
if (shrink_mmap(i, dma, 1))
|
return 1;
|
return 1;
|
state = 1;
|
state = 1;
|
case 1:
|
case 1:
|
if (can_do_io && shm_swap(i, dma))
|
if (can_do_io && shm_swap(i, dma))
|
return 1;
|
return 1;
|
state = 2;
|
state = 2;
|
default:
|
default:
|
if (swap_out(i, dma, wait, can_do_io))
|
if (swap_out(i, dma, wait, can_do_io))
|
return 1;
|
return 1;
|
state = 0;
|
state = 0;
|
i--;
|
i--;
|
} while ((i - stop) >= 0);
|
} while ((i - stop) >= 0);
|
}
|
}
|
return 0;
|
return 0;
|
}
|
}
|
|
|
/*
|
/*
|
* Before we start the kernel thread, print out the
|
* Before we start the kernel thread, print out the
|
* kswapd initialization message (otherwise the init message
|
* kswapd initialization message (otherwise the init message
|
* may be printed in the middle of another driver's init
|
* may be printed in the middle of another driver's init
|
* message). It looks very bad when that happens.
|
* message). It looks very bad when that happens.
|
*/
|
*/
|
void kswapd_setup(void)
|
void kswapd_setup(void)
|
{
|
{
|
int i;
|
int i;
|
char *revision="$Revision: 1.1 $", *s, *e;
|
char *revision="$Revision: 1.1 $", *s, *e;
|
|
|
if ((s = strchr(revision, ':')) &&
|
if ((s = strchr(revision, ':')) &&
|
(e = strchr(s, '$')))
|
(e = strchr(s, '$')))
|
s++, i = e - s;
|
s++, i = e - s;
|
else
|
else
|
s = revision, i = -1;
|
s = revision, i = -1;
|
printk ("Starting kswapd v%.*s\n", i, s);
|
printk ("Starting kswapd v%.*s\n", i, s);
|
}
|
}
|
|
|
/*
|
/*
|
* The background pageout daemon.
|
* The background pageout daemon.
|
* Started as a kernel thread from the init process.
|
* Started as a kernel thread from the init process.
|
*/
|
*/
|
int kswapd(void *unused)
|
int kswapd(void *unused)
|
{
|
{
|
int i, reserved_pages;
|
int i, reserved_pages;
|
|
|
current->session = 1;
|
current->session = 1;
|
current->pgrp = 1;
|
current->pgrp = 1;
|
sprintf(current->comm, "kswapd");
|
sprintf(current->comm, "kswapd");
|
current->blocked = ~0UL;
|
current->blocked = ~0UL;
|
|
|
/*
|
/*
|
* As a kernel thread we want to tamper with system buffers
|
* As a kernel thread we want to tamper with system buffers
|
* and other internals and thus be subject to the SMP locking
|
* and other internals and thus be subject to the SMP locking
|
* rules. (On a uniprocessor box this does nothing).
|
* rules. (On a uniprocessor box this does nothing).
|
*/
|
*/
|
|
|
#ifdef __SMP__
|
#ifdef __SMP__
|
lock_kernel();
|
lock_kernel();
|
syscall_count++;
|
syscall_count++;
|
#endif
|
#endif
|
|
|
/* Give kswapd a realtime priority. */
|
/* Give kswapd a realtime priority. */
|
current->policy = SCHED_FIFO;
|
current->policy = SCHED_FIFO;
|
current->priority = 32; /* Fixme --- we need to standardise our
|
current->priority = 32; /* Fixme --- we need to standardise our
|
namings for POSIX.4 realtime scheduling
|
namings for POSIX.4 realtime scheduling
|
priorities. */
|
priorities. */
|
|
|
init_swap_timer();
|
init_swap_timer();
|
|
|
while (1) {
|
while (1) {
|
/* low on memory, we need to start swapping soon */
|
/* low on memory, we need to start swapping soon */
|
next_swap_jiffies = jiffies +
|
next_swap_jiffies = jiffies +
|
(last_wakeup_low ? swapout_interval >> 1 : swapout_interval);
|
(last_wakeup_low ? swapout_interval >> 1 : swapout_interval);
|
kswapd_awake = 0;
|
kswapd_awake = 0;
|
current->signal = 0;
|
current->signal = 0;
|
run_task_queue(&tq_disk);
|
run_task_queue(&tq_disk);
|
interruptible_sleep_on(&kswapd_wait);
|
interruptible_sleep_on(&kswapd_wait);
|
kswapd_awake = 1;
|
kswapd_awake = 1;
|
swapstats.wakeups++;
|
swapstats.wakeups++;
|
/* Protect our reserved pages: */
|
/* Protect our reserved pages: */
|
i = 0;
|
i = 0;
|
reserved_pages = min_free_pages;
|
reserved_pages = min_free_pages;
|
if (min_free_pages >= 48)
|
if (min_free_pages >= 48)
|
reserved_pages -= (12 + (reserved_pages>>3));
|
reserved_pages -= (12 + (reserved_pages>>3));
|
if (nr_free_pages <= reserved_pages)
|
if (nr_free_pages <= reserved_pages)
|
i = (1+reserved_pages) - nr_free_pages;
|
i = (1+reserved_pages) - nr_free_pages;
|
/* Do the background pageout: */
|
/* Do the background pageout: */
|
for (i += kswapd_ctl.maxpages; i > 0; i--)
|
for (i += kswapd_ctl.maxpages; i > 0; i--)
|
try_to_free_page(GFP_KERNEL, 0,
|
try_to_free_page(GFP_KERNEL, 0,
|
(nr_free_pages <= min_free_pages));
|
(nr_free_pages <= min_free_pages));
|
}
|
}
|
}
|
}
|
|
|
/*
|
/*
|
* The swap_tick function gets called on every clock tick.
|
* The swap_tick function gets called on every clock tick.
|
*/
|
*/
|
|
|
void swap_tick(void)
|
void swap_tick(void)
|
{
|
{
|
int want_wakeup = 0;
|
int want_wakeup = 0;
|
|
|
if ((nr_free_pages + nr_async_pages) < free_pages_low) {
|
if ((nr_free_pages + nr_async_pages) < free_pages_low) {
|
if (last_wakeup_low)
|
if (last_wakeup_low)
|
want_wakeup = (jiffies >= next_swap_jiffies);
|
want_wakeup = (jiffies >= next_swap_jiffies);
|
else
|
else
|
last_wakeup_low = want_wakeup = 1;
|
last_wakeup_low = want_wakeup = 1;
|
}
|
}
|
else if (((nr_free_pages + nr_async_pages) < free_pages_high) &&
|
else if (((nr_free_pages + nr_async_pages) < free_pages_high) &&
|
(jiffies >= next_swap_jiffies)) {
|
(jiffies >= next_swap_jiffies)) {
|
last_wakeup_low = 0;
|
last_wakeup_low = 0;
|
want_wakeup = 1;
|
want_wakeup = 1;
|
}
|
}
|
|
|
if (want_wakeup) {
|
if (want_wakeup) {
|
if (!kswapd_awake && kswapd_ctl.maxpages > 0) {
|
if (!kswapd_awake && kswapd_ctl.maxpages > 0) {
|
wake_up(&kswapd_wait);
|
wake_up(&kswapd_wait);
|
need_resched = 1;
|
need_resched = 1;
|
}
|
}
|
}
|
}
|
timer_active |= (1<<SWAP_TIMER);
|
timer_active |= (1<<SWAP_TIMER);
|
}
|
}
|
|
|
|
|
/*
|
/*
|
* Initialise the swap timer
|
* Initialise the swap timer
|
*/
|
*/
|
|
|
void init_swap_timer(void)
|
void init_swap_timer(void)
|
{
|
{
|
timer_table[SWAP_TIMER].expires = 0;
|
timer_table[SWAP_TIMER].expires = 0;
|
timer_table[SWAP_TIMER].fn = swap_tick;
|
timer_table[SWAP_TIMER].fn = swap_tick;
|
timer_active |= (1<<SWAP_TIMER);
|
timer_active |= (1<<SWAP_TIMER);
|
}
|
}
|
|
|
